NEWS
author wenzelm
Tue Aug 07 20:19:48 2007 +0200 (2007-08-07)
changeset 24172 06e42cf7df4e
parent 24110 4ab3084e311c
child 24187 8bdf5ca5871f
permissions -rw-r--r--
theory loader: added use_thys, removed obsolete update_thy;
various global ML references of Pure and HOL have been turned into configuration options;
     1 Isabelle NEWS -- history user-relevant changes
     2 ==============================================
     3 
     4 New in this Isabelle version
     5 ----------------------------
     6 
     7 *** General ***
     8 
     9 * More uniform information about legacy features, notably a
    10 warning/error of "Legacy feature: ...", depending on the state of the
    11 tolerate_legacy_features flag (default true). FUTURE INCOMPATIBILITY:
    12 legacy features will disappear eventually.
    13 
    14 * Theory syntax: the header format ``theory A = B + C:'' has been
    15 discontinued in favour of ``theory A imports B C begin''.  Use isatool
    16 fixheaders to convert existing theory files.  INCOMPATIBILITY.
    17 
    18 * Theory syntax: the old non-Isar theory file format has been
    19 discontinued altogether.  Note that ML proof scripts may still be used
    20 with Isar theories; migration is usually quite simple with the ML
    21 function use_legacy_bindings.  INCOMPATIBILITY.
    22 
    23 * Theory syntax: some popular names (e.g. 'class', 'declaration',
    24 'fun', 'help', 'if') are now keywords.  INCOMPATIBILITY, use double
    25 quotes.
    26 
    27 * Theory loader: be more serious about observing the static theory
    28 header specifications (including optional directories), but not the
    29 accidental file locations of previously successful loads.  The strict
    30 update policy of former update_thy is now already performed by
    31 use_thy, so the former has been removed; use_thys updates several
    32 theories simultaneously, just as 'imports' within a theory header
    33 specification, but without merging the results.  Potential
    34 INCOMPATIBILITY: may need to refine theory headers and commands
    35 ROOT.ML which depend on load order.
    36 
    37 * Theory loader: optional support for content-based file
    38 identification, instead of the traditional scheme of full physical
    39 path plus date stamp; configured by the ISABELLE_FILE_IDENT setting
    40 (cf. the system manual).  The new scheme allows to work with
    41 non-finished theories in persistent session images, such that source
    42 files may be moved later on without requiring reloads.
    43 
    44 * Legacy goal package: reduced interface to the bare minimum required
    45 to keep existing proof scripts running.  Most other user-level
    46 functions are now part of the OldGoals structure, which is *not* open
    47 by default (consider isatool expandshort before open OldGoals).
    48 Removed top_sg, prin, printyp, pprint_term/typ altogether, because
    49 these tend to cause confusion about the actual goal (!) context being
    50 used here, which is not necessarily the same as the_context().
    51 
    52 * Command 'find_theorems': supports "*" wild-card in "name:"
    53 criterion; "with_dups" option.  Certain ProofGeneral versions might
    54 support a specific search form (see ProofGeneral/CHANGES).
    55 
    56 * The ``prems limit'' option (cf. ProofContext.prems_limit) is now -1
    57 by default, which means that "prems" (and also "fixed variables") are
    58 suppressed from proof state output.  Note that the ProofGeneral
    59 settings mechanism allows to change and save options persistently, but
    60 older versions of Isabelle will fail to start up if a negative prems
    61 limit is imposed.
    62 
    63 * Local theory targets may be specified by non-nested blocks of
    64 ``context/locale/class ... begin'' followed by ``end''.  The body may
    65 contain definitions, theorems etc., including any derived mechanism
    66 that has been implemented on top of these primitives.  This concept
    67 generalizes the existing ``theorem (in ...)'' towards more versatility
    68 and scalability.
    69 
    70 * Proof General interface: proper undo of final 'end' command;
    71 discontinued Isabelle/classic mode (ML proof scripts).
    72 
    73 
    74 *** Document preparation ***
    75 
    76 * Added antiquotation @{theory name} which prints the given name,
    77 after checking that it refers to a valid ancestor theory in the
    78 current context.
    79 
    80 * Added antiquotations @{ML_type text} and @{ML_struct text} which
    81 check the given source text as ML type/structure, printing verbatim.
    82 
    83 * Added antiquotation @{abbrev "c args"} which prints the abbreviation
    84 "c args == rhs" given in the current context.  (Any number of
    85 arguments may be given on the LHS.)
    86 
    87 
    88 
    89 *** Pure ***
    90 
    91 * code generator: consts in 'consts_code' Isar commands are now referred
    92   to by usual term syntax (including optional type annotations).
    93 
    94 * code generator: 
    95   - Isar 'definition's, 'constdef's and primitive instance definitions are added
    96     explicitly to the table of defining equations
    97   - primitive definitions are not used as defining equations by default any longer
    98   - defining equations are now definitly restricted to meta "==" and object
    99         equality "="
   100   - HOL theories have been adopted accordingly
   101 
   102 * class_package.ML offers a combination of axclasses and locales to
   103 achieve Haskell-like type classes in Isabelle.  See
   104 HOL/ex/Classpackage.thy for examples.
   105 
   106 * Yet another code generator framework allows to generate executable
   107 code for ML and Haskell (including "class"es).  A short usage sketch:
   108 
   109     internal compilation:
   110         code_gen <list of constants (term syntax)> in SML
   111     writing SML code to a file:
   112         code_gen <list of constants (term syntax)> in SML <filename>
   113     writing OCaml code to a file:
   114         code_gen <list of constants (term syntax)> in OCaml <filename>
   115     writing Haskell code to a bunch of files:
   116         code_gen <list of constants (term syntax)> in Haskell <filename>
   117 
   118 Reasonable default setup of framework in HOL/Main.
   119 
   120 Theorem attributs for selecting and transforming function equations theorems:
   121 
   122     [code fun]:        select a theorem as function equation for a specific constant
   123     [code fun del]:    deselect a theorem as function equation for a specific constant
   124     [code inline]:     select an equation theorem for unfolding (inlining) in place
   125     [code inline del]: deselect an equation theorem for unfolding (inlining) in place
   126 
   127 User-defined serializations (target in {SML, OCaml, Haskell}):
   128 
   129     code_const <and-list of constants (term syntax)>
   130       {(target) <and-list of const target syntax>}+
   131 
   132     code_type <and-list of type constructors>
   133       {(target) <and-list of type target syntax>}+
   134 
   135     code_instance <and-list of instances>
   136       {(target)}+
   137         where instance ::= <type constructor> :: <class>
   138 
   139     code_class <and_list of classes>
   140       {(target) <and-list of class target syntax>}+
   141         where class target syntax ::= <class name> {where {<classop> == <target syntax>}+}?
   142 
   143 code_instance and code_class only apply to target Haskell.
   144 
   145 See HOL theories and HOL/ex/Codegenerator*.thy for usage examples.
   146 Doc/Isar/Advanced/Codegen/ provides a tutorial.
   147 
   148 * Command 'no_translations' removes translation rules from theory
   149 syntax.
   150 
   151 * Overloaded definitions are now actually checked for acyclic
   152 dependencies.  The overloading scheme is slightly more general than
   153 that of Haskell98, although Isabelle does not demand an exact
   154 correspondence to type class and instance declarations.
   155 INCOMPATIBILITY, use ``defs (unchecked overloaded)'' to admit more
   156 exotic versions of overloading -- at the discretion of the user!
   157 
   158 Polymorphic constants are represented via type arguments, i.e. the
   159 instantiation that matches an instance against the most general
   160 declaration given in the signature.  For example, with the declaration
   161 c :: 'a => 'a => 'a, an instance c :: nat => nat => nat is represented
   162 as c(nat).  Overloading is essentially simultaneous structural
   163 recursion over such type arguments.  Incomplete specification patterns
   164 impose global constraints on all occurrences, e.g. c('a * 'a) on the
   165 LHS means that more general c('a * 'b) will be disallowed on any RHS.
   166 Command 'print_theory' outputs the normalized system of recursive
   167 equations, see section "definitions".
   168 
   169 * Configuration options are maintained within the theory or proof
   170 context (with name and type bool/int/string), providing a very simple
   171 interface to a poor-man's version of general context data.  Tools may
   172 declare options in ML (e.g. using Attrib.config_int) and then refer to
   173 these values using Config.get etc.  Users may change options via an
   174 associated attribute of the same name.  This form of context
   175 declaration works particularly well with commands 'declare' or
   176 'using', for example ``declare [[foo = 42]]''.  Thus it has become
   177 very easy to avoid global references, which would not observe Isar
   178 toplevel undo/redo and fail to work with multithreading.
   179 
   180 Various global ML references of Pure and HOL have been turned into
   181 configuration options:
   182 
   183   Unify.search_bound		unify_search_bound
   184   Unify.trace_bound		unify_trace_bound
   185   Unify.trace_simp		unify_trace_simp
   186   Unify.trace_types		unify_trace_types
   187   Simplifier.simp_depth_limit	simp_depth_limit
   188   Blast.depth_limit		blast_depth_limit
   189   DatatypeProp.dtK		datatype_distinctness_limit
   190   fast_arith_neq_limit  	fast_arith_neq_limit
   191   fast_arith_split_limit	fast_arith_split_limit
   192 
   193 * Named collections of theorems may be easily installed as context
   194 data using the functor NamedThmsFun (see
   195 src/Pure/Tools/named_thms.ML).  The user may add or delete facts via
   196 attributes; there is also a toplevel print command.  This facility is
   197 just a common case of general context data, which is the preferred way
   198 for anything more complex than just a list of facts in canonical
   199 order.
   200 
   201 * Isar: command 'declaration' augments a local theory by generic
   202 declaration functions written in ML.  This enables arbitrary content
   203 being added to the context, depending on a morphism that tells the
   204 difference of the original declaration context wrt. the application
   205 context encountered later on.
   206 
   207 * Isar: proper interfaces for simplification procedures.  Command
   208 'simproc_setup' declares named simprocs (with match patterns, and body
   209 text in ML).  Attribute "simproc" adds/deletes simprocs in the current
   210 context.  ML antiquotation @{simproc name} retrieves named simprocs.
   211 
   212 * Isar: an extra pair of brackets around attribute declarations
   213 abbreviates a theorem reference involving an internal dummy fact,
   214 which will be ignored later --- only the effect of the attribute on
   215 the background context will persist.  This form of in-place
   216 declarations is particularly useful with commands like 'declare' and
   217 'using', for example ``have A using [[simproc a]] by simp''.
   218 
   219 * Isar: method "assumption" (and implicit closing of subproofs) now
   220 takes simple non-atomic goal assumptions into account: after applying
   221 an assumption as a rule the resulting subgoals are solved by atomic
   222 assumption steps.  This is particularly useful to finish 'obtain'
   223 goals, such as "!!x. (!!x. P x ==> thesis) ==> P x ==> thesis",
   224 without referring to the original premise "!!x. P x ==> thesis" in the
   225 Isar proof context.  POTENTIAL INCOMPATIBILITY: method "assumption" is
   226 more permissive.
   227 
   228 * Isar: implicit use of prems from the Isar proof context is
   229 considered a legacy feature.  Common applications like ``have A .''
   230 may be replaced by ``have A by fact'' or ``note `A`''.  In general,
   231 referencing facts explicitly here improves readability and
   232 maintainability of proof texts.
   233 
   234 * Isar: improper proof element 'guess' is like 'obtain', but derives
   235 the obtained context from the course of reasoning!  For example:
   236 
   237   assume "EX x y. A x & B y"   -- "any previous fact"
   238   then guess x and y by clarify
   239 
   240 This technique is potentially adventurous, depending on the facts and
   241 proof tools being involved here.
   242 
   243 * Isar: known facts from the proof context may be specified as literal
   244 propositions, using ASCII back-quote syntax.  This works wherever
   245 named facts used to be allowed so far, in proof commands, proof
   246 methods, attributes etc.  Literal facts are retrieved from the context
   247 according to unification of type and term parameters.  For example,
   248 provided that "A" and "A ==> B" and "!!x. P x ==> Q x" are known
   249 theorems in the current context, then these are valid literal facts:
   250 `A` and `A ==> B` and `!!x. P x ==> Q x" as well as `P a ==> Q a` etc.
   251 
   252 There is also a proof method "fact" which does the same composition
   253 for explicit goal states, e.g. the following proof texts coincide with
   254 certain special cases of literal facts:
   255 
   256   have "A" by fact                 ==  note `A`
   257   have "A ==> B" by fact           ==  note `A ==> B`
   258   have "!!x. P x ==> Q x" by fact  ==  note `!!x. P x ==> Q x`
   259   have "P a ==> Q a" by fact       ==  note `P a ==> Q a`
   260 
   261 * Isar: ":" (colon) is no longer a symbolic identifier character in
   262 outer syntax.  Thus symbolic identifiers may be used without
   263 additional white space in declarations like this: ``assume *: A''.
   264 
   265 * Isar: 'print_facts' prints all local facts of the current context,
   266 both named and unnamed ones.
   267 
   268 * Isar: 'def' now admits simultaneous definitions, e.g.:
   269 
   270   def x == "t" and y == "u"
   271 
   272 * Isar: added command 'unfolding', which is structurally similar to
   273 'using', but affects both the goal state and facts by unfolding given
   274 rewrite rules.  Thus many occurrences of the 'unfold' method or
   275 'unfolded' attribute may be replaced by first-class proof text.
   276 
   277 * Isar: methods 'unfold' / 'fold', attributes 'unfolded' / 'folded',
   278 and command 'unfolding' now all support object-level equalities
   279 (potentially conditional).  The underlying notion of rewrite rule is
   280 analogous to the 'rule_format' attribute, but *not* that of the
   281 Simplifier (which is usually more generous).
   282 
   283 * Isar: the goal restriction operator [N] (default N = 1) evaluates a
   284 method expression within a sandbox consisting of the first N
   285 sub-goals, which need to exist.  For example, ``simp_all [3]''
   286 simplifies the first three sub-goals, while (rule foo, simp_all)[]
   287 simplifies all new goals that emerge from applying rule foo to the
   288 originally first one.
   289 
   290 * Isar: schematic goals are no longer restricted to higher-order
   291 patterns; e.g. ``lemma "?P(?x)" by (rule TrueI)'' now works as
   292 expected.
   293 
   294 * Isar: the conclusion of a long theorem statement is now either
   295 'shows' (a simultaneous conjunction, as before), or 'obtains'
   296 (essentially a disjunction of cases with local parameters and
   297 assumptions).  The latter allows to express general elimination rules
   298 adequately; in this notation common elimination rules look like this:
   299 
   300   lemma exE:    -- "EX x. P x ==> (!!x. P x ==> thesis) ==> thesis"
   301     assumes "EX x. P x"
   302     obtains x where "P x"
   303 
   304   lemma conjE:  -- "A & B ==> (A ==> B ==> thesis) ==> thesis"
   305     assumes "A & B"
   306     obtains A and B
   307 
   308   lemma disjE:  -- "A | B ==> (A ==> thesis) ==> (B ==> thesis) ==> thesis"
   309     assumes "A | B"
   310     obtains
   311       A
   312     | B
   313 
   314 The subsequent classical rules even refer to the formal "thesis"
   315 explicitly:
   316 
   317   lemma classical:     -- "(~ thesis ==> thesis) ==> thesis"
   318     obtains "~ thesis"
   319 
   320   lemma Peirce's_Law:  -- "((thesis ==> something) ==> thesis) ==> thesis"
   321     obtains "thesis ==> something"
   322 
   323 The actual proof of an 'obtains' statement is analogous to that of the
   324 Isar proof element 'obtain', only that there may be several cases.
   325 Optional case names may be specified in parentheses; these will be
   326 available both in the present proof and as annotations in the
   327 resulting rule, for later use with the 'cases' method (cf. attribute
   328 case_names).
   329 
   330 * Isar: the assumptions of a long theorem statement are available as
   331 "assms" fact in the proof context.  This is more appropriate than the
   332 (historical) "prems", which refers to all assumptions of the current
   333 context, including those from the target locale, proof body etc.
   334 
   335 * Isar: 'print_statement' prints theorems from the current theory or
   336 proof context in long statement form, according to the syntax of a
   337 top-level lemma.
   338 
   339 * Isar: 'obtain' takes an optional case name for the local context
   340 introduction rule (default "that").
   341 
   342 * Isar: removed obsolete 'concl is' patterns.  INCOMPATIBILITY, use
   343 explicit (is "_ ==> ?foo") in the rare cases where this still happens
   344 to occur.
   345 
   346 * Pure: syntax "CONST name" produces a fully internalized constant
   347 according to the current context.  This is particularly useful for
   348 syntax translations that should refer to internal constant
   349 representations independently of name spaces.
   350 
   351 * Pure: syntax constant for foo (binder "FOO ") is called "foo_binder"
   352 instead of "FOO ". This allows multiple binder declarations to coexist
   353 in the same context.  INCOMPATIBILITY.
   354 
   355 * Isar/locales: 'notation' provides a robust interface to the 'syntax'
   356 primitive that also works in a locale context (both for constants and
   357 fixed variables).  Type declaration and internal syntactic
   358 representation of given constants retrieved from the context.
   359 
   360 * Isar/locales: new derived specification elements 'axiomatization',
   361 'definition', 'abbreviation', which support type-inference, admit
   362 object-level specifications (equality, equivalence).  See also the
   363 isar-ref manual.  Examples:
   364 
   365   axiomatization
   366     eq  (infix "===" 50) where
   367     eq_refl: "x === x" and eq_subst: "x === y ==> P x ==> P y"
   368 
   369   definition "f x y = x + y + 1"
   370   definition g where "g x = f x x"
   371 
   372   abbreviation
   373     neq  (infix "=!=" 50) where
   374     "x =!= y == ~ (x === y)"
   375 
   376 These specifications may be also used in a locale context.  Then the
   377 constants being introduced depend on certain fixed parameters, and the
   378 constant name is qualified by the locale base name.  An internal
   379 abbreviation takes care for convenient input and output, making the
   380 parameters implicit and using the original short name.  See also
   381 HOL/ex/Abstract_NAT.thy for an example of deriving polymorphic
   382 entities from a monomorphic theory.
   383 
   384 Presently, abbreviations are only available 'in' a target locale, but
   385 not inherited by general import expressions.  Also note that
   386 'abbreviation' may be used as a type-safe replacement for 'syntax' +
   387 'translations' in common applications.
   388 
   389 Concrete syntax is attached to specified constants in internal form,
   390 independently of name spaces.  The parse tree representation is
   391 slightly different -- use 'notation' instead of raw 'syntax', and
   392 'translations' with explicit "CONST" markup to accommodate this.
   393 
   394 * Pure: command 'print_abbrevs' prints all constant abbreviations of
   395 the current context.  Print mode "no_abbrevs" prevents inversion of
   396 abbreviations on output.
   397 
   398 * Isar/locales: improved parameter handling:
   399 - use of locales "var" and "struct" no longer necessary;
   400 - parameter renamings are no longer required to be injective.
   401   This enables, for example, to define a locale for endomorphisms thus:
   402   locale endom = homom mult mult h.
   403 
   404 * Isar/locales: changed the way locales with predicates are defined.
   405 Instead of accumulating the specification, the imported expression is
   406 now an interpretation.  INCOMPATIBILITY: different normal form of
   407 locale expressions.  In particular, in interpretations of locales with
   408 predicates, goals repesenting already interpreted fragments are not
   409 removed automatically.  Use methods `intro_locales' and
   410 `unfold_locales'; see below.
   411 
   412 * Isar/locales: new methods `intro_locales' and `unfold_locales'
   413 provide backward reasoning on locales predicates.  The methods are
   414 aware of interpretations and discharge corresponding goals.
   415 `intro_locales' is less aggressive then `unfold_locales' and does not
   416 unfold predicates to assumptions.
   417 
   418 * Isar/locales: the order in which locale fragments are accumulated
   419 has changed.  This enables to override declarations from fragments due
   420 to interpretations -- for example, unwanted simp rules.
   421 
   422 * Isar/locales: interpretation in theories and proof contexts has been
   423 extended.  One may now specify (and prove) equations, which are
   424 unfolded in interpreted theorems.  This is useful for replacing
   425 defined concepts (constants depending on locale parameters) by
   426 concepts already existing in the target context.  Example:
   427 
   428   interpretation partial_order ["op <= :: [int, int] => bool"]
   429     where "partial_order.less (op <=) (x::int) y = (x < y)"
   430 
   431 Typically, the constant `partial_order.less' is created by a definition
   432 specification element in the context of locale partial_order.
   433 
   434 * Provers/induct: improved internal context management to support
   435 local fixes and defines on-the-fly.  Thus explicit meta-level
   436 connectives !! and ==> are rarely required anymore in inductive goals
   437 (using object-logic connectives for this purpose has been long
   438 obsolete anyway).  The subsequent proof patterns illustrate advanced
   439 techniques of natural induction; general datatypes and inductive sets
   440 work analogously (see also src/HOL/Lambda for realistic examples).
   441 
   442 (1) This is how to ``strengthen'' an inductive goal wrt. certain
   443 parameters:
   444 
   445   lemma
   446     fixes n :: nat and x :: 'a
   447     assumes a: "A n x"
   448     shows "P n x"
   449     using a                     -- {* make induct insert fact a *}
   450   proof (induct n arbitrary: x) -- {* generalize goal to "!!x. A n x ==> P n x" *}
   451     case 0
   452     show ?case sorry
   453   next
   454     case (Suc n)
   455     note `!!x. A n x ==> P n x` -- {* induction hypothesis, according to induction rule *}
   456     note `A (Suc n) x`          -- {* induction premise, stemming from fact a *}
   457     show ?case sorry
   458   qed
   459 
   460 (2) This is how to perform induction over ``expressions of a certain
   461 form'', using a locally defined inductive parameter n == "a x"
   462 together with strengthening (the latter is usually required to get
   463 sufficiently flexible induction hypotheses):
   464 
   465   lemma
   466     fixes a :: "'a => nat"
   467     assumes a: "A (a x)"
   468     shows "P (a x)"
   469     using a
   470   proof (induct n == "a x" arbitrary: x)
   471     ...
   472 
   473 See also HOL/Isar_examples/Puzzle.thy for an application of the this
   474 particular technique.
   475 
   476 (3) This is how to perform existential reasoning ('obtains' or
   477 'obtain') by induction, while avoiding explicit object-logic
   478 encodings:
   479 
   480   lemma
   481     fixes n :: nat
   482     obtains x :: 'a where "P n x" and "Q n x"
   483   proof (induct n arbitrary: thesis)
   484     case 0
   485     obtain x where "P 0 x" and "Q 0 x" sorry
   486     then show thesis by (rule 0)
   487   next
   488     case (Suc n)
   489     obtain x where "P n x" and "Q n x" by (rule Suc.hyps)
   490     obtain x where "P (Suc n) x" and "Q (Suc n) x" sorry
   491     then show thesis by (rule Suc.prems)
   492   qed
   493 
   494 Here the 'arbitrary: thesis' specification essentially modifies the
   495 scope of the formal thesis parameter, in order to the get the whole
   496 existence statement through the induction as expected.
   497 
   498 * Provers/induct: mutual induction rules are now specified as a list
   499 of rule sharing the same induction cases.  HOL packages usually
   500 provide foo_bar.inducts for mutually defined items foo and bar
   501 (e.g. inductive sets or datatypes).  INCOMPATIBILITY, users need to
   502 specify mutual induction rules differently, i.e. like this:
   503 
   504   (induct rule: foo_bar.inducts)
   505   (induct set: foo bar)
   506   (induct type: foo bar)
   507 
   508 The ML function ProjectRule.projections turns old-style rules into the
   509 new format.
   510 
   511 * Provers/induct: improved handling of simultaneous goals.  Instead of
   512 introducing object-level conjunction, the statement is now split into
   513 several conclusions, while the corresponding symbolic cases are
   514 nested accordingly.  INCOMPATIBILITY, proofs need to be structured
   515 explicitly.  For example:
   516 
   517   lemma
   518     fixes n :: nat
   519     shows "P n" and "Q n"
   520   proof (induct n)
   521     case 0 case 1
   522     show "P 0" sorry
   523   next
   524     case 0 case 2
   525     show "Q 0" sorry
   526   next
   527     case (Suc n) case 1
   528     note `P n` and `Q n`
   529     show "P (Suc n)" sorry
   530   next
   531     case (Suc n) case 2
   532     note `P n` and `Q n`
   533     show "Q (Suc n)" sorry
   534   qed
   535 
   536 The split into subcases may be deferred as follows -- this is
   537 particularly relevant for goal statements with local premises.
   538 
   539   lemma
   540     fixes n :: nat
   541     shows "A n ==> P n" and "B n ==> Q n"
   542   proof (induct n)
   543     case 0
   544     {
   545       case 1
   546       note `A 0`
   547       show "P 0" sorry
   548     next
   549       case 2
   550       note `B 0`
   551       show "Q 0" sorry
   552     }
   553   next
   554     case (Suc n)
   555     note `A n ==> P n` and `B n ==> Q n`
   556     {
   557       case 1
   558       note `A (Suc n)`
   559       show "P (Suc n)" sorry
   560     next
   561       case 2
   562       note `B (Suc n)`
   563       show "Q (Suc n)" sorry
   564     }
   565   qed
   566 
   567 If simultaneous goals are to be used with mutual rules, the statement
   568 needs to be structured carefully as a two-level conjunction, using
   569 lists of propositions separated by 'and':
   570 
   571   lemma
   572     shows "a : A ==> P1 a"
   573           "a : A ==> P2 a"
   574       and "b : B ==> Q1 b"
   575           "b : B ==> Q2 b"
   576           "b : B ==> Q3 b"
   577   proof (induct set: A B)
   578 
   579 * Provers/induct: support coinduction as well.  See
   580 src/HOL/Library/Coinductive_List.thy for various examples.
   581 
   582 * Attribute "symmetric" produces result with standardized schematic
   583 variables (index 0).  Potential INCOMPATIBILITY.
   584 
   585 * Simplifier: by default the simplifier trace only shows top level
   586 rewrites now. That is, trace_simp_depth_limit is set to 1 by
   587 default. Thus there is less danger of being flooded by the trace. The
   588 trace indicates where parts have been suppressed.
   589   
   590 * Provers/classical: removed obsolete classical version of elim_format
   591 attribute; classical elim/dest rules are now treated uniformly when
   592 manipulating the claset.
   593 
   594 * Provers/classical: stricter checks to ensure that supplied intro,
   595 dest and elim rules are well-formed; dest and elim rules must have at
   596 least one premise.
   597 
   598 * Provers/classical: attributes dest/elim/intro take an optional
   599 weight argument for the rule (just as the Pure versions).  Weights are
   600 ignored by automated tools, but determine the search order of single
   601 rule steps.
   602 
   603 * Syntax: input syntax now supports dummy variable binding "%_. b",
   604 where the body does not mention the bound variable.  Note that dummy
   605 patterns implicitly depend on their context of bounds, which makes
   606 "{_. _}" match any set comprehension as expected.  Potential
   607 INCOMPATIBILITY -- parse translations need to cope with syntactic
   608 constant "_idtdummy" in the binding position.
   609 
   610 * Syntax: removed obsolete syntactic constant "_K" and its associated
   611 parse translation.  INCOMPATIBILITY -- use dummy abstraction instead,
   612 for example "A -> B" => "Pi A (%_. B)".
   613 
   614 * Pure: 'class_deps' command visualizes the subclass relation, using
   615 the graph browser tool.
   616 
   617 * Pure: 'print_theory' now suppresses entities with internal name
   618 (trailing "_") by default; use '!' option for full details.
   619 
   620 
   621 *** HOL ***
   622 
   623 * Code generator library theories:
   624   * Pretty_Int represents HOL integers by big integer literals in target
   625     languages.
   626   * Pretty_Char represents HOL characters by character literals in target
   627     languages.
   628   * Pretty_Char_chr like Pretty_Char, but also offers treatment of character
   629     codes; includes Pretty_Int.
   630   * Executable_Set allows to generate code for finite sets using lists.
   631   * Executable_Rat implements rational numbers as triples (sign, enumerator,
   632     denominator).
   633   * Executable_Real implements a subset of real numbers, namly those
   634     representable by rational numbers.
   635   * Efficient_Nat implements natural numbers by integers, which in general will
   636     result in higher efficency; pattern matching with 0/Suc is eliminated;
   637     includes Pretty_Int.
   638   * ML_String provides an additional datatype ml_string; in the HOL default
   639     setup, strings in HOL are mapped to lists of HOL characters in SML; values
   640     of type ml_string are mapped to strings in SML.
   641   * ML_Int provides an additional datatype ml_int which is mapped to to SML
   642     built-in integers.
   643 
   644 * New package for inductive predicates
   645 
   646   An n-ary predicate p with m parameters z_1, ..., z_m can now be defined via
   647 
   648     inductive
   649       p :: "U_1 => ... => U_m => T_1 => ... => T_n => bool"
   650       for z_1 :: U_1 and ... and z_n :: U_m
   651     where
   652       rule_1: "... ==> p z_1 ... z_m t_1_1 ... t_1_n"
   653     | ...
   654 
   655   rather than
   656 
   657     consts s :: "U_1 => ... => U_m => (T_1 * ... * T_n) set"
   658 
   659     abbreviation p :: "U_1 => ... => U_m => T_1 => ... => T_n => bool"
   660     where "p z_1 ... z_m x_1 ... x_n == (x_1, ..., x_n) : s z_1 ... z_m"
   661 
   662     inductive "s z_1 ... z_m"
   663     intros
   664       rule_1: "... ==> (t_1_1, ..., t_1_n) : s z_1 ... z_m"
   665       ...
   666 
   667   For backward compatibility, there is a wrapper allowing inductive
   668   sets to be defined with the new package via
   669 
   670     inductive_set
   671       s :: "U_1 => ... => U_m => (T_1 * ... * T_n) set"
   672       for z_1 :: U_1 and ... and z_n :: U_m
   673     where
   674       rule_1: "... ==> (t_1_1, ..., t_1_n) : s z_1 ... z_m"
   675     | ...
   676 
   677   or
   678 
   679     inductive_set
   680       s :: "U_1 => ... => U_m => (T_1 * ... * T_n) set"
   681       and p :: "U_1 => ... => U_m => T_1 => ... => T_n => bool"
   682       for z_1 :: U_1 and ... and z_n :: U_m
   683     where
   684       "p z_1 ... z_m x_1 ... x_n == (x_1, ..., x_n) : s z_1 ... z_m"
   685     | rule_1: "... ==> p z_1 ... z_m t_1_1 ... t_1_n"
   686     | ...
   687 
   688   if the additional syntax "p ..." is required.
   689 
   690   Many examples can be found in the subdirectories Auth, Bali, Induct,
   691   or MicroJava.
   692 
   693   INCOMPATIBILITIES:
   694 
   695   - Since declaration and definition of inductive sets or predicates
   696     is no longer separated, abbreviations involving the newly introduced
   697     sets or predicates must be specified together with the introduction
   698     rules after the "where" keyword (see example above), rather than before
   699     the actual inductive definition.
   700 
   701   - The variables in induction and elimination rules are now quantified
   702     in the order of their occurrence in the introduction rules, rather than
   703     in alphabetical order. Since this may break some proofs, these proofs
   704     either have to be repaired, e.g. by reordering the variables
   705     a_i_1 ... a_i_{k_i} in Isar "case" statements of the form
   706 
   707       case (rule_i a_i_1 ... a_i_{k_i})
   708 
   709     or the old order of quantification has to be restored by explicitly adding
   710     meta-level quantifiers in the introduction rules, i.e.
   711 
   712       | rule_i: "!!a_i_1 ... a_i_{k_i}. ... ==> p z_1 ... z_m t_i_1 ... t_i_n"
   713 
   714   - The format of the elimination rules is now
   715 
   716       p z_1 ... z_m x_1 ... x_n ==>
   717         (!!a_1_1 ... a_1_{k_1}. x_1 = t_1_1 ==> ... ==> x_n = t_1_n ==> ... ==> P)
   718         ==> ... ==> P
   719 
   720     for predicates and
   721 
   722       (x_1, ..., x_n) : s z_1 ... z_m ==>
   723         (!!a_1_1 ... a_1_{k_1}. x_1 = t_1_1 ==> ... ==> x_n = t_1_n ==> ... ==> P)
   724         ==> ... ==> P
   725 
   726     for sets rather than
   727 
   728       x : s z_1 ... z_m ==>
   729         (!!a_1_1 ... a_1_{k_1}. x = (t_1_1, ..., t_1_n) ==> ... ==> P)
   730         ==> ... ==> P
   731 
   732     This may require terms in goals to be expanded to n-tuples (e.g. using case_tac
   733     or simplification with the split_paired_all rule) before the above elimination
   734     rule is applicable.
   735 
   736   - The elimination or case analysis rules for (mutually) inductive sets or
   737     predicates are now called "p_1.cases" ... "p_k.cases". The list of rules
   738     "p_1_..._p_k.elims" is no longer available.
   739 
   740 * Method "metis" proves goals by applying the Metis general-purpose
   741 resolution prover.  Examples are in the directory MetisExamples.  See
   742 also http://gilith.com/software/metis/
   743   
   744 * Command 'sledgehammer' invokes external automatic theorem provers as
   745 background processes.  It generates calls to the "metis" method if
   746 successful. These can be pasted into the proof.  Users do not have to
   747 wait for the automatic provers to return.
   748 
   749 * Case-expressions allow arbitrary constructor-patterns (including "_") and
   750   take their order into account, like in functional programming.
   751   Internally, this is translated into nested case-expressions; missing cases
   752   are added and mapped to the predefined constant "undefined". In complicated
   753   cases printing may no longer show the original input but the internal
   754   form. Lambda-abstractions allow the same form of pattern matching:
   755   "% pat1 => e1 | ..." is an abbreviation for
   756   "%x. case x of pat1 => e1 | ..." where x is a new variable.
   757 
   758 * IntDef: The constant "int :: nat => int" has been removed; now "int"
   759   is an abbreviation for "of_nat :: nat => int". The simplification rules
   760   for "of_nat" have been changed to work like "int" did previously.
   761   (potential INCOMPATIBILITY)
   762   - "of_nat (Suc m)" simplifies to "1 + of_nat m" instead of "of_nat m + 1"
   763   - of_nat_diff and of_nat_mult are no longer default simp rules
   764 
   765 * Method "algebra" solves polynomial equations over (semi)rings using
   766   Groebner bases. The (semi)ring structure is defined by locales and
   767   the tool setup depends on that generic context. Installing the
   768   method for a specific type involves instantiating the locale and
   769   possibly adding declarations for computation on the coefficients.
   770   The method is already instantiated for natural numbers and for the
   771   axiomatic class of idoms with numerals.  See also the paper by
   772   Chaieb and Wenzel at CALCULEMUS 2007 for the general principles
   773   underlying this architecture of context-aware proof-tools.
   774 
   775 * constant "List.op @" now named "List.append".  Use ML antiquotations
   776 @{const_name List.append} or @{term " ... @ ... "} to circumvent
   777 possible incompatibilities when working on ML level.
   778 
   779 * Constant renames due to introduction of canonical name prefixing for
   780   class package:
   781 
   782     HOL.abs ~> HOL.minus_class.abs
   783     HOL.divide ~> HOL.divide_class.divide
   784     Nat.power ~> Nat.power_class.power
   785     Nat.size ~> Nat.size_class.size
   786     Numeral.number_of ~> Numeral.number_class.number_of
   787     FixedPoint.Inf ~> FixedPoint.complete_lattice_class.Inf
   788 
   789 * Rudimentary class target mechanism involves constant renames:
   790 
   791     Orderings.min ~> Orderings.ord_class.min
   792     Orderings.max ~> Orderings.ord_class.max
   793     FixedPoint.Sup ~> FixedPoint.complete_lattice_class.Sup
   794 
   795 * primrec: missing cases mapped to "undefined" instead of "arbitrary"
   796 
   797 * new constant "undefined" with axiom "undefined x = undefined"
   798 
   799 * new class "default" with associated constant "default"
   800 
   801 * new function listsum :: 'a list => 'a for arbitrary monoids.
   802   Special syntax: "SUM x <- xs. f x" (and latex variants)
   803 
   804 * new (input only) syntax for Haskell-like list comprehension, eg
   805   [(x,y). x <- xs, y <- ys, x ~= y]
   806   For details see List.thy.
   807 
   808 * The special syntax for function "filter" has changed from [x : xs. P] to
   809   [x <- xs. P] to avoid an ambiguity caused by list comprehension syntax,
   810   and for uniformity. INCOMPATIBILITY
   811 
   812 * Lemma "set_take_whileD" renamed to "set_takeWhileD"
   813 
   814 * New lemma collection field_simps (an extension of ring_simps)
   815   for manipulating (in)equations involving division. Multiplies
   816   with all denominators that can be proved to be non-zero (in equations)
   817   or positive/negative (in inequations).
   818 
   819 * Lemma collections ring_eq_simps, group_eq_simps and ring_distrib
   820   have been improved and renamed to ring_simps, group_simps and ring_distribs.
   821   Removed lemmas field_xyz in Ring_and_Field
   822   because they were subsumed by lemmas xyz.
   823 INCOMPATIBILITY.
   824 
   825 * Library/Pretty_Int.thy: maps HOL numerals on target language integer literals
   826   when generating code.
   827 
   828 * Library/Pretty_Char.thy: maps HOL characters on target language character literals
   829   when generating code.
   830 
   831 * Library/Commutative_Ring.thy: switched from recdef to function package;
   832   constants add, mul, pow now curried.  Infix syntax for algebraic operations.
   833 
   834 * Some steps towards more uniform lattice theory development in HOL.
   835 
   836     constants "meet" and "join" now named "inf" and "sup"
   837     constant "Meet" now named "Inf"
   838 
   839     classes "meet_semilorder" and "join_semilorder" now named
   840       "lower_semilattice" and "upper_semilattice"
   841     class "lorder" now named "lattice"
   842     class "comp_lat" now named "complete_lattice"
   843 
   844     Instantiation of lattice classes allows explicit definitions
   845     for "inf" and "sup" operations.
   846 
   847   INCOMPATIBILITY.  Theorem renames:
   848 
   849     meet_left_le            ~> inf_le1
   850     meet_right_le           ~> inf_le2
   851     join_left_le            ~> sup_ge1
   852     join_right_le           ~> sup_ge2
   853     meet_join_le            ~> inf_sup_ord
   854     le_meetI                ~> le_infI
   855     join_leI                ~> le_supI
   856     le_meet                 ~> le_inf_iff
   857     le_join                 ~> ge_sup_conv
   858     meet_idempotent         ~> inf_idem
   859     join_idempotent         ~> sup_idem
   860     meet_comm               ~> inf_commute
   861     join_comm               ~> sup_commute
   862     meet_leI1               ~> le_infI1
   863     meet_leI2               ~> le_infI2
   864     le_joinI1               ~> le_supI1
   865     le_joinI2               ~> le_supI2
   866     meet_assoc              ~> inf_assoc
   867     join_assoc              ~> sup_assoc
   868     meet_left_comm          ~> inf_left_commute
   869     meet_left_idempotent    ~> inf_left_idem
   870     join_left_comm          ~> sup_left_commute
   871     join_left_idempotent    ~> sup_left_idem
   872     meet_aci                ~> inf_aci
   873     join_aci                ~> sup_aci
   874     le_def_meet             ~> le_iff_inf
   875     le_def_join             ~> le_iff_sup
   876     join_absorp2            ~> sup_absorb2
   877     join_absorp1            ~> sup_absorb1
   878     meet_absorp1            ~> inf_absorb1
   879     meet_absorp2            ~> inf_absorb2
   880     meet_join_absorp        ~> inf_sup_absorb
   881     join_meet_absorp        ~> sup_inf_absorb
   882     distrib_join_le         ~> distrib_sup_le
   883     distrib_meet_le         ~> distrib_inf_le
   884 
   885     add_meet_distrib_left   ~> add_inf_distrib_left
   886     add_join_distrib_left   ~> add_sup_distrib_left
   887     is_join_neg_meet        ~> is_join_neg_inf
   888     is_meet_neg_join        ~> is_meet_neg_sup
   889     add_meet_distrib_right  ~> add_inf_distrib_right
   890     add_join_distrib_right  ~> add_sup_distrib_right
   891     add_meet_join_distribs  ~> add_sup_inf_distribs
   892     join_eq_neg_meet        ~> sup_eq_neg_inf
   893     meet_eq_neg_join        ~> inf_eq_neg_sup
   894     add_eq_meet_join        ~> add_eq_inf_sup
   895     meet_0_imp_0            ~> inf_0_imp_0
   896     join_0_imp_0            ~> sup_0_imp_0
   897     meet_0_eq_0             ~> inf_0_eq_0
   898     join_0_eq_0             ~> sup_0_eq_0
   899     neg_meet_eq_join        ~> neg_inf_eq_sup
   900     neg_join_eq_meet        ~> neg_sup_eq_inf
   901     join_eq_if              ~> sup_eq_if
   902 
   903     mono_meet               ~> mono_inf
   904     mono_join               ~> mono_sup
   905     meet_bool_eq            ~> inf_bool_eq
   906     join_bool_eq            ~> sup_bool_eq
   907     meet_fun_eq             ~> inf_fun_eq
   908     join_fun_eq             ~> sup_fun_eq
   909     meet_set_eq             ~> inf_set_eq
   910     join_set_eq             ~> sup_set_eq
   911     meet1_iff               ~> inf1_iff
   912     meet2_iff               ~> inf2_iff
   913     meet1I                  ~> inf1I
   914     meet2I                  ~> inf2I
   915     meet1D1                 ~> inf1D1
   916     meet2D1                 ~> inf2D1
   917     meet1D2                 ~> inf1D2
   918     meet2D2                 ~> inf2D2
   919     meet1E                  ~> inf1E
   920     meet2E                  ~> inf2E
   921     join1_iff               ~> sup1_iff
   922     join2_iff               ~> sup2_iff
   923     join1I1                 ~> sup1I1
   924     join2I1                 ~> sup2I1
   925     join1I1                 ~> sup1I1
   926     join2I2                 ~> sup1I2
   927     join1CI                 ~> sup1CI
   928     join2CI                 ~> sup2CI
   929     join1E                  ~> sup1E
   930     join2E                  ~> sup2E
   931 
   932     is_meet_Meet            ~> is_meet_Inf
   933     Meet_bool_def           ~> Inf_bool_def
   934     Meet_fun_def            ~> Inf_fun_def
   935     Meet_greatest           ~> Inf_greatest
   936     Meet_lower              ~> Inf_lower
   937     Meet_set_def            ~> Inf_set_def
   938 
   939     listsp_meetI            ~> listsp_infI
   940     listsp_meet_eq          ~> listsp_inf_eq
   941 
   942     meet_min                ~> inf_min
   943     join_max                ~> sup_max
   944 
   945 * Classes "order" and "linorder": facts "refl", "trans" and
   946 "cases" renamed ro "order_refl", "order_trans" and "linorder_cases", to
   947 avoid clashes with HOL "refl" and "trans". INCOMPATIBILITY.
   948 
   949 * Classes "order" and "linorder": 
   950 potential INCOMPATIBILITY: order of proof goals in order/linorder instance
   951 proofs changed.
   952 
   953 * Dropped lemma duplicate def_imp_eq in favor of meta_eq_to_obj_eq.
   954 INCOMPATIBILITY.
   955 
   956 * Dropped lemma duplicate if_def2 in favor of if_bool_eq_conj.
   957 INCOMPATIBILITY.
   958 
   959 * Added syntactic class "size"; overloaded constant "size" now has
   960 type "'a::size ==> bool"
   961 
   962 * Renamed constants "Divides.op div", "Divides.op mod" and "Divides.op
   963 dvd" to "Divides.div_class.div", "Divides.div_class.mod" and "Divides.dvd". INCOMPATIBILITY.
   964 
   965 * Added method "lexicographic_order" automatically synthesizes
   966 termination relations as lexicographic combinations of size measures
   967 -- 'function' package.
   968 
   969 * HOL/records: generalised field-update to take a function on the
   970 field rather than the new value: r(|A := x|) is translated to A_update
   971 (K x) r The K-combinator that is internally used is called K_record.
   972 INCOMPATIBILITY: Usage of the plain update functions has to be
   973 adapted.
   974  
   975 * axclass "semiring_0" now contains annihilation axioms x * 0 = 0 and
   976 0 * x = 0, which are required for a semiring.  Richer structures do
   977 not inherit from semiring_0 anymore, because this property is a
   978 theorem there, not an axiom.  INCOMPATIBILITY: In instances of
   979 semiring_0, there is more to prove, but this is mostly trivial.
   980 
   981 * axclass "recpower" was generalized to arbitrary monoids, not just
   982 commutative semirings.  INCOMPATIBILITY: If you use recpower and need
   983 commutativity or a semiring property, add the corresponding classes.
   984 
   985 * Unified locale partial_order with class definition (cf. theory
   986 Orderings), added parameter ``less''.  INCOMPATIBILITY.
   987 
   988 * Constant "List.list_all2" in List.thy now uses authentic syntax.
   989 INCOMPATIBILITY: translations containing list_all2 may go wrong.  On
   990 Isar level, use abbreviations instead.
   991 
   992 * Renamed constant "List.op mem" to "List.memberl" INCOMPATIBILITY:
   993 rarely occuring name references (e.g. ``List.op mem.simps'') require
   994 renaming (e.g. ``List.memberl.simps'').
   995 
   996 * Renamed constants "0" to "HOL.zero_class.zero" and "1" to "HOL.one_class.one".
   997 INCOMPATIBILITY.
   998 
   999 * Added class "HOL.eq", allowing for code generation with polymorphic equality.
  1000 
  1001 * Numeral syntax: type 'bin' which was a mere type copy of 'int' has
  1002 been abandoned in favour of plain 'int'. INCOMPATIBILITY --
  1003 significant changes for setting up numeral syntax for types:
  1004 
  1005   - new constants Numeral.pred and Numeral.succ instead
  1006       of former Numeral.bin_pred and Numeral.bin_succ.
  1007   - Use integer operations instead of bin_add, bin_mult and so on.
  1008   - Numeral simplification theorems named Numeral.numeral_simps instead of Bin_simps.
  1009   - ML structure Bin_Simprocs now named Int_Numeral_Base_Simprocs.
  1010 
  1011 See HOL/Integ/IntArith.thy for an example setup.
  1012 
  1013 * New top level command 'normal_form' computes the normal form of a
  1014 term that may contain free variables. For example ``normal_form
  1015 "rev[a,b,c]"'' produces ``[b,c,a]'' (without proof).  This command is
  1016 suitable for heavy-duty computations because the functions are
  1017 compiled to ML first.
  1018 
  1019 * Alternative iff syntax "A <-> B" for equality on bool (with priority
  1020 25 like -->); output depends on the "iff" print_mode, the default is
  1021 "A = B" (with priority 50).
  1022 
  1023 * Renamed constants in HOL.thy and Orderings.thy:
  1024     op +   ~> HOL.plus_class.plus
  1025     op -   ~> HOL.minus_class.minus
  1026     uminus ~> HOL.minus_class.uminus
  1027     abs    ~> HOL.abs_class.abs
  1028     op *   ~> HOL.times_class.times
  1029     op <   ~> HOL.ord_class.less
  1030     op <=  ~> HOL.ord_class.less_eq
  1031 
  1032 Adaptions may be required in the following cases:
  1033 
  1034 a) User-defined constants using any of the names "plus", "minus", "times",
  1035 "less" or "less_eq". The standard syntax translations for "+", "-" and "*"
  1036 may go wrong.
  1037 INCOMPATIBILITY: use more specific names.
  1038 
  1039 b) Variables named "plus", "minus", "times", "less", "less_eq"
  1040 INCOMPATIBILITY: use more specific names.
  1041 
  1042 c) Permutative equations (e.g. "a + b = b + a")
  1043 Since the change of names also changes the order of terms, permutative
  1044 rewrite rules may get applied in a different order. Experience shows that
  1045 this is rarely the case (only two adaptions in the whole Isabelle
  1046 distribution).
  1047 INCOMPATIBILITY: rewrite proofs
  1048 
  1049 d) ML code directly refering to constant names
  1050 This in general only affects hand-written proof tactics, simprocs and so on.
  1051 INCOMPATIBILITY: grep your sourcecode and replace names.  Consider use
  1052 of const_name ML antiquotations.
  1053 
  1054 * Relations less (<) and less_eq (<=) are also available on type bool.
  1055 Modified syntax to disallow nesting without explicit parentheses,
  1056 e.g. "(x < y) < z" or "x < (y < z)", but NOT "x < y < z".
  1057 
  1058 * "LEAST x:A. P" expands to "LEAST x. x:A & P" (input only).
  1059 
  1060 * Relation composition operator "op O" now has precedence 75 and binds
  1061 stronger than union and intersection. INCOMPATIBILITY.
  1062 
  1063 * The old set interval syntax "{m..n(}" (and relatives) has been
  1064 removed.  Use "{m..<n}" (and relatives) instead.
  1065 
  1066 * In the context of the assumption "~(s = t)" the Simplifier rewrites
  1067 "t = s" to False (by simproc "neq_simproc").  For backward
  1068 compatibility this can be disabled by ML "reset use_neq_simproc".
  1069 
  1070 * "m dvd n" where m and n are numbers is evaluated to True/False by
  1071 simp.
  1072 
  1073 * Theorem Cons_eq_map_conv no longer declared as ``simp''.
  1074 
  1075 * Theorem setsum_mult renamed to setsum_right_distrib.
  1076 
  1077 * Prefer ex1I over ex_ex1I in single-step reasoning, e.g. by the
  1078 ``rule'' method.
  1079 
  1080 * Reimplemented methods ``sat'' and ``satx'', with several
  1081 improvements: goals no longer need to be stated as "<prems> ==>
  1082 False", equivalences (i.e. "=" on type bool) are handled, variable
  1083 names of the form "lit_<n>" are no longer reserved, significant
  1084 speedup.
  1085 
  1086 * Methods ``sat'' and ``satx'' can now replay MiniSat proof traces.
  1087 zChaff is still supported as well.
  1088 
  1089 * 'inductive' and 'datatype': provide projections of mutual rules,
  1090 bundled as foo_bar.inducts;
  1091 
  1092 * Library: moved theories Parity, GCD, Binomial, Infinite_Set to
  1093 Library.
  1094 
  1095 * Library: moved theory Accessible_Part to main HOL.
  1096 
  1097 * Library: added theory Coinductive_List of potentially infinite lists
  1098 as greatest fixed-point.
  1099 
  1100 * Library: added theory AssocList which implements (finite) maps as
  1101 association lists.
  1102 
  1103 * Added proof method ``evaluation'' for efficiently solving a goal
  1104 (i.e. a boolean expression) by compiling it to ML. The goal is
  1105 "proved" (via an oracle) if it evaluates to True.
  1106 
  1107 * Linear arithmetic now splits certain operators (e.g. min, max, abs)
  1108 also when invoked by the simplifier.  This results in the simplifier
  1109 being more powerful on arithmetic goals.  INCOMPATIBILITY.  Set
  1110 fast_arith_split_limit to 0 to obtain the old behavior.
  1111 
  1112 * Support for hex (0x20) and binary (0b1001) numerals.
  1113 
  1114 * New method: reify eqs (t), where eqs are equations for an
  1115 interpretation I :: 'a list => 'b => 'c and t::'c is an optional
  1116 parameter, computes a term s::'b and a list xs::'a list and proves the
  1117 theorem I xs s = t. This is also known as reification or quoting. The
  1118 resulting theorem is applied to the subgoal to substitute t with I xs
  1119 s.  If t is omitted, the subgoal itself is reified.
  1120 
  1121 * New method: reflection corr_thm eqs (t). The parameters eqs and (t)
  1122 are as explained above. corr_thm is a theorem for I vs (f t) = I vs t,
  1123 where f is supposed to be a computable function (in the sense of code
  1124 generattion). The method uses reify to compute s and xs as above then
  1125 applies corr_thm and uses normalization by evaluation to "prove" f s =
  1126 r and finally gets the theorem t = r, which is again applied to the
  1127 subgoal. An Example is available in HOL/ex/ReflectionEx.thy.
  1128 
  1129 * Reflection: Automatic reification now handels binding, an example
  1130 is available in HOL/ex/ReflectionEx.thy
  1131 
  1132 
  1133 *** HOL-Algebra ***
  1134 
  1135 * Formalisation of ideals and the quotient construction over rings.
  1136 
  1137 * Order and lattice theory no longer based on records.
  1138 INCOMPATIBILITY.
  1139 
  1140 * Renamed lemmas least_carrier -> least_closed and greatest_carrier ->
  1141 greatest_closed.  INCOMPATIBILITY.
  1142 
  1143 * Method algebra is now set up via an attribute.  For examples see
  1144 Ring.thy.  INCOMPATIBILITY: the method is now weaker on combinations
  1145 of algebraic structures.
  1146 
  1147 * Renamed theory CRing to Ring.
  1148 
  1149 
  1150 *** HOL-Complex ***
  1151 
  1152 * Theory Real: new method ferrack implements quantifier elimination
  1153 for linear arithmetic over the reals. The quantifier elimination
  1154 feature is used only for decision, for compatibility with arith. This
  1155 means a goal is either solved or left unchanged, no simplification.
  1156 
  1157 * Hyperreal: Functions root and sqrt are now defined on negative real
  1158 inputs so that root n (- x) = - root n x and sqrt (- x) = - sqrt x.
  1159 Nonnegativity side conditions have been removed from many lemmas, so
  1160 that more subgoals may now be solved by simplification; potential
  1161 INCOMPATIBILITY.
  1162 
  1163 * Real: New axiomatic classes formalize real normed vector spaces and
  1164 algebras, using new overloaded constants scaleR :: real => 'a => 'a
  1165 and norm :: 'a => real.
  1166 
  1167 * Real: New constant of_real :: real => 'a::real_algebra_1 injects
  1168 from reals into other types. The overloaded constant Reals :: 'a set
  1169 is now defined as range of_real; potential INCOMPATIBILITY.
  1170 
  1171 * Real: ML code generation is supported now and hence also quickcheck.
  1172 Reals are implemented as arbitrary precision rationals.
  1173 
  1174 * Hyperreal: Several constants that previously worked only for the
  1175 reals have been generalized, so they now work over arbitrary vector
  1176 spaces. Type annotations may need to be added in some cases; potential
  1177 INCOMPATIBILITY.
  1178 
  1179   Infinitesimal  :: ('a::real_normed_vector) star set
  1180   HFinite        :: ('a::real_normed_vector) star set
  1181   HInfinite      :: ('a::real_normed_vector) star set
  1182   approx         :: ('a::real_normed_vector) star => 'a star => bool
  1183   monad          :: ('a::real_normed_vector) star => 'a star set
  1184   galaxy         :: ('a::real_normed_vector) star => 'a star set
  1185   (NS)LIMSEQ     :: [nat => 'a::real_normed_vector, 'a] => bool
  1186   (NS)convergent :: (nat => 'a::real_normed_vector) => bool
  1187   (NS)Bseq       :: (nat => 'a::real_normed_vector) => bool
  1188   (NS)Cauchy     :: (nat => 'a::real_normed_vector) => bool
  1189   (NS)LIM        :: ['a::real_normed_vector => 'b::real_normed_vector, 'a, 'b] => bool
  1190   is(NS)Cont     :: ['a::real_normed_vector => 'b::real_normed_vector, 'a] => bool
  1191   deriv          :: ['a::real_normed_field => 'a, 'a, 'a] => bool
  1192   sgn            :: 'a::real_normed_vector => 'a
  1193   exp            :: 'a::{recpower,real_normed_field,banach} => 'a
  1194 
  1195 * Complex: Some complex-specific constants are now abbreviations for
  1196 overloaded ones: complex_of_real = of_real, cmod = norm, hcmod =
  1197 hnorm.  Other constants have been entirely removed in favor of the
  1198 polymorphic versions (INCOMPATIBILITY):
  1199 
  1200   approx        <-- capprox
  1201   HFinite       <-- CFinite
  1202   HInfinite     <-- CInfinite
  1203   Infinitesimal <-- CInfinitesimal
  1204   monad         <-- cmonad
  1205   galaxy        <-- cgalaxy
  1206   (NS)LIM       <-- (NS)CLIM, (NS)CRLIM
  1207   is(NS)Cont    <-- is(NS)Contc, is(NS)contCR
  1208   (ns)deriv     <-- (ns)cderiv
  1209 
  1210 
  1211 *** ML ***
  1212 
  1213 * Generic arithmetic modules: Tools/integer.ML, Tools/rat.ML, Tools/float.ML
  1214 
  1215 * Context data interfaces (Theory/Proof/GenericDataFun): removed
  1216 name/print, uninitialized data defaults to ad-hoc copy of empty value,
  1217 init only required for impure data.  INCOMPATIBILITY: empty really
  1218 need to be empty (no dependencies on theory content!)
  1219 
  1220 * ML within Isar: antiquotations allow to embed statically-checked
  1221 formal entities in the source, referring to the context available at
  1222 compile-time.  For example:
  1223 
  1224 ML {* @{typ "'a => 'b"} *}
  1225 ML {* @{term "%x. x"} *}
  1226 ML {* @{prop "x == y"} *}
  1227 ML {* @{ctyp "'a => 'b"} *}
  1228 ML {* @{cterm "%x. x"} *}
  1229 ML {* @{cprop "x == y"} *}
  1230 ML {* @{thm asm_rl} *}
  1231 ML {* @{thms asm_rl} *}
  1232 ML {* @{const_name c} *}
  1233 ML {* @{const_syntax c} *}
  1234 ML {* @{context} *}
  1235 ML {* @{theory} *}
  1236 ML {* @{theory Pure} *}
  1237 ML {* @{simpset} *}
  1238 ML {* @{claset} *}
  1239 ML {* @{clasimpset} *}
  1240 
  1241 The same works for sources being ``used'' within an Isar context.
  1242 
  1243 * ML in Isar: improved error reporting; extra verbosity with
  1244 Toplevel.debug enabled.
  1245 
  1246 * Pure/library:
  1247 
  1248   val burrow: ('a list -> 'b list) -> 'a list list -> 'b list list
  1249   val fold_burrow: ('a list -> 'c -> 'b list * 'd) -> 'a list list -> 'c -> 'b list list * 'd
  1250 
  1251 The semantics of "burrow" is: "take a function with *simulatanously*
  1252 transforms a list of value, and apply it *simulatanously* to a list of
  1253 list of values of the appropriate type". Compare this with "map" which
  1254 would *not* apply its argument function simulatanously but in
  1255 sequence; "fold_burrow" has an additional context.
  1256 
  1257 * Pure/library: functions map2 and fold2 with curried syntax for
  1258 simultanous mapping and folding:
  1259 
  1260     val map2: ('a -> 'b -> 'c) -> 'a list -> 'b list -> 'c list
  1261     val fold2: ('a -> 'b -> 'c -> 'c) -> 'a list -> 'b list -> 'c -> 'c
  1262 
  1263 * Pure/library: indexed lists - some functions in the Isabelle library
  1264 treating lists over 'a as finite mappings from [0...n] to 'a have been
  1265 given more convenient names and signatures reminiscent of similar
  1266 functions for alists, tables, etc:
  1267 
  1268   val nth: 'a list -> int -> 'a 
  1269   val nth_map: int -> ('a -> 'a) -> 'a list -> 'a list
  1270   val fold_index: (int * 'a -> 'b -> 'b) -> 'a list -> 'b -> 'b
  1271 
  1272 Note that fold_index starts counting at index 0, not 1 like foldln
  1273 used to.
  1274 
  1275 * Pure/library: added general ``divide_and_conquer'' combinator on
  1276 lists.
  1277 
  1278 * Pure/General/table.ML: the join operations now works via exceptions
  1279 DUP/SAME instead of type option.  This is simpler in simple cases, and
  1280 admits slightly more efficient complex applications.
  1281 
  1282 * Pure: datatype Context.generic joins theory/Proof.context and
  1283 provides some facilities for code that works in either kind of
  1284 context, notably GenericDataFun for uniform theory and proof data.
  1285 
  1286 * Pure: 'advanced' translation functions (parse_translation etc.) now
  1287 use Context.generic instead of just theory.
  1288 
  1289 * Pure: simplified internal attribute type, which is now always
  1290 Context.generic * thm -> Context.generic * thm.  Global (theory)
  1291 vs. local (Proof.context) attributes have been discontinued, while
  1292 minimizing code duplication.  Thm.rule_attribute and
  1293 Thm.declaration_attribute build canonical attributes; see also
  1294 structure Context for further operations on Context.generic, notably
  1295 GenericDataFun.  INCOMPATIBILITY, need to adapt attribute type
  1296 declarations and definitions.
  1297 
  1298 * Pure/kernel: consts certification ignores sort constraints given in
  1299 signature declarations.  (This information is not relevant to the
  1300 logic, but only for type inference.)  IMPORTANT INTERNAL CHANGE,
  1301 potential INCOMPATIBILITY.
  1302 
  1303 * Pure: axiomatic type classes are now purely definitional, with
  1304 explicit proofs of class axioms and super class relations performed
  1305 internally.  See Pure/axclass.ML for the main internal interfaces --
  1306 notably AxClass.define_class supercedes AxClass.add_axclass, and
  1307 AxClass.axiomatize_class/classrel/arity supercede
  1308 Sign.add_classes/classrel/arities.
  1309 
  1310 * Pure/Isar: Args/Attrib parsers operate on Context.generic --
  1311 global/local versions on theory vs. Proof.context have been
  1312 discontinued; Attrib.syntax and Method.syntax have been adapted
  1313 accordingly.  INCOMPATIBILITY, need to adapt parser expressions for
  1314 attributes, methods, etc.
  1315 
  1316 * Pure: several functions of signature "... -> theory -> theory * ..."
  1317 have been reoriented to "... -> theory -> ... * theory" in order to
  1318 allow natural usage in combination with the ||>, ||>>, |-> and
  1319 fold_map combinators.
  1320 
  1321 * Pure: official theorem names (closed derivations) and additional
  1322 comments (tags) are now strictly separate.  Name hints -- which are
  1323 maintained as tags -- may be attached any time without affecting the
  1324 derivation.
  1325 
  1326 * Pure: primitive rule lift_rule now takes goal cterm instead of an
  1327 actual goal state (thm).  Use Thm.lift_rule (Thm.cprem_of st i) to
  1328 achieve the old behaviour.
  1329 
  1330 * Pure: the "Goal" constant is now called "prop", supporting a
  1331 slightly more general idea of ``protecting'' meta-level rule
  1332 statements.
  1333 
  1334 * Pure: Logic.(un)varify only works in a global context, which is now
  1335 enforced instead of silently assumed.  INCOMPATIBILITY, may use
  1336 Logic.legacy_(un)varify as temporary workaround.
  1337 
  1338 * Pure: structure Name provides scalable operations for generating
  1339 internal variable names, notably Name.variants etc.  This replaces
  1340 some popular functions from term.ML:
  1341 
  1342   Term.variant		->  Name.variant
  1343   Term.variantlist	->  Name.variant_list  (*canonical argument order*)
  1344   Term.invent_names	->  Name.invent_list
  1345 
  1346 Note that low-level renaming rarely occurs in new code -- operations
  1347 from structure Variable are used instead (see below).
  1348 
  1349 * Pure: structure Variable provides fundamental operations for proper
  1350 treatment of fixed/schematic variables in a context.  For example,
  1351 Variable.import introduces fixes for schematics of given facts and
  1352 Variable.export reverses the effect (up to renaming) -- this replaces
  1353 various freeze_thaw operations.
  1354 
  1355 * Pure: structure Goal provides simple interfaces for
  1356 init/conclude/finish and tactical prove operations (replacing former
  1357 Tactic.prove).  Goal.prove is the canonical way to prove results
  1358 within a given context; Goal.prove_global is a degraded version for
  1359 theory level goals, including a global Drule.standard.  Note that
  1360 OldGoals.prove_goalw_cterm has long been obsolete, since it is
  1361 ill-behaved in a local proof context (e.g. with local fixes/assumes or
  1362 in a locale context).
  1363 
  1364 * Isar: simplified treatment of user-level errors, using exception
  1365 ERROR of string uniformly.  Function error now merely raises ERROR,
  1366 without any side effect on output channels.  The Isar toplevel takes
  1367 care of proper display of ERROR exceptions.  ML code may use plain
  1368 handle/can/try; cat_error may be used to concatenate errors like this:
  1369 
  1370   ... handle ERROR msg => cat_error msg "..."
  1371 
  1372 Toplevel ML code (run directly or through the Isar toplevel) may be
  1373 embedded into the Isar toplevel with exception display/debug like
  1374 this:
  1375 
  1376   Isar.toplevel (fn () => ...)
  1377 
  1378 INCOMPATIBILITY, removed special transform_error facilities, removed
  1379 obsolete variants of user-level exceptions (ERROR_MESSAGE,
  1380 Context.PROOF, ProofContext.CONTEXT, Proof.STATE, ProofHistory.FAIL)
  1381 -- use plain ERROR instead.
  1382 
  1383 * Isar: theory setup now has type (theory -> theory), instead of a
  1384 list.  INCOMPATIBILITY, may use #> to compose setup functions.
  1385 
  1386 * Isar: installed ML toplevel pretty printer for type Proof.context,
  1387 subject to ProofContext.debug/verbose flags.
  1388 
  1389 * Isar: Toplevel.theory_to_proof admits transactions that modify the
  1390 theory before entering a proof state.  Transactions now always see a
  1391 quasi-functional intermediate checkpoint, both in interactive and
  1392 batch mode.
  1393 
  1394 * Simplifier: the simpset of a running simplification process now
  1395 contains a proof context (cf. Simplifier.the_context), which is the
  1396 very context that the initial simpset has been retrieved from (by
  1397 simpset_of/local_simpset_of).  Consequently, all plug-in components
  1398 (solver, looper etc.) may depend on arbitrary proof data.
  1399 
  1400 * Simplifier.inherit_context inherits the proof context (plus the
  1401 local bounds) of the current simplification process; any simproc
  1402 etc. that calls the Simplifier recursively should do this!  Removed
  1403 former Simplifier.inherit_bounds, which is already included here --
  1404 INCOMPATIBILITY.  Tools based on low-level rewriting may even have to
  1405 specify an explicit context using Simplifier.context/theory_context.
  1406 
  1407 * Simplifier/Classical Reasoner: more abstract interfaces
  1408 change_simpset/claset for modifying the simpset/claset reference of a
  1409 theory; raw versions simpset/claset_ref etc. have been discontinued --
  1410 INCOMPATIBILITY.
  1411 
  1412 * Provers: more generic wrt. syntax of object-logics, avoid hardwired
  1413 "Trueprop" etc.
  1414 
  1415 
  1416 *** System ***
  1417 
  1418 * settings: ML_IDENTIFIER -- which is appended to user specific heap
  1419 locations -- now includes the Isabelle version identifier as well.
  1420 This simplifies use of multiple Isabelle installations.
  1421 
  1422 * isabelle-process: option -S (secure mode) disables some critical
  1423 operations, notably runtime compilation and evaluation of ML source
  1424 code.
  1425 
  1426 
  1427 New in Isabelle2005 (October 2005)
  1428 ----------------------------------
  1429 
  1430 *** General ***
  1431 
  1432 * Theory headers: the new header syntax for Isar theories is
  1433 
  1434   theory <name>
  1435   imports <theory1> ... <theoryN>
  1436   uses <file1> ... <fileM>
  1437   begin
  1438 
  1439 where the 'uses' part is optional.  The previous syntax
  1440 
  1441   theory <name> = <theory1> + ... + <theoryN>:
  1442 
  1443 will disappear in the next release.  Use isatool fixheaders to convert
  1444 existing theory files.  Note that there is no change in ancient
  1445 non-Isar theories now, but these will disappear soon.
  1446 
  1447 * Theory loader: parent theories can now also be referred to via
  1448 relative and absolute paths.
  1449 
  1450 * Command 'find_theorems' searches for a list of criteria instead of a
  1451 list of constants. Known criteria are: intro, elim, dest, name:string,
  1452 simp:term, and any term. Criteria can be preceded by '-' to select
  1453 theorems that do not match. Intro, elim, dest select theorems that
  1454 match the current goal, name:s selects theorems whose fully qualified
  1455 name contain s, and simp:term selects all simplification rules whose
  1456 lhs match term.  Any other term is interpreted as pattern and selects
  1457 all theorems matching the pattern. Available in ProofGeneral under
  1458 'ProofGeneral -> Find Theorems' or C-c C-f.  Example:
  1459 
  1460   C-c C-f (100) "(_::nat) + _ + _" intro -name: "HOL."
  1461 
  1462 prints the last 100 theorems matching the pattern "(_::nat) + _ + _",
  1463 matching the current goal as introduction rule and not having "HOL."
  1464 in their name (i.e. not being defined in theory HOL).
  1465 
  1466 * Command 'thms_containing' has been discontinued in favour of
  1467 'find_theorems'; INCOMPATIBILITY.
  1468 
  1469 * Communication with Proof General is now 8bit clean, which means that
  1470 Unicode text in UTF-8 encoding may be used within theory texts (both
  1471 formal and informal parts).  Cf. option -U of the Isabelle Proof
  1472 General interface.  Here are some simple examples (cf. src/HOL/ex):
  1473 
  1474   http://isabelle.in.tum.de/library/HOL/ex/Hebrew.html
  1475   http://isabelle.in.tum.de/library/HOL/ex/Chinese.html
  1476 
  1477 * Improved efficiency of the Simplifier and, to a lesser degree, the
  1478 Classical Reasoner.  Typical big applications run around 2 times
  1479 faster.
  1480 
  1481 
  1482 *** Document preparation ***
  1483 
  1484 * Commands 'display_drafts' and 'print_drafts' perform simple output
  1485 of raw sources.  Only those symbols that do not require additional
  1486 LaTeX packages (depending on comments in isabellesym.sty) are
  1487 displayed properly, everything else is left verbatim.  isatool display
  1488 and isatool print are used as front ends (these are subject to the
  1489 DVI/PDF_VIEWER and PRINT_COMMAND settings, respectively).
  1490 
  1491 * Command tags control specific markup of certain regions of text,
  1492 notably folding and hiding.  Predefined tags include "theory" (for
  1493 theory begin and end), "proof" for proof commands, and "ML" for
  1494 commands involving ML code; the additional tags "visible" and
  1495 "invisible" are unused by default.  Users may give explicit tag
  1496 specifications in the text, e.g. ''by %invisible (auto)''.  The
  1497 interpretation of tags is determined by the LaTeX job during document
  1498 preparation: see option -V of isatool usedir, or options -n and -t of
  1499 isatool document, or even the LaTeX macros \isakeeptag, \isafoldtag,
  1500 \isadroptag.
  1501 
  1502 Several document versions may be produced at the same time via isatool
  1503 usedir (the generated index.html will link all of them).  Typical
  1504 specifications include ''-V document=theory,proof,ML'' to present
  1505 theory/proof/ML parts faithfully, ''-V outline=/proof,/ML'' to fold
  1506 proof and ML commands, and ''-V mutilated=-theory,-proof,-ML'' to omit
  1507 these parts without any formal replacement text.  The Isabelle site
  1508 default settings produce ''document'' and ''outline'' versions as
  1509 specified above.
  1510 
  1511 * Several new antiquotations:
  1512 
  1513   @{term_type term} prints a term with its type annotated;
  1514 
  1515   @{typeof term} prints the type of a term;
  1516 
  1517   @{const const} is the same as @{term const}, but checks that the
  1518   argument is a known logical constant;
  1519 
  1520   @{term_style style term} and @{thm_style style thm} print a term or
  1521   theorem applying a "style" to it
  1522 
  1523   @{ML text}
  1524 
  1525 Predefined styles are 'lhs' and 'rhs' printing the lhs/rhs of
  1526 definitions, equations, inequations etc., 'concl' printing only the
  1527 conclusion of a meta-logical statement theorem, and 'prem1' .. 'prem19'
  1528 to print the specified premise.  TermStyle.add_style provides an ML
  1529 interface for introducing further styles.  See also the "LaTeX Sugar"
  1530 document practical applications.  The ML antiquotation prints
  1531 type-checked ML expressions verbatim.
  1532 
  1533 * Markup commands 'chapter', 'section', 'subsection', 'subsubsection',
  1534 and 'text' support optional locale specification '(in loc)', which
  1535 specifies the default context for interpreting antiquotations.  For
  1536 example: 'text (in lattice) {* @{thm inf_assoc}*}'.
  1537 
  1538 * Option 'locale=NAME' of antiquotations specifies an alternative
  1539 context interpreting the subsequent argument.  For example: @{thm
  1540 [locale=lattice] inf_assoc}.
  1541 
  1542 * Proper output of proof terms (@{prf ...} and @{full_prf ...}) within
  1543 a proof context.
  1544 
  1545 * Proper output of antiquotations for theory commands involving a
  1546 proof context (such as 'locale' or 'theorem (in loc) ...').
  1547 
  1548 * Delimiters of outer tokens (string etc.) now produce separate LaTeX
  1549 macros (\isachardoublequoteopen, isachardoublequoteclose etc.).
  1550 
  1551 * isatool usedir: new option -C (default true) controls whether option
  1552 -D should include a copy of the original document directory; -C false
  1553 prevents unwanted effects such as copying of administrative CVS data.
  1554 
  1555 
  1556 *** Pure ***
  1557 
  1558 * Considerably improved version of 'constdefs' command.  Now performs
  1559 automatic type-inference of declared constants; additional support for
  1560 local structure declarations (cf. locales and HOL records), see also
  1561 isar-ref manual.  Potential INCOMPATIBILITY: need to observe strictly
  1562 sequential dependencies of definitions within a single 'constdefs'
  1563 section; moreover, the declared name needs to be an identifier.  If
  1564 all fails, consider to fall back on 'consts' and 'defs' separately.
  1565 
  1566 * Improved indexed syntax and implicit structures.  First of all,
  1567 indexed syntax provides a notational device for subscripted
  1568 application, using the new syntax \<^bsub>term\<^esub> for arbitrary
  1569 expressions.  Secondly, in a local context with structure
  1570 declarations, number indexes \<^sub>n or the empty index (default
  1571 number 1) refer to a certain fixed variable implicitly; option
  1572 show_structs controls printing of implicit structures.  Typical
  1573 applications of these concepts involve record types and locales.
  1574 
  1575 * New command 'no_syntax' removes grammar declarations (and
  1576 translations) resulting from the given syntax specification, which is
  1577 interpreted in the same manner as for the 'syntax' command.
  1578 
  1579 * 'Advanced' translation functions (parse_translation etc.) may depend
  1580 on the signature of the theory context being presently used for
  1581 parsing/printing, see also isar-ref manual.
  1582 
  1583 * Improved 'oracle' command provides a type-safe interface to turn an
  1584 ML expression of type theory -> T -> term into a primitive rule of
  1585 type theory -> T -> thm (i.e. the functionality of Thm.invoke_oracle
  1586 is already included here); see also FOL/ex/IffExample.thy;
  1587 INCOMPATIBILITY.
  1588 
  1589 * axclass: name space prefix for class "c" is now "c_class" (was "c"
  1590 before); "cI" is no longer bound, use "c.intro" instead.
  1591 INCOMPATIBILITY.  This change avoids clashes of fact bindings for
  1592 axclasses vs. locales.
  1593 
  1594 * Improved internal renaming of symbolic identifiers -- attach primes
  1595 instead of base 26 numbers.
  1596 
  1597 * New flag show_question_marks controls printing of leading question
  1598 marks in schematic variable names.
  1599 
  1600 * In schematic variable names, *any* symbol following \<^isub> or
  1601 \<^isup> is now treated as part of the base name.  For example, the
  1602 following works without printing of awkward ".0" indexes:
  1603 
  1604   lemma "x\<^isub>1 = x\<^isub>2 ==> x\<^isub>2 = x\<^isub>1"
  1605     by simp
  1606 
  1607 * Inner syntax includes (*(*nested*) comments*).
  1608 
  1609 * Pretty printer now supports unbreakable blocks, specified in mixfix
  1610 annotations as "(00...)".
  1611 
  1612 * Clear separation of logical types and nonterminals, where the latter
  1613 may only occur in 'syntax' specifications or type abbreviations.
  1614 Before that distinction was only partially implemented via type class
  1615 "logic" vs. "{}".  Potential INCOMPATIBILITY in rare cases of improper
  1616 use of 'types'/'consts' instead of 'nonterminals'/'syntax'.  Some very
  1617 exotic syntax specifications may require further adaption
  1618 (e.g. Cube/Cube.thy).
  1619 
  1620 * Removed obsolete type class "logic", use the top sort {} instead.
  1621 Note that non-logical types should be declared as 'nonterminals'
  1622 rather than 'types'.  INCOMPATIBILITY for new object-logic
  1623 specifications.
  1624 
  1625 * Attributes 'induct' and 'cases': type or set names may now be
  1626 locally fixed variables as well.
  1627 
  1628 * Simplifier: can now control the depth to which conditional rewriting
  1629 is traced via the PG menu Isabelle -> Settings -> Trace Simp Depth
  1630 Limit.
  1631 
  1632 * Simplifier: simplification procedures may now take the current
  1633 simpset into account (cf. Simplifier.simproc(_i) / mk_simproc
  1634 interface), which is very useful for calling the Simplifier
  1635 recursively.  Minor INCOMPATIBILITY: the 'prems' argument of simprocs
  1636 is gone -- use prems_of_ss on the simpset instead.  Moreover, the
  1637 low-level mk_simproc no longer applies Logic.varify internally, to
  1638 allow for use in a context of fixed variables.
  1639 
  1640 * thin_tac now works even if the assumption being deleted contains !!
  1641 or ==>.  More generally, erule now works even if the major premise of
  1642 the elimination rule contains !! or ==>.
  1643 
  1644 * Method 'rules' has been renamed to 'iprover'. INCOMPATIBILITY.
  1645 
  1646 * Reorganized bootstrapping of the Pure theories; CPure is now derived
  1647 from Pure, which contains all common declarations already.  Both
  1648 theories are defined via plain Isabelle/Isar .thy files.
  1649 INCOMPATIBILITY: elements of CPure (such as the CPure.intro /
  1650 CPure.elim / CPure.dest attributes) now appear in the Pure name space;
  1651 use isatool fixcpure to adapt your theory and ML sources.
  1652 
  1653 * New syntax 'name(i-j, i-, i, ...)' for referring to specific
  1654 selections of theorems in named facts via index ranges.
  1655 
  1656 * 'print_theorems': in theory mode, really print the difference
  1657 wrt. the last state (works for interactive theory development only),
  1658 in proof mode print all local facts (cf. 'print_facts');
  1659 
  1660 * 'hide': option '(open)' hides only base names.
  1661 
  1662 * More efficient treatment of intermediate checkpoints in interactive
  1663 theory development.
  1664 
  1665 * Code generator is now invoked via code_module (incremental code
  1666 generation) and code_library (modular code generation, ML structures
  1667 for each theory).  INCOMPATIBILITY: new keywords 'file' and 'contains'
  1668 must be quoted when used as identifiers.
  1669 
  1670 * New 'value' command for reading, evaluating and printing terms using
  1671 the code generator.  INCOMPATIBILITY: command keyword 'value' must be
  1672 quoted when used as identifier.
  1673 
  1674 
  1675 *** Locales ***
  1676 
  1677 * New commands for the interpretation of locale expressions in
  1678 theories (1), locales (2) and proof contexts (3).  These generate
  1679 proof obligations from the expression specification.  After the
  1680 obligations have been discharged, theorems of the expression are added
  1681 to the theory, target locale or proof context.  The synopsis of the
  1682 commands is a follows:
  1683 
  1684   (1) interpretation expr inst
  1685   (2) interpretation target < expr
  1686   (3) interpret expr inst
  1687 
  1688 Interpretation in theories and proof contexts require a parameter
  1689 instantiation of terms from the current context.  This is applied to
  1690 specifications and theorems of the interpreted expression.
  1691 Interpretation in locales only permits parameter renaming through the
  1692 locale expression.  Interpretation is smart in that interpretations
  1693 that are active already do not occur in proof obligations, neither are
  1694 instantiated theorems stored in duplicate.  Use 'print_interps' to
  1695 inspect active interpretations of a particular locale.  For details,
  1696 see the Isar Reference manual.  Examples can be found in
  1697 HOL/Finite_Set.thy and HOL/Algebra/UnivPoly.thy.
  1698 
  1699 INCOMPATIBILITY: former 'instantiate' has been withdrawn, use
  1700 'interpret' instead.
  1701 
  1702 * New context element 'constrains' for adding type constraints to
  1703 parameters.
  1704 
  1705 * Context expressions: renaming of parameters with syntax
  1706 redeclaration.
  1707 
  1708 * Locale declaration: 'includes' disallowed.
  1709 
  1710 * Proper static binding of attribute syntax -- i.e. types / terms /
  1711 facts mentioned as arguments are always those of the locale definition
  1712 context, independently of the context of later invocations.  Moreover,
  1713 locale operations (renaming and type / term instantiation) are applied
  1714 to attribute arguments as expected.
  1715 
  1716 INCOMPATIBILITY of the ML interface: always pass Attrib.src instead of
  1717 actual attributes; rare situations may require Attrib.attribute to
  1718 embed those attributes into Attrib.src that lack concrete syntax.
  1719 Attribute implementations need to cooperate properly with the static
  1720 binding mechanism.  Basic parsers Args.XXX_typ/term/prop and
  1721 Attrib.XXX_thm etc. already do the right thing without further
  1722 intervention.  Only unusual applications -- such as "where" or "of"
  1723 (cf. src/Pure/Isar/attrib.ML), which process arguments depending both
  1724 on the context and the facts involved -- may have to assign parsed
  1725 values to argument tokens explicitly.
  1726 
  1727 * Changed parameter management in theorem generation for long goal
  1728 statements with 'includes'.  INCOMPATIBILITY: produces a different
  1729 theorem statement in rare situations.
  1730 
  1731 * Locale inspection command 'print_locale' omits notes elements.  Use
  1732 'print_locale!' to have them included in the output.
  1733 
  1734 
  1735 *** Provers ***
  1736 
  1737 * Provers/hypsubst.ML: improved version of the subst method, for
  1738 single-step rewriting: it now works in bound variable contexts. New is
  1739 'subst (asm)', for rewriting an assumption.  INCOMPATIBILITY: may
  1740 rewrite a different subterm than the original subst method, which is
  1741 still available as 'simplesubst'.
  1742 
  1743 * Provers/quasi.ML: new transitivity reasoners for transitivity only
  1744 and quasi orders.
  1745 
  1746 * Provers/trancl.ML: new transitivity reasoner for transitive and
  1747 reflexive-transitive closure of relations.
  1748 
  1749 * Provers/blast.ML: new reference depth_limit to make blast's depth
  1750 limit (previously hard-coded with a value of 20) user-definable.
  1751 
  1752 * Provers/simplifier.ML has been moved to Pure, where Simplifier.setup
  1753 is peformed already.  Object-logics merely need to finish their
  1754 initial simpset configuration as before.  INCOMPATIBILITY.
  1755 
  1756 
  1757 *** HOL ***
  1758 
  1759 * Symbolic syntax of Hilbert Choice Operator is now as follows:
  1760 
  1761   syntax (epsilon)
  1762     "_Eps" :: "[pttrn, bool] => 'a"    ("(3\<some>_./ _)" [0, 10] 10)
  1763 
  1764 The symbol \<some> is displayed as the alternative epsilon of LaTeX
  1765 and x-symbol; use option '-m epsilon' to get it actually printed.
  1766 Moreover, the mathematically important symbolic identifier \<epsilon>
  1767 becomes available as variable, constant etc.  INCOMPATIBILITY,
  1768 
  1769 * "x > y" abbreviates "y < x" and "x >= y" abbreviates "y <= x".
  1770 Similarly for all quantifiers: "ALL x > y" etc.  The x-symbol for >=
  1771 is \<ge>. New transitivity rules have been added to HOL/Orderings.thy to
  1772 support corresponding Isar calculations.
  1773 
  1774 * "{x:A. P}" abbreviates "{x. x:A & P}", and similarly for "\<in>"
  1775 instead of ":".
  1776 
  1777 * theory SetInterval: changed the syntax for open intervals:
  1778 
  1779   Old       New
  1780   {..n(}    {..<n}
  1781   {)n..}    {n<..}
  1782   {m..n(}   {m..<n}
  1783   {)m..n}   {m<..n}
  1784   {)m..n(}  {m<..<n}
  1785 
  1786 The old syntax is still supported but will disappear in the next
  1787 release.  For conversion use the following Emacs search and replace
  1788 patterns (these are not perfect but work quite well):
  1789 
  1790   {)\([^\.]*\)\.\.  ->  {\1<\.\.}
  1791   \.\.\([^(}]*\)(}  ->  \.\.<\1}
  1792 
  1793 * Theory Commutative_Ring (in Library): method comm_ring for proving
  1794 equalities in commutative rings; method 'algebra' provides a generic
  1795 interface.
  1796 
  1797 * Theory Finite_Set: changed the syntax for 'setsum', summation over
  1798 finite sets: "setsum (%x. e) A", which used to be "\<Sum>x:A. e", is
  1799 now either "SUM x:A. e" or "\<Sum>x \<in> A. e". The bound variable can
  1800 be a tuple pattern.
  1801 
  1802 Some new syntax forms are available:
  1803 
  1804   "\<Sum>x | P. e"      for     "setsum (%x. e) {x. P}"
  1805   "\<Sum>x = a..b. e"   for     "setsum (%x. e) {a..b}"
  1806   "\<Sum>x = a..<b. e"  for     "setsum (%x. e) {a..<b}"
  1807   "\<Sum>x < k. e"      for     "setsum (%x. e) {..<k}"
  1808 
  1809 The latter form "\<Sum>x < k. e" used to be based on a separate
  1810 function "Summation", which has been discontinued.
  1811 
  1812 * theory Finite_Set: in structured induction proofs, the insert case
  1813 is now 'case (insert x F)' instead of the old counterintuitive 'case
  1814 (insert F x)'.
  1815 
  1816 * The 'refute' command has been extended to support a much larger
  1817 fragment of HOL, including axiomatic type classes, constdefs and
  1818 typedefs, inductive datatypes and recursion.
  1819 
  1820 * New tactics 'sat' and 'satx' to prove propositional tautologies.
  1821 Requires zChaff with proof generation to be installed.  See
  1822 HOL/ex/SAT_Examples.thy for examples.
  1823 
  1824 * Datatype induction via method 'induct' now preserves the name of the
  1825 induction variable. For example, when proving P(xs::'a list) by
  1826 induction on xs, the induction step is now P(xs) ==> P(a#xs) rather
  1827 than P(list) ==> P(a#list) as previously.  Potential INCOMPATIBILITY
  1828 in unstructured proof scripts.
  1829 
  1830 * Reworked implementation of records.  Improved scalability for
  1831 records with many fields, avoiding performance problems for type
  1832 inference. Records are no longer composed of nested field types, but
  1833 of nested extension types. Therefore the record type only grows linear
  1834 in the number of extensions and not in the number of fields.  The
  1835 top-level (users) view on records is preserved.  Potential
  1836 INCOMPATIBILITY only in strange cases, where the theory depends on the
  1837 old record representation. The type generated for a record is called
  1838 <record_name>_ext_type.
  1839 
  1840 Flag record_quick_and_dirty_sensitive can be enabled to skip the
  1841 proofs triggered by a record definition or a simproc (if
  1842 quick_and_dirty is enabled).  Definitions of large records can take
  1843 quite long.
  1844 
  1845 New simproc record_upd_simproc for simplification of multiple record
  1846 updates enabled by default.  Moreover, trivial updates are also
  1847 removed: r(|x := x r|) = r.  INCOMPATIBILITY: old proofs break
  1848 occasionally, since simplification is more powerful by default.
  1849 
  1850 * typedef: proper support for polymorphic sets, which contain extra
  1851 type-variables in the term.
  1852 
  1853 * Simplifier: automatically reasons about transitivity chains
  1854 involving "trancl" (r^+) and "rtrancl" (r^*) by setting up tactics
  1855 provided by Provers/trancl.ML as additional solvers.  INCOMPATIBILITY:
  1856 old proofs break occasionally as simplification may now solve more
  1857 goals than previously.
  1858 
  1859 * Simplifier: converts x <= y into x = y if assumption y <= x is
  1860 present.  Works for all partial orders (class "order"), in particular
  1861 numbers and sets.  For linear orders (e.g. numbers) it treats ~ x < y
  1862 just like y <= x.
  1863 
  1864 * Simplifier: new simproc for "let x = a in f x".  If a is a free or
  1865 bound variable or a constant then the let is unfolded.  Otherwise
  1866 first a is simplified to b, and then f b is simplified to g. If
  1867 possible we abstract b from g arriving at "let x = b in h x",
  1868 otherwise we unfold the let and arrive at g.  The simproc can be
  1869 enabled/disabled by the reference use_let_simproc.  Potential
  1870 INCOMPATIBILITY since simplification is more powerful by default.
  1871 
  1872 * Classical reasoning: the meson method now accepts theorems as arguments.
  1873 
  1874 * Prover support: pre-release of the Isabelle-ATP linkup, which runs background
  1875 jobs to provide advice on the provability of subgoals.
  1876 
  1877 * Theory OrderedGroup and Ring_and_Field: various additions and
  1878 improvements to faciliate calculations involving equalities and
  1879 inequalities.
  1880 
  1881 The following theorems have been eliminated or modified
  1882 (INCOMPATIBILITY):
  1883 
  1884   abs_eq             now named abs_of_nonneg
  1885   abs_of_ge_0        now named abs_of_nonneg
  1886   abs_minus_eq       now named abs_of_nonpos
  1887   imp_abs_id         now named abs_of_nonneg
  1888   imp_abs_neg_id     now named abs_of_nonpos
  1889   mult_pos           now named mult_pos_pos
  1890   mult_pos_le        now named mult_nonneg_nonneg
  1891   mult_pos_neg_le    now named mult_nonneg_nonpos
  1892   mult_pos_neg2_le   now named mult_nonneg_nonpos2
  1893   mult_neg           now named mult_neg_neg
  1894   mult_neg_le        now named mult_nonpos_nonpos
  1895 
  1896 * The following lemmas in Ring_and_Field have been added to the simplifier:
  1897      
  1898      zero_le_square
  1899      not_square_less_zero 
  1900 
  1901   The following lemmas have been deleted from Real/RealPow:
  1902   
  1903      realpow_zero_zero
  1904      realpow_two
  1905      realpow_less
  1906      zero_le_power
  1907      realpow_two_le
  1908      abs_realpow_two
  1909      realpow_two_abs     
  1910 
  1911 * Theory Parity: added rules for simplifying exponents.
  1912 
  1913 * Theory List:
  1914 
  1915 The following theorems have been eliminated or modified
  1916 (INCOMPATIBILITY):
  1917 
  1918   list_all_Nil       now named list_all.simps(1)
  1919   list_all_Cons      now named list_all.simps(2)
  1920   list_all_conv      now named list_all_iff
  1921   set_mem_eq         now named mem_iff
  1922 
  1923 * Theories SetsAndFunctions and BigO (see HOL/Library) support
  1924 asymptotic "big O" calculations.  See the notes in BigO.thy.
  1925 
  1926 
  1927 *** HOL-Complex ***
  1928 
  1929 * Theory RealDef: better support for embedding natural numbers and
  1930 integers in the reals.
  1931 
  1932 The following theorems have been eliminated or modified
  1933 (INCOMPATIBILITY):
  1934 
  1935   exp_ge_add_one_self  now requires no hypotheses
  1936   real_of_int_add      reversed direction of equality (use [symmetric])
  1937   real_of_int_minus    reversed direction of equality (use [symmetric])
  1938   real_of_int_diff     reversed direction of equality (use [symmetric])
  1939   real_of_int_mult     reversed direction of equality (use [symmetric])
  1940 
  1941 * Theory RComplete: expanded support for floor and ceiling functions.
  1942 
  1943 * Theory Ln is new, with properties of the natural logarithm
  1944 
  1945 * Hyperreal: There is a new type constructor "star" for making
  1946 nonstandard types.  The old type names are now type synonyms:
  1947 
  1948   hypreal = real star
  1949   hypnat = nat star
  1950   hcomplex = complex star
  1951 
  1952 * Hyperreal: Many groups of similarly-defined constants have been
  1953 replaced by polymorphic versions (INCOMPATIBILITY):
  1954 
  1955   star_of <-- hypreal_of_real, hypnat_of_nat, hcomplex_of_complex
  1956 
  1957   starset      <-- starsetNat, starsetC
  1958   *s*          <-- *sNat*, *sc*
  1959   starset_n    <-- starsetNat_n, starsetC_n
  1960   *sn*         <-- *sNatn*, *scn*
  1961   InternalSets <-- InternalNatSets, InternalCSets
  1962 
  1963   starfun      <-- starfun{Nat,Nat2,C,RC,CR}
  1964   *f*          <-- *fNat*, *fNat2*, *fc*, *fRc*, *fcR*
  1965   starfun_n    <-- starfun{Nat,Nat2,C,RC,CR}_n
  1966   *fn*         <-- *fNatn*, *fNat2n*, *fcn*, *fRcn*, *fcRn*
  1967   InternalFuns <-- InternalNatFuns, InternalNatFuns2, Internal{C,RC,CR}Funs
  1968 
  1969 * Hyperreal: Many type-specific theorems have been removed in favor of
  1970 theorems specific to various axiomatic type classes (INCOMPATIBILITY):
  1971 
  1972   add_commute <-- {hypreal,hypnat,hcomplex}_add_commute
  1973   add_assoc   <-- {hypreal,hypnat,hcomplex}_add_assocs
  1974   OrderedGroup.add_0 <-- {hypreal,hypnat,hcomplex}_add_zero_left
  1975   OrderedGroup.add_0_right <-- {hypreal,hcomplex}_add_zero_right
  1976   right_minus <-- hypreal_add_minus
  1977   left_minus <-- {hypreal,hcomplex}_add_minus_left
  1978   mult_commute <-- {hypreal,hypnat,hcomplex}_mult_commute
  1979   mult_assoc <-- {hypreal,hypnat,hcomplex}_mult_assoc
  1980   mult_1_left <-- {hypreal,hypnat}_mult_1, hcomplex_mult_one_left
  1981   mult_1_right <-- hcomplex_mult_one_right
  1982   mult_zero_left <-- hcomplex_mult_zero_left
  1983   left_distrib <-- {hypreal,hypnat,hcomplex}_add_mult_distrib
  1984   right_distrib <-- hypnat_add_mult_distrib2
  1985   zero_neq_one <-- {hypreal,hypnat,hcomplex}_zero_not_eq_one
  1986   right_inverse <-- hypreal_mult_inverse
  1987   left_inverse <-- hypreal_mult_inverse_left, hcomplex_mult_inv_left
  1988   order_refl <-- {hypreal,hypnat}_le_refl
  1989   order_trans <-- {hypreal,hypnat}_le_trans
  1990   order_antisym <-- {hypreal,hypnat}_le_anti_sym
  1991   order_less_le <-- {hypreal,hypnat}_less_le
  1992   linorder_linear <-- {hypreal,hypnat}_le_linear
  1993   add_left_mono <-- {hypreal,hypnat}_add_left_mono
  1994   mult_strict_left_mono <-- {hypreal,hypnat}_mult_less_mono2
  1995   add_nonneg_nonneg <-- hypreal_le_add_order
  1996 
  1997 * Hyperreal: Separate theorems having to do with type-specific
  1998 versions of constants have been merged into theorems that apply to the
  1999 new polymorphic constants (INCOMPATIBILITY):
  2000 
  2001   STAR_UNIV_set <-- {STAR_real,NatStar_real,STARC_complex}_set
  2002   STAR_empty_set <-- {STAR,NatStar,STARC}_empty_set
  2003   STAR_Un <-- {STAR,NatStar,STARC}_Un
  2004   STAR_Int <-- {STAR,NatStar,STARC}_Int
  2005   STAR_Compl <-- {STAR,NatStar,STARC}_Compl
  2006   STAR_subset <-- {STAR,NatStar,STARC}_subset
  2007   STAR_mem <-- {STAR,NatStar,STARC}_mem
  2008   STAR_mem_Compl <-- {STAR,STARC}_mem_Compl
  2009   STAR_diff <-- {STAR,STARC}_diff
  2010   STAR_star_of_image_subset <-- {STAR_hypreal_of_real, NatStar_hypreal_of_real,
  2011     STARC_hcomplex_of_complex}_image_subset
  2012   starset_n_Un <-- starset{Nat,C}_n_Un
  2013   starset_n_Int <-- starset{Nat,C}_n_Int
  2014   starset_n_Compl <-- starset{Nat,C}_n_Compl
  2015   starset_n_diff <-- starset{Nat,C}_n_diff
  2016   InternalSets_Un <-- Internal{Nat,C}Sets_Un
  2017   InternalSets_Int <-- Internal{Nat,C}Sets_Int
  2018   InternalSets_Compl <-- Internal{Nat,C}Sets_Compl
  2019   InternalSets_diff <-- Internal{Nat,C}Sets_diff
  2020   InternalSets_UNIV_diff <-- Internal{Nat,C}Sets_UNIV_diff
  2021   InternalSets_starset_n <-- Internal{Nat,C}Sets_starset{Nat,C}_n
  2022   starset_starset_n_eq <-- starset{Nat,C}_starset{Nat,C}_n_eq
  2023   starset_n_starset <-- starset{Nat,C}_n_starset{Nat,C}
  2024   starfun_n_starfun <-- starfun{Nat,Nat2,C,RC,CR}_n_starfun{Nat,Nat2,C,RC,CR}
  2025   starfun <-- starfun{Nat,Nat2,C,RC,CR}
  2026   starfun_mult <-- starfun{Nat,Nat2,C,RC,CR}_mult
  2027   starfun_add <-- starfun{Nat,Nat2,C,RC,CR}_add
  2028   starfun_minus <-- starfun{Nat,Nat2,C,RC,CR}_minus
  2029   starfun_diff <-- starfun{C,RC,CR}_diff
  2030   starfun_o <-- starfun{NatNat2,Nat2,_stafunNat,C,C_starfunRC,_starfunCR}_o
  2031   starfun_o2 <-- starfun{NatNat2,_stafunNat,C,C_starfunRC,_starfunCR}_o2
  2032   starfun_const_fun <-- starfun{Nat,Nat2,C,RC,CR}_const_fun
  2033   starfun_inverse <-- starfun{Nat,C,RC,CR}_inverse
  2034   starfun_eq <-- starfun{Nat,Nat2,C,RC,CR}_eq
  2035   starfun_eq_iff <-- starfun{C,RC,CR}_eq_iff
  2036   starfun_Id <-- starfunC_Id
  2037   starfun_approx <-- starfun{Nat,CR}_approx
  2038   starfun_capprox <-- starfun{C,RC}_capprox
  2039   starfun_abs <-- starfunNat_rabs
  2040   starfun_lambda_cancel <-- starfun{C,CR,RC}_lambda_cancel
  2041   starfun_lambda_cancel2 <-- starfun{C,CR,RC}_lambda_cancel2
  2042   starfun_mult_HFinite_approx <-- starfunCR_mult_HFinite_capprox
  2043   starfun_mult_CFinite_capprox <-- starfun{C,RC}_mult_CFinite_capprox
  2044   starfun_add_capprox <-- starfun{C,RC}_add_capprox
  2045   starfun_add_approx <-- starfunCR_add_approx
  2046   starfun_inverse_inverse <-- starfunC_inverse_inverse
  2047   starfun_divide <-- starfun{C,CR,RC}_divide
  2048   starfun_n <-- starfun{Nat,C}_n
  2049   starfun_n_mult <-- starfun{Nat,C}_n_mult
  2050   starfun_n_add <-- starfun{Nat,C}_n_add
  2051   starfun_n_add_minus <-- starfunNat_n_add_minus
  2052   starfun_n_const_fun <-- starfun{Nat,C}_n_const_fun
  2053   starfun_n_minus <-- starfun{Nat,C}_n_minus
  2054   starfun_n_eq <-- starfun{Nat,C}_n_eq
  2055 
  2056   star_n_add <-- {hypreal,hypnat,hcomplex}_add
  2057   star_n_minus <-- {hypreal,hcomplex}_minus
  2058   star_n_diff <-- {hypreal,hcomplex}_diff
  2059   star_n_mult <-- {hypreal,hcomplex}_mult
  2060   star_n_inverse <-- {hypreal,hcomplex}_inverse
  2061   star_n_le <-- {hypreal,hypnat}_le
  2062   star_n_less <-- {hypreal,hypnat}_less
  2063   star_n_zero_num <-- {hypreal,hypnat,hcomplex}_zero_num
  2064   star_n_one_num <-- {hypreal,hypnat,hcomplex}_one_num
  2065   star_n_abs <-- hypreal_hrabs
  2066   star_n_divide <-- hcomplex_divide
  2067 
  2068   star_of_add <-- {hypreal_of_real,hypnat_of_nat,hcomplex_of_complex}_add
  2069   star_of_minus <-- {hypreal_of_real,hcomplex_of_complex}_minus
  2070   star_of_diff <-- hypreal_of_real_diff
  2071   star_of_mult <-- {hypreal_of_real,hypnat_of_nat,hcomplex_of_complex}_mult
  2072   star_of_one <-- {hypreal_of_real,hcomplex_of_complex}_one
  2073   star_of_zero <-- {hypreal_of_real,hypnat_of_nat,hcomplex_of_complex}_zero
  2074   star_of_le <-- {hypreal_of_real,hypnat_of_nat}_le_iff
  2075   star_of_less <-- {hypreal_of_real,hypnat_of_nat}_less_iff
  2076   star_of_eq <-- {hypreal_of_real,hypnat_of_nat,hcomplex_of_complex}_eq_iff
  2077   star_of_inverse <-- {hypreal_of_real,hcomplex_of_complex}_inverse
  2078   star_of_divide <-- {hypreal_of_real,hcomplex_of_complex}_divide
  2079   star_of_of_nat <-- {hypreal_of_real,hcomplex_of_complex}_of_nat
  2080   star_of_of_int <-- {hypreal_of_real,hcomplex_of_complex}_of_int
  2081   star_of_number_of <-- {hypreal,hcomplex}_number_of
  2082   star_of_number_less <-- number_of_less_hypreal_of_real_iff
  2083   star_of_number_le <-- number_of_le_hypreal_of_real_iff
  2084   star_of_eq_number <-- hypreal_of_real_eq_number_of_iff
  2085   star_of_less_number <-- hypreal_of_real_less_number_of_iff
  2086   star_of_le_number <-- hypreal_of_real_le_number_of_iff
  2087   star_of_power <-- hypreal_of_real_power
  2088   star_of_eq_0 <-- hcomplex_of_complex_zero_iff
  2089 
  2090 * Hyperreal: new method "transfer" that implements the transfer
  2091 principle of nonstandard analysis. With a subgoal that mentions
  2092 nonstandard types like "'a star", the command "apply transfer"
  2093 replaces it with an equivalent one that mentions only standard types.
  2094 To be successful, all free variables must have standard types; non-
  2095 standard variables must have explicit universal quantifiers.
  2096 
  2097 * Hyperreal: A theory of Taylor series.
  2098 
  2099 
  2100 *** HOLCF ***
  2101 
  2102 * Discontinued special version of 'constdefs' (which used to support
  2103 continuous functions) in favor of the general Pure one with full
  2104 type-inference.
  2105 
  2106 * New simplification procedure for solving continuity conditions; it
  2107 is much faster on terms with many nested lambda abstractions (cubic
  2108 instead of exponential time).
  2109 
  2110 * New syntax for domain package: selector names are now optional.
  2111 Parentheses should be omitted unless argument is lazy, for example:
  2112 
  2113   domain 'a stream = cons "'a" (lazy "'a stream")
  2114 
  2115 * New command 'fixrec' for defining recursive functions with pattern
  2116 matching; defining multiple functions with mutual recursion is also
  2117 supported.  Patterns may include the constants cpair, spair, up, sinl,
  2118 sinr, or any data constructor defined by the domain package. The given
  2119 equations are proven as rewrite rules. See HOLCF/ex/Fixrec_ex.thy for
  2120 syntax and examples.
  2121 
  2122 * New commands 'cpodef' and 'pcpodef' for defining predicate subtypes
  2123 of cpo and pcpo types. Syntax is exactly like the 'typedef' command,
  2124 but the proof obligation additionally includes an admissibility
  2125 requirement. The packages generate instances of class cpo or pcpo,
  2126 with continuity and strictness theorems for Rep and Abs.
  2127 
  2128 * HOLCF: Many theorems have been renamed according to a more standard naming
  2129 scheme (INCOMPATIBILITY):
  2130 
  2131   foo_inject:  "foo$x = foo$y ==> x = y"
  2132   foo_eq:      "(foo$x = foo$y) = (x = y)"
  2133   foo_less:    "(foo$x << foo$y) = (x << y)"
  2134   foo_strict:  "foo$UU = UU"
  2135   foo_defined: "... ==> foo$x ~= UU"
  2136   foo_defined_iff: "(foo$x = UU) = (x = UU)"
  2137 
  2138 
  2139 *** ZF ***
  2140 
  2141 * ZF/ex: theories Group and Ring provide examples in abstract algebra,
  2142 including the First Isomorphism Theorem (on quotienting by the kernel
  2143 of a homomorphism).
  2144 
  2145 * ZF/Simplifier: install second copy of type solver that actually
  2146 makes use of TC rules declared to Isar proof contexts (or locales);
  2147 the old version is still required for ML proof scripts.
  2148 
  2149 
  2150 *** Cube ***
  2151 
  2152 * Converted to Isar theory format; use locales instead of axiomatic
  2153 theories.
  2154 
  2155 
  2156 *** ML ***
  2157 
  2158 * Pure/library.ML: added ##>, ##>>, #>> -- higher-order counterparts
  2159 for ||>, ||>>, |>>,
  2160 
  2161 * Pure/library.ML no longer defines its own option datatype, but uses
  2162 that of the SML basis, which has constructors NONE and SOME instead of
  2163 None and Some, as well as exception Option.Option instead of OPTION.
  2164 The functions the, if_none, is_some, is_none have been adapted
  2165 accordingly, while Option.map replaces apsome.
  2166 
  2167 * Pure/library.ML: the exception LIST has been given up in favour of
  2168 the standard exceptions Empty and Subscript, as well as
  2169 Library.UnequalLengths.  Function like Library.hd and Library.tl are
  2170 superceded by the standard hd and tl functions etc.
  2171 
  2172 A number of basic list functions are no longer exported to the ML
  2173 toplevel, as they are variants of predefined functions.  The following
  2174 suggests how one can translate existing code:
  2175 
  2176     rev_append xs ys = List.revAppend (xs, ys)
  2177     nth_elem (i, xs) = List.nth (xs, i)
  2178     last_elem xs = List.last xs
  2179     flat xss = List.concat xss
  2180     seq fs = List.app fs
  2181     partition P xs = List.partition P xs
  2182     mapfilter f xs = List.mapPartial f xs
  2183 
  2184 * Pure/library.ML: several combinators for linear functional
  2185 transformations, notably reverse application and composition:
  2186 
  2187   x |> f                f #> g
  2188   (x, y) |-> f          f #-> g
  2189 
  2190 * Pure/library.ML: introduced/changed precedence of infix operators:
  2191 
  2192   infix 1 |> |-> ||> ||>> |>> |>>> #> #->;
  2193   infix 2 ?;
  2194   infix 3 o oo ooo oooo;
  2195   infix 4 ~~ upto downto;
  2196 
  2197 Maybe INCOMPATIBILITY when any of those is used in conjunction with other
  2198 infix operators.
  2199 
  2200 * Pure/library.ML: natural list combinators fold, fold_rev, and
  2201 fold_map support linear functional transformations and nesting.  For
  2202 example:
  2203 
  2204   fold f [x1, ..., xN] y =
  2205     y |> f x1 |> ... |> f xN
  2206 
  2207   (fold o fold) f [xs1, ..., xsN] y =
  2208     y |> fold f xs1 |> ... |> fold f xsN
  2209 
  2210   fold f [x1, ..., xN] =
  2211     f x1 #> ... #> f xN
  2212 
  2213   (fold o fold) f [xs1, ..., xsN] =
  2214     fold f xs1 #> ... #> fold f xsN
  2215 
  2216 * Pure/library.ML: the following selectors on type 'a option are
  2217 available:
  2218 
  2219   the:               'a option -> 'a  (*partial*)
  2220   these:             'a option -> 'a  where 'a = 'b list
  2221   the_default: 'a -> 'a option -> 'a
  2222   the_list:          'a option -> 'a list
  2223 
  2224 * Pure/General: structure AList (cf. Pure/General/alist.ML) provides
  2225 basic operations for association lists, following natural argument
  2226 order; moreover the explicit equality predicate passed here avoids
  2227 potentially expensive polymorphic runtime equality checks.
  2228 The old functions may be expressed as follows:
  2229 
  2230   assoc = uncurry (AList.lookup (op =))
  2231   assocs = these oo AList.lookup (op =)
  2232   overwrite = uncurry (AList.update (op =)) o swap
  2233 
  2234 * Pure/General: structure AList (cf. Pure/General/alist.ML) provides
  2235 
  2236   val make: ('a -> 'b) -> 'a list -> ('a * 'b) list
  2237   val find: ('a * 'b -> bool) -> ('c * 'b) list -> 'a -> 'c list
  2238 
  2239 replacing make_keylist and keyfilter (occassionally used)
  2240 Naive rewrites:
  2241 
  2242   make_keylist = AList.make
  2243   keyfilter = AList.find (op =)
  2244 
  2245 * eq_fst and eq_snd now take explicit equality parameter, thus
  2246   avoiding eqtypes. Naive rewrites:
  2247 
  2248     eq_fst = eq_fst (op =)
  2249     eq_snd = eq_snd (op =)
  2250 
  2251 * Removed deprecated apl and apr (rarely used).
  2252   Naive rewrites:
  2253 
  2254     apl (n, op) =>>= curry op n
  2255     apr (op, m) =>>= fn n => op (n, m)
  2256 
  2257 * Pure/General: structure OrdList (cf. Pure/General/ord_list.ML)
  2258 provides a reasonably efficient light-weight implementation of sets as
  2259 lists.
  2260 
  2261 * Pure/General: generic tables (cf. Pure/General/table.ML) provide a
  2262 few new operations; existing lookup and update are now curried to
  2263 follow natural argument order (for use with fold etc.);
  2264 INCOMPATIBILITY, use (uncurry Symtab.lookup) etc. as last resort.
  2265 
  2266 * Pure/General: output via the Isabelle channels of
  2267 writeln/warning/error etc. is now passed through Output.output, with a
  2268 hook for arbitrary transformations depending on the print_mode
  2269 (cf. Output.add_mode -- the first active mode that provides a output
  2270 function wins).  Already formatted output may be embedded into further
  2271 text via Output.raw; the result of Pretty.string_of/str_of and derived
  2272 functions (string_of_term/cterm/thm etc.) is already marked raw to
  2273 accommodate easy composition of diagnostic messages etc.  Programmers
  2274 rarely need to care about Output.output or Output.raw at all, with
  2275 some notable exceptions: Output.output is required when bypassing the
  2276 standard channels (writeln etc.), or in token translations to produce
  2277 properly formatted results; Output.raw is required when capturing
  2278 already output material that will eventually be presented to the user
  2279 a second time.  For the default print mode, both Output.output and
  2280 Output.raw have no effect.
  2281 
  2282 * Pure/General: Output.time_accumulator NAME creates an operator ('a
  2283 -> 'b) -> 'a -> 'b to measure runtime and count invocations; the
  2284 cumulative results are displayed at the end of a batch session.
  2285 
  2286 * Pure/General: File.sysify_path and File.quote_sysify path have been
  2287 replaced by File.platform_path and File.shell_path (with appropriate
  2288 hooks).  This provides a clean interface for unusual systems where the
  2289 internal and external process view of file names are different.
  2290 
  2291 * Pure: more efficient orders for basic syntactic entities: added
  2292 fast_string_ord, fast_indexname_ord, fast_term_ord; changed sort_ord
  2293 and typ_ord to use fast_string_ord and fast_indexname_ord (term_ord is
  2294 NOT affected); structures Symtab, Vartab, Typtab, Termtab use the fast
  2295 orders now -- potential INCOMPATIBILITY for code that depends on a
  2296 particular order for Symtab.keys, Symtab.dest, etc. (consider using
  2297 Library.sort_strings on result).
  2298 
  2299 * Pure/term.ML: combinators fold_atyps, fold_aterms, fold_term_types,
  2300 fold_types traverse types/terms from left to right, observing natural
  2301 argument order.  Supercedes previous foldl_XXX versions, add_frees,
  2302 add_vars etc. have been adapted as well: INCOMPATIBILITY.
  2303 
  2304 * Pure: name spaces have been refined, with significant changes of the
  2305 internal interfaces -- INCOMPATIBILITY.  Renamed cond_extern(_table)
  2306 to extern(_table).  The plain name entry path is superceded by a
  2307 general 'naming' context, which also includes the 'policy' to produce
  2308 a fully qualified name and external accesses of a fully qualified
  2309 name; NameSpace.extend is superceded by context dependent
  2310 Sign.declare_name.  Several theory and proof context operations modify
  2311 the naming context.  Especially note Theory.restore_naming and
  2312 ProofContext.restore_naming to get back to a sane state; note that
  2313 Theory.add_path is no longer sufficient to recover from
  2314 Theory.absolute_path in particular.
  2315 
  2316 * Pure: new flags short_names (default false) and unique_names
  2317 (default true) for controlling output of qualified names.  If
  2318 short_names is set, names are printed unqualified.  If unique_names is
  2319 reset, the name prefix is reduced to the minimum required to achieve
  2320 the original result when interning again, even if there is an overlap
  2321 with earlier declarations.
  2322 
  2323 * Pure/TheoryDataFun: change of the argument structure; 'prep_ext' is
  2324 now 'extend', and 'merge' gets an additional Pretty.pp argument
  2325 (useful for printing error messages).  INCOMPATIBILITY.
  2326 
  2327 * Pure: major reorganization of the theory context.  Type Sign.sg and
  2328 Theory.theory are now identified, referring to the universal
  2329 Context.theory (see Pure/context.ML).  Actual signature and theory
  2330 content is managed as theory data.  The old code and interfaces were
  2331 spread over many files and structures; the new arrangement introduces
  2332 considerable INCOMPATIBILITY to gain more clarity:
  2333 
  2334   Context -- theory management operations (name, identity, inclusion,
  2335     parents, ancestors, merge, etc.), plus generic theory data;
  2336 
  2337   Sign -- logical signature and syntax operations (declaring consts,
  2338     types, etc.), plus certify/read for common entities;
  2339 
  2340   Theory -- logical theory operations (stating axioms, definitions,
  2341     oracles), plus a copy of logical signature operations (consts,
  2342     types, etc.); also a few basic management operations (Theory.copy,
  2343     Theory.merge, etc.)
  2344 
  2345 The most basic sign_of operations (Theory.sign_of, Thm.sign_of_thm
  2346 etc.) as well as the sign field in Thm.rep_thm etc. have been retained
  2347 for convenience -- they merely return the theory.
  2348 
  2349 * Pure: type Type.tsig is superceded by theory in most interfaces.
  2350 
  2351 * Pure: the Isar proof context type is already defined early in Pure
  2352 as Context.proof (note that ProofContext.context and Proof.context are
  2353 aliases, where the latter is the preferred name).  This enables other
  2354 Isabelle components to refer to that type even before Isar is present.
  2355 
  2356 * Pure/sign/theory: discontinued named name spaces (i.e. classK,
  2357 typeK, constK, axiomK, oracleK), but provide explicit operations for
  2358 any of these kinds.  For example, Sign.intern typeK is now
  2359 Sign.intern_type, Theory.hide_space Sign.typeK is now
  2360 Theory.hide_types.  Also note that former
  2361 Theory.hide_classes/types/consts are now
  2362 Theory.hide_classes_i/types_i/consts_i, while the non '_i' versions
  2363 internalize their arguments!  INCOMPATIBILITY.
  2364 
  2365 * Pure: get_thm interface (of PureThy and ProofContext) expects
  2366 datatype thmref (with constructors Name and NameSelection) instead of
  2367 plain string -- INCOMPATIBILITY;
  2368 
  2369 * Pure: cases produced by proof methods specify options, where NONE
  2370 means to remove case bindings -- INCOMPATIBILITY in
  2371 (RAW_)METHOD_CASES.
  2372 
  2373 * Pure: the following operations retrieve axioms or theorems from a
  2374 theory node or theory hierarchy, respectively:
  2375 
  2376   Theory.axioms_of: theory -> (string * term) list
  2377   Theory.all_axioms_of: theory -> (string * term) list
  2378   PureThy.thms_of: theory -> (string * thm) list
  2379   PureThy.all_thms_of: theory -> (string * thm) list
  2380 
  2381 * Pure: print_tac now outputs the goal through the trace channel.
  2382 
  2383 * Isar toplevel: improved diagnostics, mostly for Poly/ML only.
  2384 Reference Toplevel.debug (default false) controls detailed printing
  2385 and tracing of low-level exceptions; Toplevel.profiling (default 0)
  2386 controls execution profiling -- set to 1 for time and 2 for space
  2387 (both increase the runtime).
  2388 
  2389 * Isar session: The initial use of ROOT.ML is now always timed,
  2390 i.e. the log will show the actual process times, in contrast to the
  2391 elapsed wall-clock time that the outer shell wrapper produces.
  2392 
  2393 * Simplifier: improved handling of bound variables (nameless
  2394 representation, avoid allocating new strings).  Simprocs that invoke
  2395 the Simplifier recursively should use Simplifier.inherit_bounds to
  2396 avoid local name clashes.  Failure to do so produces warnings
  2397 "Simplifier: renamed bound variable ..."; set Simplifier.debug_bounds
  2398 for further details.
  2399 
  2400 * ML functions legacy_bindings and use_legacy_bindings produce ML fact
  2401 bindings for all theorems stored within a given theory; this may help
  2402 in porting non-Isar theories to Isar ones, while keeping ML proof
  2403 scripts for the time being.
  2404 
  2405 * ML operator HTML.with_charset specifies the charset begin used for
  2406 generated HTML files.  For example:
  2407 
  2408   HTML.with_charset "utf-8" use_thy "Hebrew";
  2409   HTML.with_charset "utf-8" use_thy "Chinese";
  2410 
  2411 
  2412 *** System ***
  2413 
  2414 * Allow symlinks to all proper Isabelle executables (Isabelle,
  2415 isabelle, isatool etc.).
  2416 
  2417 * ISABELLE_DOC_FORMAT setting specifies preferred document format (for
  2418 isatool doc, isatool mkdir, display_drafts etc.).
  2419 
  2420 * isatool usedir: option -f allows specification of the ML file to be
  2421 used by Isabelle; default is ROOT.ML.
  2422 
  2423 * New isatool version outputs the version identifier of the Isabelle
  2424 distribution being used.
  2425 
  2426 * HOL: new isatool dimacs2hol converts files in DIMACS CNF format
  2427 (containing Boolean satisfiability problems) into Isabelle/HOL
  2428 theories.
  2429 
  2430 
  2431 
  2432 New in Isabelle2004 (April 2004)
  2433 --------------------------------
  2434 
  2435 *** General ***
  2436 
  2437 * Provers/order.ML:  new efficient reasoner for partial and linear orders.
  2438   Replaces linorder.ML.
  2439 
  2440 * Pure: Greek letters (except small lambda, \<lambda>), as well as Gothic
  2441   (\<aa>...\<zz>\<AA>...\<ZZ>), calligraphic (\<A>...\<Z>), and Euler
  2442   (\<a>...\<z>), are now considered normal letters, and can therefore
  2443   be used anywhere where an ASCII letter (a...zA...Z) has until
  2444   now. COMPATIBILITY: This obviously changes the parsing of some
  2445   terms, especially where a symbol has been used as a binder, say
  2446   '\<Pi>x. ...', which is now a type error since \<Pi>x will be parsed
  2447   as an identifier.  Fix it by inserting a space around former
  2448   symbols.  Call 'isatool fixgreek' to try to fix parsing errors in
  2449   existing theory and ML files.
  2450 
  2451 * Pure: Macintosh and Windows line-breaks are now allowed in theory files.
  2452 
  2453 * Pure: single letter sub/superscripts (\<^isub> and \<^isup>) are now
  2454   allowed in identifiers. Similar to Greek letters \<^isub> is now considered
  2455   a normal (but invisible) letter. For multiple letter subscripts repeat
  2456   \<^isub> like this: x\<^isub>1\<^isub>2.
  2457 
  2458 * Pure: There are now sub-/superscripts that can span more than one
  2459   character. Text between \<^bsub> and \<^esub> is set in subscript in
  2460   ProofGeneral and LaTeX, text between \<^bsup> and \<^esup> in
  2461   superscript. The new control characters are not identifier parts.
  2462 
  2463 * Pure: Control-symbols of the form \<^raw:...> will literally print the
  2464   content of "..." to the latex file instead of \isacntrl... . The "..."
  2465   may consist of any printable characters excluding the end bracket >.
  2466 
  2467 * Pure: Using new Isar command "finalconsts" (or the ML functions
  2468   Theory.add_finals or Theory.add_finals_i) it is now possible to
  2469   declare constants "final", which prevents their being given a definition
  2470   later.  It is useful for constants whose behaviour is fixed axiomatically
  2471   rather than definitionally, such as the meta-logic connectives.
  2472 
  2473 * Pure: 'instance' now handles general arities with general sorts
  2474   (i.e. intersections of classes),
  2475 
  2476 * Presentation: generated HTML now uses a CSS style sheet to make layout
  2477   (somewhat) independent of content. It is copied from lib/html/isabelle.css.
  2478   It can be changed to alter the colors/layout of generated pages.
  2479 
  2480 
  2481 *** Isar ***
  2482 
  2483 * Tactic emulation methods rule_tac, erule_tac, drule_tac, frule_tac,
  2484   cut_tac, subgoal_tac and thin_tac:
  2485   - Now understand static (Isar) contexts.  As a consequence, users of Isar
  2486     locales are no longer forced to write Isar proof scripts.
  2487     For details see Isar Reference Manual, paragraph 4.3.2: Further tactic
  2488     emulations.
  2489   - INCOMPATIBILITY: names of variables to be instantiated may no
  2490     longer be enclosed in quotes.  Instead, precede variable name with `?'.
  2491     This is consistent with the instantiation attribute "where".
  2492 
  2493 * Attributes "where" and "of":
  2494   - Now take type variables of instantiated theorem into account when reading
  2495     the instantiation string.  This fixes a bug that caused instantiated
  2496     theorems to have too special types in some circumstances.
  2497   - "where" permits explicit instantiations of type variables.
  2498 
  2499 * Calculation commands "moreover" and "also" no longer interfere with
  2500   current facts ("this"), admitting arbitrary combinations with "then"
  2501   and derived forms.
  2502 
  2503 * Locales:
  2504   - Goal statements involving the context element "includes" no longer
  2505     generate theorems with internal delta predicates (those ending on
  2506     "_axioms") in the premise.
  2507     Resolve particular premise with <locale>.intro to obtain old form.
  2508   - Fixed bug in type inference ("unify_frozen") that prevented mix of target
  2509     specification and "includes" elements in goal statement.
  2510   - Rule sets <locale>.intro and <locale>.axioms no longer declared as
  2511     [intro?] and [elim?] (respectively) by default.
  2512   - Experimental command for instantiation of locales in proof contexts:
  2513         instantiate <label>[<attrs>]: <loc>
  2514     Instantiates locale <loc> and adds all its theorems to the current context
  2515     taking into account their attributes.  Label and attrs are optional
  2516     modifiers, like in theorem declarations.  If present, names of
  2517     instantiated theorems are qualified with <label>, and the attributes
  2518     <attrs> are applied after any attributes these theorems might have already.
  2519       If the locale has assumptions, a chained fact of the form
  2520     "<loc> t1 ... tn" is expected from which instantiations of the parameters
  2521     are derived.  The command does not support old-style locales declared
  2522     with "locale (open)".
  2523       A few (very simple) examples can be found in FOL/ex/LocaleInst.thy.
  2524 
  2525 * HOL: Tactic emulation methods induct_tac and case_tac understand static
  2526   (Isar) contexts.
  2527 
  2528 
  2529 *** HOL ***
  2530 
  2531 * Proof import: new image HOL4 contains the imported library from
  2532   the HOL4 system with about 2500 theorems. It is imported by
  2533   replaying proof terms produced by HOL4 in Isabelle. The HOL4 image
  2534   can be used like any other Isabelle image.  See
  2535   HOL/Import/HOL/README for more information.
  2536 
  2537 * Simplifier:
  2538   - Much improved handling of linear and partial orders.
  2539     Reasoners for linear and partial orders are set up for type classes
  2540     "linorder" and "order" respectively, and are added to the default simpset
  2541     as solvers.  This means that the simplifier can build transitivity chains
  2542     to solve goals from the assumptions.
  2543   - INCOMPATIBILITY: old proofs break occasionally.  Typically, applications
  2544     of blast or auto after simplification become unnecessary because the goal
  2545     is solved by simplification already.
  2546 
  2547 * Numerics: new theory Ring_and_Field contains over 250 basic numerical laws,
  2548     all proved in axiomatic type classes for semirings, rings and fields.
  2549 
  2550 * Numerics:
  2551   - Numeric types (nat, int, and in HOL-Complex rat, real, complex, etc.) are
  2552     now formalized using the Ring_and_Field theory mentioned above.
  2553   - INCOMPATIBILITY: simplification and arithmetic behaves somewhat differently
  2554     than before, because now they are set up once in a generic manner.
  2555   - INCOMPATIBILITY: many type-specific arithmetic laws have gone.
  2556     Look for the general versions in Ring_and_Field (and Power if they concern
  2557     exponentiation).
  2558 
  2559 * Type "rat" of the rational numbers is now available in HOL-Complex.
  2560 
  2561 * Records:
  2562   - Record types are now by default printed with their type abbreviation
  2563     instead of the list of all field types. This can be configured via
  2564     the reference "print_record_type_abbr".
  2565   - Simproc "record_upd_simproc" for simplification of multiple updates added
  2566     (not enabled by default).
  2567   - Simproc "record_ex_sel_eq_simproc" to simplify EX x. sel r = x resp.
  2568     EX x. x = sel r to True (not enabled by default).
  2569   - Tactic "record_split_simp_tac" to split and simplify records added.
  2570 
  2571 * 'specification' command added, allowing for definition by
  2572   specification.  There is also an 'ax_specification' command that
  2573   introduces the new constants axiomatically.
  2574 
  2575 * arith(_tac) is now able to generate counterexamples for reals as well.
  2576 
  2577 * HOL-Algebra: new locale "ring" for non-commutative rings.
  2578 
  2579 * HOL-ex: InductiveInvariant_examples illustrates advanced recursive function
  2580   definitions, thanks to Sava Krsti\'{c} and John Matthews.
  2581 
  2582 * HOL-Matrix: a first theory for matrices in HOL with an application of
  2583   matrix theory to linear programming.
  2584 
  2585 * Unions and Intersections:
  2586   The latex output syntax of UN and INT has been changed
  2587   from "\Union x \in A. B" to "\Union_{x \in A} B"
  2588   i.e. the index formulae has become a subscript.
  2589   Similarly for "\Union x. B", and for \Inter instead of \Union.
  2590 
  2591 * Unions and Intersections over Intervals:
  2592   There is new short syntax "UN i<=n. A" for "UN i:{0..n}. A". There is
  2593   also an x-symbol version with subscripts "\<Union>\<^bsub>i <= n\<^esub>. A"
  2594   like in normal math, and corresponding versions for < and for intersection.
  2595 
  2596 * HOL/List: Ordering "lexico" is renamed "lenlex" and the standard
  2597   lexicographic dictonary ordering has been added as "lexord".
  2598 
  2599 * ML: the legacy theory structures Int and List have been removed. They had
  2600   conflicted with ML Basis Library structures having the same names.
  2601 
  2602 * 'refute' command added to search for (finite) countermodels.  Only works
  2603   for a fragment of HOL.  The installation of an external SAT solver is
  2604   highly recommended.  See "HOL/Refute.thy" for details.
  2605 
  2606 * 'quickcheck' command: Allows to find counterexamples by evaluating
  2607   formulae under an assignment of free variables to random values.
  2608   In contrast to 'refute', it can deal with inductive datatypes,
  2609   but cannot handle quantifiers. See "HOL/ex/Quickcheck_Examples.thy"
  2610   for examples.
  2611 
  2612 
  2613 *** HOLCF ***
  2614 
  2615 * Streams now come with concatenation and are part of the HOLCF image
  2616 
  2617 
  2618 
  2619 New in Isabelle2003 (May 2003)
  2620 ------------------------------
  2621 
  2622 *** General ***
  2623 
  2624 * Provers/simplifier:
  2625 
  2626   - Completely reimplemented method simp (ML: Asm_full_simp_tac):
  2627     Assumptions are now subject to complete mutual simplification,
  2628     not just from left to right. The simplifier now preserves
  2629     the order of assumptions.
  2630 
  2631     Potential INCOMPATIBILITY:
  2632 
  2633     -- simp sometimes diverges where the old version did
  2634        not, e.g. invoking simp on the goal
  2635 
  2636         [| P (f x); y = x; f x = f y |] ==> Q
  2637 
  2638        now gives rise to the infinite reduction sequence
  2639 
  2640         P(f x) --(f x = f y)--> P(f y) --(y = x)--> P(f x) --(f x = f y)--> ...
  2641 
  2642        Using "simp (asm_lr)" (ML: Asm_lr_simp_tac) instead often solves this
  2643        kind of problem.
  2644 
  2645     -- Tactics combining classical reasoner and simplification (such as auto)
  2646        are also affected by this change, because many of them rely on
  2647        simp. They may sometimes diverge as well or yield a different numbers
  2648        of subgoals. Try to use e.g. force, fastsimp, or safe instead of auto
  2649        in case of problems. Sometimes subsequent calls to the classical
  2650        reasoner will fail because a preceeding call to the simplifier too
  2651        eagerly simplified the goal, e.g. deleted redundant premises.
  2652 
  2653   - The simplifier trace now shows the names of the applied rewrite rules
  2654 
  2655   - You can limit the number of recursive invocations of the simplifier
  2656     during conditional rewriting (where the simplifie tries to solve the
  2657     conditions before applying the rewrite rule):
  2658     ML "simp_depth_limit := n"
  2659     where n is an integer. Thus you can force termination where previously
  2660     the simplifier would diverge.
  2661 
  2662   - Accepts free variables as head terms in congruence rules.  Useful in Isar.
  2663 
  2664   - No longer aborts on failed congruence proof.  Instead, the
  2665     congruence is ignored.
  2666 
  2667 * Pure: New generic framework for extracting programs from constructive
  2668   proofs. See HOL/Extraction.thy for an example instantiation, as well
  2669   as HOL/Extraction for some case studies.
  2670 
  2671 * Pure: The main goal of the proof state is no longer shown by default, only
  2672 the subgoals. This behaviour is controlled by a new flag.
  2673    PG menu: Isabelle/Isar -> Settings -> Show Main Goal
  2674 (ML: Proof.show_main_goal).
  2675 
  2676 * Pure: You can find all matching introduction rules for subgoal 1, i.e. all
  2677 rules whose conclusion matches subgoal 1:
  2678       PG menu: Isabelle/Isar -> Show me -> matching rules
  2679 The rules are ordered by how closely they match the subgoal.
  2680 In particular, rules that solve a subgoal outright are displayed first
  2681 (or rather last, the way they are printed).
  2682 (ML: ProofGeneral.print_intros())
  2683 
  2684 * Pure: New flag trace_unify_fail causes unification to print
  2685 diagnostic information (PG: in trace buffer) when it fails. This is
  2686 useful for figuring out why single step proofs like rule, erule or
  2687 assumption failed.
  2688 
  2689 * Pure: Locale specifications now produce predicate definitions
  2690 according to the body of text (covering assumptions modulo local
  2691 definitions); predicate "loc_axioms" covers newly introduced text,
  2692 while "loc" is cumulative wrt. all included locale expressions; the
  2693 latter view is presented only on export into the global theory
  2694 context; potential INCOMPATIBILITY, use "(open)" option to fall back
  2695 on the old view without predicates;
  2696 
  2697 * Pure: predefined locales "var" and "struct" are useful for sharing
  2698 parameters (as in CASL, for example); just specify something like
  2699 ``var x + var y + struct M'' as import;
  2700 
  2701 * Pure: improved thms_containing: proper indexing of facts instead of
  2702 raw theorems; check validity of results wrt. current name space;
  2703 include local facts of proof configuration (also covers active
  2704 locales), cover fixed variables in index; may use "_" in term
  2705 specification; an optional limit for the number of printed facts may
  2706 be given (the default is 40);
  2707 
  2708 * Pure: disallow duplicate fact bindings within new-style theory files
  2709 (batch-mode only);
  2710 
  2711 * Provers: improved induct method: assumptions introduced by case
  2712 "foo" are split into "foo.hyps" (from the rule) and "foo.prems" (from
  2713 the goal statement); "foo" still refers to all facts collectively;
  2714 
  2715 * Provers: the function blast.overloaded has been removed: all constants
  2716 are regarded as potentially overloaded, which improves robustness in exchange
  2717 for slight decrease in efficiency;
  2718 
  2719 * Provers/linorder: New generic prover for transitivity reasoning over
  2720 linear orders.  Note: this prover is not efficient!
  2721 
  2722 * Isar: preview of problems to finish 'show' now produce an error
  2723 rather than just a warning (in interactive mode);
  2724 
  2725 
  2726 *** HOL ***
  2727 
  2728 * arith(_tac)
  2729 
  2730  - Produces a counter example if it cannot prove a goal.
  2731    Note that the counter example may be spurious if the goal is not a formula
  2732    of quantifier-free linear arithmetic.
  2733    In ProofGeneral the counter example appears in the trace buffer.
  2734 
  2735  - Knows about div k and mod k where k is a numeral of type nat or int.
  2736 
  2737  - Calls full Presburger arithmetic (by Amine Chaieb) if quantifier-free
  2738    linear arithmetic fails. This takes account of quantifiers and divisibility.
  2739    Presburger arithmetic can also be called explicitly via presburger(_tac).
  2740 
  2741 * simp's arithmetic capabilities have been enhanced a bit: it now
  2742 takes ~= in premises into account (by performing a case split);
  2743 
  2744 * simp reduces "m*(n div m) + n mod m" to n, even if the two summands
  2745 are distributed over a sum of terms;
  2746 
  2747 * New tactic "trans_tac" and method "trans" instantiate
  2748 Provers/linorder.ML for axclasses "order" and "linorder" (predicates
  2749 "<=", "<" and "=").
  2750 
  2751 * function INCOMPATIBILITIES: Pi-sets have been redefined and moved from main
  2752 HOL to Library/FuncSet; constant "Fun.op o" is now called "Fun.comp";
  2753 
  2754 * 'typedef' command has new option "open" to suppress the set
  2755 definition;
  2756 
  2757 * functions Min and Max on finite sets have been introduced (theory
  2758 Finite_Set);
  2759 
  2760 * attribute [symmetric] now works for relations as well; it turns
  2761 (x,y) : R^-1 into (y,x) : R, and vice versa;
  2762 
  2763 * induct over a !!-quantified statement (say !!x1..xn):
  2764   each "case" automatically performs "fix x1 .. xn" with exactly those names.
  2765 
  2766 * Map: `empty' is no longer a constant but a syntactic abbreviation for
  2767 %x. None. Warning: empty_def now refers to the previously hidden definition
  2768 of the empty set.
  2769 
  2770 * Algebra: formalization of classical algebra.  Intended as base for
  2771 any algebraic development in Isabelle.  Currently covers group theory
  2772 (up to Sylow's theorem) and ring theory (Universal Property of
  2773 Univariate Polynomials).  Contributions welcome;
  2774 
  2775 * GroupTheory: deleted, since its material has been moved to Algebra;
  2776 
  2777 * Complex: new directory of the complex numbers with numeric constants,
  2778 nonstandard complex numbers, and some complex analysis, standard and
  2779 nonstandard (Jacques Fleuriot);
  2780 
  2781 * HOL-Complex: new image for analysis, replacing HOL-Real and HOL-Hyperreal;
  2782 
  2783 * Hyperreal: introduced Gauge integration and hyperreal logarithms (Jacques
  2784 Fleuriot);
  2785 
  2786 * Real/HahnBanach: updated and adapted to locales;
  2787 
  2788 * NumberTheory: added Gauss's law of quadratic reciprocity (by Avigad,
  2789 Gray and Kramer);
  2790 
  2791 * UNITY: added the Meier-Sanders theory of progress sets;
  2792 
  2793 * MicroJava: bytecode verifier and lightweight bytecode verifier
  2794 as abstract algorithms, instantiated to the JVM;
  2795 
  2796 * Bali: Java source language formalization. Type system, operational
  2797 semantics, axiomatic semantics. Supported language features:
  2798 classes, interfaces, objects,virtual methods, static methods,
  2799 static/instance fields, arrays, access modifiers, definite
  2800 assignment, exceptions.
  2801 
  2802 
  2803 *** ZF ***
  2804 
  2805 * ZF/Constructible: consistency proof for AC (Gdel's constructible
  2806 universe, etc.);
  2807 
  2808 * Main ZF: virtually all theories converted to new-style format;
  2809 
  2810 
  2811 *** ML ***
  2812 
  2813 * Pure: Tactic.prove provides sane interface for internal proofs;
  2814 omits the infamous "standard" operation, so this is more appropriate
  2815 than prove_goalw_cterm in many situations (e.g. in simprocs);
  2816 
  2817 * Pure: improved error reporting of simprocs;
  2818 
  2819 * Provers: Simplifier.simproc(_i) provides sane interface for setting
  2820 up simprocs;
  2821 
  2822 
  2823 *** Document preparation ***
  2824 
  2825 * uses \par instead of \\ for line breaks in theory text. This may
  2826 shift some page breaks in large documents. To get the old behaviour
  2827 use \renewcommand{\isanewline}{\mbox{}\\\mbox{}} in root.tex.
  2828 
  2829 * minimized dependencies of isabelle.sty and isabellesym.sty on
  2830 other packages
  2831 
  2832 * \<euro> now needs package babel/greek instead of marvosym (which
  2833 broke \Rightarrow)
  2834 
  2835 * normal size for \<zero>...\<nine> (uses \mathbf instead of
  2836 textcomp package)
  2837 
  2838 
  2839 
  2840 New in Isabelle2002 (March 2002)
  2841 --------------------------------
  2842 
  2843 *** Document preparation ***
  2844 
  2845 * greatly simplified document preparation setup, including more
  2846 graceful interpretation of isatool usedir -i/-d/-D options, and more
  2847 instructive isatool mkdir; users should basically be able to get
  2848 started with "isatool mkdir HOL Test && isatool make"; alternatively,
  2849 users may run a separate document processing stage manually like this:
  2850 "isatool usedir -D output HOL Test && isatool document Test/output";
  2851 
  2852 * theory dependency graph may now be incorporated into documents;
  2853 isatool usedir -g true will produce session_graph.eps/.pdf for use
  2854 with \includegraphics of LaTeX;
  2855 
  2856 * proper spacing of consecutive markup elements, especially text
  2857 blocks after section headings;
  2858 
  2859 * support bold style (for single symbols only), input syntax is like
  2860 this: "\<^bold>\<alpha>" or "\<^bold>A";
  2861 
  2862 * \<bullet> is now output as bold \cdot by default, which looks much
  2863 better in printed text;
  2864 
  2865 * added default LaTeX bindings for \<tturnstile> and \<TTurnstile>;
  2866 note that these symbols are currently unavailable in Proof General /
  2867 X-Symbol; new symbols \<zero>, \<one>, ..., \<nine>, and \<euro>;
  2868 
  2869 * isatool latex no longer depends on changed TEXINPUTS, instead
  2870 isatool document copies the Isabelle style files to the target
  2871 location;
  2872 
  2873 
  2874 *** Isar ***
  2875 
  2876 * Pure/Provers: improved proof by cases and induction;
  2877   - 'case' command admits impromptu naming of parameters (such as
  2878     "case (Suc n)");
  2879   - 'induct' method divinates rule instantiation from the inductive
  2880     claim; no longer requires excessive ?P bindings for proper
  2881     instantiation of cases;
  2882   - 'induct' method properly enumerates all possibilities of set/type
  2883     rules; as a consequence facts may be also passed through *type*
  2884     rules without further ado;
  2885   - 'induct' method now derives symbolic cases from the *rulified*
  2886     rule (before it used to rulify cases stemming from the internal
  2887     atomized version); this means that the context of a non-atomic
  2888     statement becomes is included in the hypothesis, avoiding the
  2889     slightly cumbersome show "PROP ?case" form;
  2890   - 'induct' may now use elim-style induction rules without chaining
  2891     facts, using ``missing'' premises from the goal state; this allows
  2892     rules stemming from inductive sets to be applied in unstructured
  2893     scripts, while still benefitting from proper handling of non-atomic
  2894     statements; NB: major inductive premises need to be put first, all
  2895     the rest of the goal is passed through the induction;
  2896   - 'induct' proper support for mutual induction involving non-atomic
  2897     rule statements (uses the new concept of simultaneous goals, see
  2898     below);
  2899   - append all possible rule selections, but only use the first
  2900     success (no backtracking);
  2901   - removed obsolete "(simplified)" and "(stripped)" options of methods;
  2902   - undeclared rule case names default to numbers 1, 2, 3, ...;
  2903   - added 'print_induct_rules' (covered by help item in recent Proof
  2904     General versions);
  2905   - moved induct/cases attributes to Pure, methods to Provers;
  2906   - generic method setup instantiated for FOL and HOL;
  2907 
  2908 * Pure: support multiple simultaneous goal statements, for example
  2909 "have a: A and b: B" (same for 'theorem' etc.); being a pure
  2910 meta-level mechanism, this acts as if several individual goals had
  2911 been stated separately; in particular common proof methods need to be
  2912 repeated in order to cover all claims; note that a single elimination
  2913 step is *not* sufficient to establish the two conjunctions, so this
  2914 fails:
  2915 
  2916   assume "A & B" then have A and B ..   (*".." fails*)
  2917 
  2918 better use "obtain" in situations as above; alternative refer to
  2919 multi-step methods like 'auto', 'simp_all', 'blast+' etc.;
  2920 
  2921 * Pure: proper integration with ``locales''; unlike the original
  2922 version by Florian Kammller, Isar locales package high-level proof
  2923 contexts rather than raw logical ones (e.g. we admit to include
  2924 attributes everywhere); operations on locales include merge and
  2925 rename; support for implicit arguments (``structures''); simultaneous
  2926 type-inference over imports and text; see also HOL/ex/Locales.thy for
  2927 some examples;
  2928 
  2929 * Pure: the following commands have been ``localized'', supporting a
  2930 target locale specification "(in name)": 'lemma', 'theorem',
  2931 'corollary', 'lemmas', 'theorems', 'declare'; the results will be
  2932 stored both within the locale and at the theory level (exported and
  2933 qualified by the locale name);
  2934 
  2935 * Pure: theory goals may now be specified in ``long'' form, with
  2936 ad-hoc contexts consisting of arbitrary locale elements. for example
  2937 ``lemma foo: fixes x assumes "A x" shows "B x"'' (local syntax and
  2938 definitions may be given, too); the result is a meta-level rule with
  2939 the context elements being discharged in the obvious way;
  2940 
  2941 * Pure: new proof command 'using' allows to augment currently used
  2942 facts after a goal statement ('using' is syntactically analogous to
  2943 'apply', but acts on the goal's facts only); this allows chained facts
  2944 to be separated into parts given before and after a claim, as in
  2945 ``from a and b have C using d and e <proof>'';
  2946 
  2947 * Pure: renamed "antecedent" case to "rule_context";
  2948 
  2949 * Pure: new 'judgment' command records explicit information about the
  2950 object-logic embedding (used by several tools internally); no longer
  2951 use hard-wired "Trueprop";
  2952 
  2953 * Pure: added 'corollary' command;
  2954 
  2955 * Pure: fixed 'token_translation' command;
  2956 
  2957 * Pure: removed obsolete 'exported' attribute;
  2958 
  2959 * Pure: dummy pattern "_" in is/let is now automatically lifted over
  2960 bound variables: "ALL x. P x --> Q x" (is "ALL x. _ --> ?C x")
  2961 supersedes more cumbersome ... (is "ALL x. _ x --> ?C x");
  2962 
  2963 * Pure: method 'atomize' presents local goal premises as object-level
  2964 statements (atomic meta-level propositions); setup controlled via
  2965 rewrite rules declarations of 'atomize' attribute; example
  2966 application: 'induct' method with proper rule statements in improper
  2967 proof *scripts*;
  2968 
  2969 * Pure: emulation of instantiation tactics (rule_tac, cut_tac, etc.)
  2970 now consider the syntactic context of assumptions, giving a better
  2971 chance to get type-inference of the arguments right (this is
  2972 especially important for locales);
  2973 
  2974 * Pure: "sorry" no longer requires quick_and_dirty in interactive
  2975 mode;
  2976 
  2977 * Pure/obtain: the formal conclusion "thesis", being marked as
  2978 ``internal'', may no longer be reference directly in the text;
  2979 potential INCOMPATIBILITY, may need to use "?thesis" in rare
  2980 situations;
  2981 
  2982 * Pure: generic 'sym' attribute which declares a rule both as pure
  2983 'elim?' and for the 'symmetric' operation;
  2984 
  2985 * Pure: marginal comments ``--'' may now occur just anywhere in the
  2986 text; the fixed correlation with particular command syntax has been
  2987 discontinued;
  2988 
  2989 * Pure: new method 'rules' is particularly well-suited for proof
  2990 search in intuitionistic logic; a bit slower than 'blast' or 'fast',
  2991 but often produces more compact proof terms with less detours;
  2992 
  2993 * Pure/Provers/classical: simplified integration with pure rule
  2994 attributes and methods; the classical "intro?/elim?/dest?"
  2995 declarations coincide with the pure ones; the "rule" method no longer
  2996 includes classically swapped intros; "intro" and "elim" methods no
  2997 longer pick rules from the context; also got rid of ML declarations
  2998 AddXIs/AddXEs/AddXDs; all of this has some potential for
  2999 INCOMPATIBILITY;
  3000 
  3001 * Provers/classical: attribute 'swapped' produces classical inversions
  3002 of introduction rules;
  3003 
  3004 * Provers/simplifier: 'simplified' attribute may refer to explicit
  3005 rules instead of full simplifier context; 'iff' attribute handles
  3006 conditional rules;
  3007 
  3008 * HOL: 'typedef' now allows alternative names for Rep/Abs morphisms;
  3009 
  3010 * HOL: 'recdef' now fails on unfinished automated proofs, use
  3011 "(permissive)" option to recover old behavior;
  3012 
  3013 * HOL: 'inductive' no longer features separate (collective) attributes
  3014 for 'intros' (was found too confusing);
  3015 
  3016 * HOL: properly declared induction rules less_induct and
  3017 wf_induct_rule;
  3018 
  3019 
  3020 *** HOL ***
  3021 
  3022 * HOL: moved over to sane numeral syntax; the new policy is as
  3023 follows:
  3024 
  3025   - 0 and 1 are polymorphic constants, which are defined on any
  3026   numeric type (nat, int, real etc.);
  3027 
  3028   - 2, 3, 4, ... and -1, -2, -3, ... are polymorphic numerals, based
  3029   binary representation internally;
  3030 
  3031   - type nat has special constructor Suc, and generally prefers Suc 0
  3032   over 1::nat and Suc (Suc 0) over 2::nat;
  3033 
  3034 This change may cause significant problems of INCOMPATIBILITY; here
  3035 are some hints on converting existing sources:
  3036 
  3037   - due to the new "num" token, "-0" and "-1" etc. are now atomic
  3038   entities, so expressions involving "-" (unary or binary minus) need
  3039   to be spaced properly;
  3040 
  3041   - existing occurrences of "1" may need to be constraint "1::nat" or
  3042   even replaced by Suc 0; similar for old "2";
  3043 
  3044   - replace "#nnn" by "nnn", and "#-nnn" by "-nnn";
  3045 
  3046   - remove all special provisions on numerals in proofs;
  3047 
  3048 * HOL: simp rules nat_number expand numerals on nat to Suc/0
  3049 representation (depends on bin_arith_simps in the default context);
  3050 
  3051 * HOL: symbolic syntax for x^2 (numeral 2);
  3052 
  3053 * HOL: the class of all HOL types is now called "type" rather than
  3054 "term"; INCOMPATIBILITY, need to adapt references to this type class
  3055 in axclass/classes, instance/arities, and (usually rare) occurrences
  3056 in typings (of consts etc.); internally the class is called
  3057 "HOL.type", ML programs should refer to HOLogic.typeS;
  3058 
  3059 * HOL/record package improvements:
  3060   - new derived operations "fields" to build a partial record section,
  3061     "extend" to promote a fixed record to a record scheme, and
  3062     "truncate" for the reverse; cf. theorems "xxx.defs", which are *not*
  3063     declared as simp by default;
  3064   - shared operations ("more", "fields", etc.) now need to be always
  3065     qualified) --- potential INCOMPATIBILITY;
  3066   - removed "make_scheme" operations (use "make" with "extend") --
  3067     INCOMPATIBILITY;
  3068   - removed "more" class (simply use "term") -- INCOMPATIBILITY;
  3069   - provides cases/induct rules for use with corresponding Isar
  3070     methods (for concrete records, record schemes, concrete more
  3071     parts, and schematic more parts -- in that order);
  3072   - internal definitions directly based on a light-weight abstract
  3073     theory of product types over typedef rather than datatype;
  3074 
  3075 * HOL: generic code generator for generating executable ML code from
  3076 specifications; specific support for HOL constructs such as inductive
  3077 datatypes and sets, as well as recursive functions; can be invoked
  3078 via 'generate_code' theory section;
  3079 
  3080 * HOL: canonical cases/induct rules for n-tuples (n = 3..7);
  3081 
  3082 * HOL: consolidated and renamed several theories.  In particular:
  3083         Ord.thy has been absorbed into HOL.thy
  3084         String.thy has been absorbed into List.thy
  3085 
  3086 * HOL: concrete setsum syntax "\<Sum>i:A. b" == "setsum (%i. b) A"
  3087 (beware of argument permutation!);
  3088 
  3089 * HOL: linorder_less_split superseded by linorder_cases;
  3090 
  3091 * HOL/List: "nodups" renamed to "distinct";
  3092 
  3093 * HOL: added "The" definite description operator; move Hilbert's "Eps"
  3094 to peripheral theory "Hilbert_Choice"; some INCOMPATIBILITIES:
  3095   - Ex_def has changed, now need to use some_eq_ex
  3096 
  3097 * HOL: made split_all_tac safe; EXISTING PROOFS MAY FAIL OR LOOP, so
  3098 in this (rare) case use:
  3099 
  3100   delSWrapper "split_all_tac"
  3101   addSbefore ("unsafe_split_all_tac", unsafe_split_all_tac)
  3102 
  3103 * HOL: added safe wrapper "split_conv_tac" to claset; EXISTING PROOFS
  3104 MAY FAIL;
  3105 
  3106 * HOL: introduced f^n = f o ... o f; warning: due to the limits of
  3107 Isabelle's type classes, ^ on functions and relations has too general
  3108 a domain, namely ('a * 'b) set and 'a => 'b; this means that it may be
  3109 necessary to attach explicit type constraints;
  3110 
  3111 * HOL/Relation: the prefix name of the infix "O" has been changed from
  3112 "comp" to "rel_comp"; INCOMPATIBILITY: a few theorems have been
  3113 renamed accordingly (eg "compI" -> "rel_compI").
  3114 
  3115 * HOL: syntax translations now work properly with numerals and records
  3116 expressions;
  3117 
  3118 * HOL: bounded abstraction now uses syntax "%" / "\<lambda>" instead
  3119 of "lam" -- INCOMPATIBILITY;
  3120 
  3121 * HOL: got rid of some global declarations (potential INCOMPATIBILITY
  3122 for ML tools): const "()" renamed "Product_Type.Unity", type "unit"
  3123 renamed "Product_Type.unit";
  3124 
  3125 * HOL: renamed rtrancl_into_rtrancl2 to converse_rtrancl_into_rtrancl
  3126 
  3127 * HOL: removed obsolete theorem "optionE" (use "option.exhaust", or
  3128 the "cases" method);
  3129 
  3130 * HOL/GroupTheory: group theory examples including Sylow's theorem (by
  3131 Florian Kammller);
  3132 
  3133 * HOL/IMP: updated and converted to new-style theory format; several
  3134 parts turned into readable document, with proper Isar proof texts and
  3135 some explanations (by Gerwin Klein);
  3136 
  3137 * HOL-Real: added Complex_Numbers (by Gertrud Bauer);
  3138 
  3139 * HOL-Hyperreal is now a logic image;
  3140 
  3141 
  3142 *** HOLCF ***
  3143 
  3144 * Isar: consts/constdefs supports mixfix syntax for continuous
  3145 operations;
  3146 
  3147 * Isar: domain package adapted to new-style theory format, e.g. see
  3148 HOLCF/ex/Dnat.thy;
  3149 
  3150 * theory Lift: proper use of rep_datatype lift instead of ML hacks --
  3151 potential INCOMPATIBILITY; now use plain induct_tac instead of former
  3152 lift.induct_tac, always use UU instead of Undef;
  3153 
  3154 * HOLCF/IMP: updated and converted to new-style theory;
  3155 
  3156 
  3157 *** ZF ***
  3158 
  3159 * Isar: proper integration of logic-specific tools and packages,
  3160 including theory commands '(co)inductive', '(co)datatype',
  3161 'rep_datatype', 'inductive_cases', as well as methods 'ind_cases',
  3162 'induct_tac', 'case_tac', and 'typecheck' (with attribute 'TC');
  3163 
  3164 * theory Main no longer includes AC; for the Axiom of Choice, base
  3165 your theory on Main_ZFC;
  3166 
  3167 * the integer library now covers quotients and remainders, with many
  3168 laws relating division to addition, multiplication, etc.;
  3169 
  3170 * ZF/UNITY: Chandy and Misra's UNITY is now available in ZF, giving a
  3171 typeless version of the formalism;
  3172 
  3173 * ZF/AC, Coind, IMP, Resid: updated and converted to new-style theory
  3174 format;
  3175 
  3176 * ZF/Induct: new directory for examples of inductive definitions,
  3177 including theory Multiset for multiset orderings; converted to
  3178 new-style theory format;
  3179 
  3180 * ZF: many new theorems about lists, ordinals, etc.;
  3181 
  3182 
  3183 *** General ***
  3184 
  3185 * Pure/kernel: meta-level proof terms (by Stefan Berghofer); reference
  3186 variable proof controls level of detail: 0 = no proofs (only oracle
  3187 dependencies), 1 = lemma dependencies, 2 = compact proof terms; see
  3188 also ref manual for further ML interfaces;
  3189 
  3190 * Pure/axclass: removed obsolete ML interface
  3191 goal_subclass/goal_arity;
  3192 
  3193 * Pure/syntax: new token syntax "num" for plain numerals (without "#"
  3194 of "xnum"); potential INCOMPATIBILITY, since -0, -1 etc. are now
  3195 separate tokens, so expressions involving minus need to be spaced
  3196 properly;
  3197 
  3198 * Pure/syntax: support non-oriented infixes, using keyword "infix"
  3199 rather than "infixl" or "infixr";
  3200 
  3201 * Pure/syntax: concrete syntax for dummy type variables admits genuine
  3202 sort constraint specifications in type inference; e.g. "x::_::foo"
  3203 ensures that the type of "x" is of sort "foo" (but not necessarily a
  3204 type variable);
  3205 
  3206 * Pure/syntax: print modes "type_brackets" and "no_type_brackets"
  3207 control output of nested => (types); the default behavior is
  3208 "type_brackets";
  3209 
  3210 * Pure/syntax: builtin parse translation for "_constify" turns valued
  3211 tokens into AST constants;
  3212 
  3213 * Pure/syntax: prefer later declarations of translations and print
  3214 translation functions; potential INCOMPATIBILITY: need to reverse
  3215 multiple declarations for same syntax element constant;
  3216 
  3217 * Pure/show_hyps reset by default (in accordance to existing Isar
  3218 practice);
  3219 
  3220 * Provers/classical: renamed addaltern to addafter, addSaltern to
  3221 addSafter;
  3222 
  3223 * Provers/clasimp: ``iff'' declarations now handle conditional rules
  3224 as well;
  3225 
  3226 * system: tested support for MacOS X; should be able to get Isabelle +
  3227 Proof General to work in a plain Terminal after installing Poly/ML
  3228 (e.g. from the Isabelle distribution area) and GNU bash alone
  3229 (e.g. from http://www.apple.com); full X11, XEmacs and X-Symbol
  3230 support requires further installations, e.g. from
  3231 http://fink.sourceforge.net/);
  3232 
  3233 * system: support Poly/ML 4.1.1 (able to manage larger heaps);
  3234 
  3235 * system: reduced base memory usage by Poly/ML (approx. 20 MB instead
  3236 of 40 MB), cf. ML_OPTIONS;
  3237 
  3238 * system: Proof General keywords specification is now part of the
  3239 Isabelle distribution (see etc/isar-keywords.el);
  3240 
  3241 * system: support for persistent Proof General sessions (refrain from
  3242 outdating all loaded theories on startup); user may create writable
  3243 logic images like this: ``isabelle -q HOL Test'';
  3244 
  3245 * system: smart selection of Isabelle process versus Isabelle
  3246 interface, accommodates case-insensitive file systems (e.g. HFS+); may
  3247 run both "isabelle" and "Isabelle" even if file names are badly
  3248 damaged (executable inspects the case of the first letter of its own
  3249 name); added separate "isabelle-process" and "isabelle-interface";
  3250 
  3251 * system: refrain from any attempt at filtering input streams; no
  3252 longer support ``8bit'' encoding of old isabelle font, instead proper
  3253 iso-latin characters may now be used; the related isatools
  3254 "symbolinput" and "nonascii" have disappeared as well;
  3255 
  3256 * system: removed old "xterm" interface (the print modes "xterm" and
  3257 "xterm_color" are still available for direct use in a suitable
  3258 terminal);
  3259 
  3260 
  3261 
  3262 New in Isabelle99-2 (February 2001)
  3263 -----------------------------------
  3264 
  3265 *** Overview of INCOMPATIBILITIES ***
  3266 
  3267 * HOL: please note that theories in the Library and elsewhere often use the
  3268 new-style (Isar) format; to refer to their theorems in an ML script you must
  3269 bind them to ML identifers by e.g.      val thm_name = thm "thm_name";
  3270 
  3271 * HOL: inductive package no longer splits induction rule aggressively,
  3272 but only as far as specified by the introductions given; the old
  3273 format may be recovered via ML function complete_split_rule or attribute
  3274 'split_rule (complete)';
  3275 
  3276 * HOL: induct renamed to lfp_induct, lfp_Tarski to lfp_unfold,
  3277 gfp_Tarski to gfp_unfold;
  3278 
  3279 * HOL: contrapos, contrapos2 renamed to contrapos_nn, contrapos_pp;
  3280 
  3281 * HOL: infix "dvd" now has priority 50 rather than 70 (because it is a
  3282 relation); infix "^^" has been renamed "``"; infix "``" has been
  3283 renamed "`"; "univalent" has been renamed "single_valued";
  3284 
  3285 * HOL/Real: "rinv" and "hrinv" replaced by overloaded "inverse"
  3286 operation;
  3287 
  3288 * HOLCF: infix "`" has been renamed "$"; the symbol syntax is \<cdot>;
  3289 
  3290 * Isar: 'obtain' no longer declares "that" fact as simp/intro;
  3291 
  3292 * Isar/HOL: method 'induct' now handles non-atomic goals; as a
  3293 consequence, it is no longer monotonic wrt. the local goal context
  3294 (which is now passed through the inductive cases);
  3295 
  3296 * Document preparation: renamed standard symbols \<ll> to \<lless> and
  3297 \<gg> to \<ggreater>;
  3298 
  3299 
  3300 *** Document preparation ***
  3301 
  3302 * \isabellestyle{NAME} selects version of Isabelle output (currently
  3303 available: are "it" for near math-mode best-style output, "sl" for
  3304 slanted text style, and "tt" for plain type-writer; if no
  3305 \isabellestyle command is given, output is according to slanted
  3306 type-writer);
  3307 
  3308 * support sub/super scripts (for single symbols only), input syntax is
  3309 like this: "A\<^sup>*" or "A\<^sup>\<star>";
  3310 
  3311 * some more standard symbols; see Appendix A of the system manual for
  3312 the complete list of symbols defined in isabellesym.sty;
  3313 
  3314 * improved isabelle style files; more abstract symbol implementation
  3315 (should now use \isamath{...} and \isatext{...} in custom symbol
  3316 definitions);
  3317 
  3318 * antiquotation @{goals} and @{subgoals} for output of *dynamic* goals
  3319 state; Note that presentation of goal states does not conform to
  3320 actual human-readable proof documents.  Please do not include goal
  3321 states into document output unless you really know what you are doing!
  3322 
  3323 * proper indentation of antiquoted output with proportional LaTeX
  3324 fonts;
  3325 
  3326 * no_document ML operator temporarily disables LaTeX document
  3327 generation;
  3328 
  3329 * isatool unsymbolize tunes sources for plain ASCII communication;
  3330 
  3331 
  3332 *** Isar ***
  3333 
  3334 * Pure: Isar now suffers initial goal statements to contain unbound
  3335 schematic variables (this does not conform to actual readable proof
  3336 documents, due to unpredictable outcome and non-compositional proof
  3337 checking); users who know what they are doing may use schematic goals
  3338 for Prolog-style synthesis of proven results;
  3339 
  3340 * Pure: assumption method (an implicit finishing) now handles actual
  3341 rules as well;
  3342 
  3343 * Pure: improved 'obtain' --- moved to Pure, insert "that" into
  3344 initial goal, declare "that" only as Pure intro (only for single
  3345 steps); the "that" rule assumption may now be involved in implicit
  3346 finishing, thus ".." becomes a feasible for trivial obtains;
  3347 
  3348 * Pure: default proof step now includes 'intro_classes'; thus trivial
  3349 instance proofs may be performed by "..";
  3350 
  3351 * Pure: ?thesis / ?this / "..." now work for pure meta-level
  3352 statements as well;
  3353 
  3354 * Pure: more robust selection of calculational rules;
  3355 
  3356 * Pure: the builtin notion of 'finished' goal now includes the ==-refl
  3357 rule (as well as the assumption rule);
  3358 
  3359 * Pure: 'thm_deps' command visualizes dependencies of theorems and
  3360 lemmas, using the graph browser tool;
  3361 
  3362 * Pure: predict failure of "show" in interactive mode;
  3363 
  3364 * Pure: 'thms_containing' now takes actual terms as arguments;
  3365 
  3366 * HOL: improved method 'induct' --- now handles non-atomic goals
  3367 (potential INCOMPATIBILITY); tuned error handling;
  3368 
  3369 * HOL: cases and induct rules now provide explicit hints about the
  3370 number of facts to be consumed (0 for "type" and 1 for "set" rules);
  3371 any remaining facts are inserted into the goal verbatim;
  3372 
  3373 * HOL: local contexts (aka cases) may now contain term bindings as
  3374 well; the 'cases' and 'induct' methods new provide a ?case binding for
  3375 the result to be shown in each case;
  3376 
  3377 * HOL: added 'recdef_tc' command;
  3378 
  3379 * isatool convert assists in eliminating legacy ML scripts;
  3380 
  3381 
  3382 *** HOL ***
  3383 
  3384 * HOL/Library: a collection of generic theories to be used together
  3385 with main HOL; the theory loader path already includes this directory
  3386 by default; the following existing theories have been moved here:
  3387 HOL/Induct/Multiset, HOL/Induct/Acc (as Accessible_Part), HOL/While
  3388 (as While_Combinator), HOL/Lex/Prefix (as List_Prefix);
  3389 
  3390 * HOL/Unix: "Some aspects of Unix file-system security", a typical
  3391 modelling and verification task performed in Isabelle/HOL +
  3392 Isabelle/Isar + Isabelle document preparation (by Markus Wenzel).
  3393 
  3394 * HOL/Algebra: special summation operator SUM no longer exists, it has
  3395 been replaced by setsum; infix 'assoc' now has priority 50 (like
  3396 'dvd'); axiom 'one_not_zero' has been moved from axclass 'ring' to
  3397 'domain', this makes the theory consistent with mathematical
  3398 literature;
  3399 
  3400 * HOL basics: added overloaded operations "inverse" and "divide"
  3401 (infix "/"), syntax for generic "abs" operation, generic summation
  3402 operator \<Sum>;
  3403 
  3404 * HOL/typedef: simplified package, provide more useful rules (see also
  3405 HOL/subset.thy);
  3406 
  3407 * HOL/datatype: induction rule for arbitrarily branching datatypes is
  3408 now expressed as a proper nested rule (old-style tactic scripts may
  3409 require atomize_strip_tac to cope with non-atomic premises);
  3410 
  3411 * HOL: renamed theory "Prod" to "Product_Type", renamed "split" rule
  3412 to "split_conv" (old name still available for compatibility);
  3413 
  3414 * HOL: improved concrete syntax for strings (e.g. allows translation
  3415 rules with string literals);
  3416 
  3417 * HOL-Real-Hyperreal: this extends HOL-Real with the hyperreals
  3418  and Fleuriot's mechanization of analysis, including the transcendental
  3419  functions for the reals;
  3420 
  3421 * HOL/Real, HOL/Hyperreal: improved arithmetic simplification;
  3422 
  3423 
  3424 *** CTT ***
  3425 
  3426 * CTT: x-symbol support for Pi, Sigma, -->, : (membership); note that
  3427 "lam" is displayed as TWO lambda-symbols
  3428 
  3429 * CTT: theory Main now available, containing everything (that is, Bool
  3430 and Arith);
  3431 
  3432 
  3433 *** General ***
  3434 
  3435 * Pure: the Simplifier has been implemented properly as a derived rule
  3436 outside of the actual kernel (at last!); the overall performance
  3437 penalty in practical applications is about 50%, while reliability of
  3438 the Isabelle inference kernel has been greatly improved;
  3439 
  3440 * print modes "brackets" and "no_brackets" control output of nested =>
  3441 (types) and ==> (props); the default behaviour is "brackets";
  3442 
  3443 * Provers: fast_tac (and friends) now handle actual object-logic rules
  3444 as assumptions as well;
  3445 
  3446 * system: support Poly/ML 4.0;
  3447 
  3448 * system: isatool install handles KDE version 1 or 2;
  3449 
  3450 
  3451 
  3452 New in Isabelle99-1 (October 2000)
  3453 ----------------------------------
  3454 
  3455 *** Overview of INCOMPATIBILITIES ***
  3456 
  3457 * HOL: simplification of natural numbers is much changed; to partly
  3458 recover the old behaviour (e.g. to prevent n+n rewriting to #2*n)
  3459 issue the following ML commands:
  3460 
  3461   Delsimprocs Nat_Numeral_Simprocs.cancel_numerals;
  3462   Delsimprocs [Nat_Numeral_Simprocs.combine_numerals];
  3463 
  3464 * HOL: simplification no longer dives into case-expressions; this is
  3465 controlled by "t.weak_case_cong" for each datatype t;
  3466 
  3467 * HOL: nat_less_induct renamed to less_induct;
  3468 
  3469 * HOL: systematic renaming of the SOME (Eps) rules, may use isatool
  3470 fixsome to patch .thy and .ML sources automatically;
  3471 
  3472   select_equality  -> some_equality
  3473   select_eq_Ex     -> some_eq_ex
  3474   selectI2EX       -> someI2_ex
  3475   selectI2         -> someI2
  3476   selectI          -> someI
  3477   select1_equality -> some1_equality
  3478   Eps_sym_eq       -> some_sym_eq_trivial
  3479   Eps_eq           -> some_eq_trivial
  3480 
  3481 * HOL: exhaust_tac on datatypes superceded by new generic case_tac;
  3482 
  3483 * HOL: removed obsolete theorem binding expand_if (refer to split_if
  3484 instead);
  3485 
  3486 * HOL: the recursion equations generated by 'recdef' are now called
  3487 f.simps instead of f.rules;
  3488 
  3489 * HOL: qed_spec_mp now also handles bounded ALL as well;
  3490 
  3491 * HOL: 0 is now overloaded, so the type constraint ":: nat" may
  3492 sometimes be needed;
  3493 
  3494 * HOL: the constant for "f``x" is now "image" rather than "op ``";
  3495 
  3496 * HOL: the constant for "f-``x" is now "vimage" rather than "op -``";
  3497 
  3498 * HOL: the disjoint sum is now "<+>" instead of "Plus"; the cartesian
  3499 product is now "<*>" instead of "Times"; the lexicographic product is
  3500 now "<*lex*>" instead of "**";
  3501 
  3502 * HOL: theory Sexp is now in HOL/Induct examples (it used to be part
  3503 of main HOL, but was unused); better use HOL's datatype package;
  3504 
  3505 * HOL: removed "symbols" syntax for constant "override" of theory Map;
  3506 the old syntax may be recovered as follows:
  3507 
  3508   syntax (symbols)
  3509     override  :: "('a ~=> 'b) => ('a ~=> 'b) => ('a ~=> 'b)"
  3510       (infixl "\\<oplus>" 100)
  3511 
  3512 * HOL/Real: "rabs" replaced by overloaded "abs" function;
  3513 
  3514 * HOL/ML: even fewer consts are declared as global (see theories Ord,
  3515 Lfp, Gfp, WF); this only affects ML packages that refer to const names
  3516 internally;
  3517 
  3518 * HOL and ZF: syntax for quotienting wrt an equivalence relation
  3519 changed from A/r to A//r;
  3520 
  3521 * ZF: new treatment of arithmetic (nat & int) may break some old
  3522 proofs;
  3523 
  3524 * Isar: renamed some attributes (RS -> THEN, simplify -> simplified,
  3525 rulify -> rule_format, elimify -> elim_format, ...);
  3526 
  3527 * Isar/Provers: intro/elim/dest attributes changed; renamed
  3528 intro/intro!/intro!! flags to intro!/intro/intro? (in most cases, one
  3529 should have to change intro!! to intro? only); replaced "delrule" by
  3530 "rule del";
  3531 
  3532 * Isar/HOL: renamed "intrs" to "intros" in inductive definitions;
  3533 
  3534 * Provers: strengthened force_tac by using new first_best_tac;
  3535 
  3536 * LaTeX document preparation: several changes of isabelle.sty (see
  3537 lib/texinputs);
  3538 
  3539 
  3540 *** Document preparation ***
  3541 
  3542 * formal comments (text blocks etc.) in new-style theories may now
  3543 contain antiquotations of thm/prop/term/typ/text to be presented
  3544 according to latex print mode; concrete syntax is like this:
  3545 @{term[show_types] "f(x) = a + x"};
  3546 
  3547 * isatool mkdir provides easy setup of Isabelle session directories,
  3548 including proper document sources;
  3549 
  3550 * generated LaTeX sources are now deleted after successful run
  3551 (isatool document -c); may retain a copy somewhere else via -D option
  3552 of isatool usedir;
  3553 
  3554 * isatool usedir -D now lets isatool latex -o sty update the Isabelle
  3555 style files, achieving self-contained LaTeX sources and simplifying
  3556 LaTeX debugging;
  3557 
  3558 * old-style theories now produce (crude) LaTeX output as well;
  3559 
  3560 * browser info session directories are now self-contained (may be put
  3561 on WWW server seperately); improved graphs of nested sessions; removed
  3562 graph for 'all sessions';
  3563 
  3564 * several improvements in isabelle style files; \isabellestyle{it}
  3565 produces fake math mode output; \isamarkupheader is now \section by
  3566 default; see lib/texinputs/isabelle.sty etc.;
  3567 
  3568 
  3569 *** Isar ***
  3570 
  3571 * Isar/Pure: local results and corresponding term bindings are now
  3572 subject to Hindley-Milner polymorphism (similar to ML); this
  3573 accommodates incremental type-inference very nicely;
  3574 
  3575 * Isar/Pure: new derived language element 'obtain' supports
  3576 generalized existence reasoning;
  3577 
  3578 * Isar/Pure: new calculational elements 'moreover' and 'ultimately'
  3579 support accumulation of results, without applying any rules yet;
  3580 useful to collect intermediate results without explicit name
  3581 references, and for use with transitivity rules with more than 2
  3582 premises;
  3583 
  3584 * Isar/Pure: scalable support for case-analysis type proofs: new
  3585 'case' language element refers to local contexts symbolically, as
  3586 produced by certain proof methods; internally, case names are attached
  3587 to theorems as "tags";
  3588 
  3589 * Isar/Pure: theory command 'hide' removes declarations from
  3590 class/type/const name spaces;
  3591 
  3592 * Isar/Pure: theory command 'defs' supports option "(overloaded)" to
  3593 indicate potential overloading;
  3594 
  3595 * Isar/Pure: changed syntax of local blocks from {{ }} to { };
  3596 
  3597 * Isar/Pure: syntax of sorts made 'inner', i.e. have to write
  3598 "{a,b,c}" instead of {a,b,c};
  3599 
  3600 * Isar/Pure now provides its own version of intro/elim/dest
  3601 attributes; useful for building new logics, but beware of confusion
  3602 with the version in Provers/classical;
  3603 
  3604 * Isar/Pure: the local context of (non-atomic) goals is provided via
  3605 case name 'antecedent';
  3606 
  3607 * Isar/Pure: removed obsolete 'transfer' attribute (transfer of thms
  3608 to the current context is now done automatically);
  3609 
  3610 * Isar/Pure: theory command 'method_setup' provides a simple interface
  3611 for definining proof methods in ML;
  3612 
  3613 * Isar/Provers: intro/elim/dest attributes changed; renamed
  3614 intro/intro!/intro!! flags to intro!/intro/intro? (INCOMPATIBILITY, in
  3615 most cases, one should have to change intro!! to intro? only);
  3616 replaced "delrule" by "rule del";
  3617 
  3618 * Isar/Provers: new 'hypsubst' method, plain 'subst' method and
  3619 'symmetric' attribute (the latter supercedes [RS sym]);
  3620 
  3621 * Isar/Provers: splitter support (via 'split' attribute and 'simp'
  3622 method modifier); 'simp' method: 'only:' modifier removes loopers as
  3623 well (including splits);
  3624 
  3625 * Isar/Provers: Simplifier and Classical methods now support all kind
  3626 of modifiers used in the past, including 'cong', 'iff', etc.
  3627 
  3628 * Isar/Provers: added 'fastsimp' and 'clarsimp' methods (combination
  3629 of Simplifier and Classical reasoner);
  3630 
  3631 * Isar/HOL: new proof method 'cases' and improved version of 'induct'
  3632 now support named cases; major packages (inductive, datatype, primrec,
  3633 recdef) support case names and properly name parameters;
  3634 
  3635 * Isar/HOL: new transitivity rules for substitution in inequalities --
  3636 monotonicity conditions are extracted to be proven at end of
  3637 calculations;
  3638 
  3639 * Isar/HOL: removed 'case_split' thm binding, should use 'cases' proof
  3640 method anyway;
  3641 
  3642 * Isar/HOL: removed old expand_if = split_if; theorems if_splits =
  3643 split_if split_if_asm; datatype package provides theorems foo.splits =
  3644 foo.split foo.split_asm for each datatype;
  3645 
  3646 * Isar/HOL: tuned inductive package, rename "intrs" to "intros"
  3647 (potential INCOMPATIBILITY), emulation of mk_cases feature for proof
  3648 scripts: new 'inductive_cases' command and 'ind_cases' method; (Note:
  3649 use "(cases (simplified))" method in proper proof texts);
  3650 
  3651 * Isar/HOL: added global 'arith_split' attribute for 'arith' method;
  3652 
  3653 * Isar: names of theorems etc. may be natural numbers as well;
  3654 
  3655 * Isar: 'pr' command: optional arguments for goals_limit and
  3656 ProofContext.prems_limit; no longer prints theory contexts, but only
  3657 proof states;
  3658 
  3659 * Isar: diagnostic commands 'pr', 'thm', 'prop', 'term', 'typ' admit
  3660 additional print modes to be specified; e.g. "pr(latex)" will print
  3661 proof state according to the Isabelle LaTeX style;
  3662 
  3663 * Isar: improved support for emulating tactic scripts, including proof
  3664 methods 'rule_tac' etc., 'cut_tac', 'thin_tac', 'subgoal_tac',
  3665 'rename_tac', 'rotate_tac', 'tactic', and 'case_tac' / 'induct_tac'
  3666 (for HOL datatypes);
  3667 
  3668 * Isar: simplified (more robust) goal selection of proof methods: 1st
  3669 goal, all goals, or explicit goal specifier (tactic emulation); thus
  3670 'proof method scripts' have to be in depth-first order;
  3671 
  3672 * Isar: tuned 'let' syntax: replaced 'as' keyword by 'and';
  3673 
  3674 * Isar: removed 'help' command, which hasn't been too helpful anyway;
  3675 should instead use individual commands for printing items
  3676 (print_commands, print_methods etc.);
  3677 
  3678 * Isar: added 'nothing' --- the empty list of theorems;
  3679 
  3680 
  3681 *** HOL ***
  3682 
  3683 * HOL/MicroJava: formalization of a fragment of Java, together with a
  3684 corresponding virtual machine and a specification of its bytecode
  3685 verifier and a lightweight bytecode verifier, including proofs of
  3686 type-safety; by Gerwin Klein, Tobias Nipkow, David von Oheimb, and
  3687 Cornelia Pusch (see also the homepage of project Bali at
  3688 http://isabelle.in.tum.de/Bali/);
  3689 
  3690 * HOL/Algebra: new theory of rings and univariate polynomials, by
  3691 Clemens Ballarin;
  3692 
  3693 * HOL/NumberTheory: fundamental Theorem of Arithmetic, Chinese
  3694 Remainder Theorem, Fermat/Euler Theorem, Wilson's Theorem, by Thomas M
  3695 Rasmussen;
  3696 
  3697 * HOL/Lattice: fundamental concepts of lattice theory and order
  3698 structures, including duals, properties of bounds versus algebraic
  3699 laws, lattice operations versus set-theoretic ones, the Knaster-Tarski
  3700 Theorem for complete lattices etc.; may also serve as a demonstration
  3701 for abstract algebraic reasoning using axiomatic type classes, and
  3702 mathematics-style proof in Isabelle/Isar; by Markus Wenzel;
  3703 
  3704 * HOL/Prolog: a (bare-bones) implementation of Lambda-Prolog, by David
  3705 von Oheimb;
  3706 
  3707 * HOL/IMPP: extension of IMP with local variables and mutually
  3708 recursive procedures, by David von Oheimb;
  3709 
  3710 * HOL/Lambda: converted into new-style theory and document;
  3711 
  3712 * HOL/ex/Multiquote: example of multiple nested quotations and
  3713 anti-quotations -- basically a generalized version of de-Bruijn
  3714 representation; very useful in avoiding lifting of operations;
  3715 
  3716 * HOL/record: added general record equality rule to simpset; fixed
  3717 select-update simplification procedure to handle extended records as
  3718 well; admit "r" as field name;
  3719 
  3720 * HOL: 0 is now overloaded over the new sort "zero", allowing its use with
  3721 other numeric types and also as the identity of groups, rings, etc.;
  3722 
  3723 * HOL: new axclass plus_ac0 for addition with the AC-laws and 0 as identity.
  3724 Types nat and int belong to this axclass;
  3725 
  3726 * HOL: greatly improved simplification involving numerals of type nat, int, real:
  3727    (i + #8 + j) = Suc k simplifies to  #7 + (i + j) = k
  3728    i*j + k + j*#3*i     simplifies to  #4*(i*j) + k
  3729   two terms #m*u and #n*u are replaced by #(m+n)*u
  3730     (where #m, #n and u can implicitly be 1; this is simproc combine_numerals)
  3731   and the term/formula #m*u+x ~~ #n*u+y simplifies simplifies to #(m-n)+x ~~ y
  3732     or x ~~ #(n-m)+y, where ~~ is one of = < <= or - (simproc cancel_numerals);
  3733 
  3734 * HOL: meson_tac is available (previously in ex/meson.ML); it is a
  3735 powerful prover for predicate logic but knows nothing of clasets; see
  3736 ex/mesontest.ML and ex/mesontest2.ML for example applications;
  3737 
  3738 * HOL: new version of "case_tac" subsumes both boolean case split and
  3739 "exhaust_tac" on datatypes; INCOMPATIBILITY: exhaust_tac no longer
  3740 exists, may define val exhaust_tac = case_tac for ad-hoc portability;
  3741 
  3742 * HOL: simplification no longer dives into case-expressions: only the
  3743 selector expression is simplified, but not the remaining arms; to
  3744 enable full simplification of case-expressions for datatype t, you may
  3745 remove t.weak_case_cong from the simpset, either globally (Delcongs
  3746 [thm"t.weak_case_cong"];) or locally (delcongs [...]).
  3747 
  3748 * HOL/recdef: the recursion equations generated by 'recdef' for
  3749 function 'f' are now called f.simps instead of f.rules; if all
  3750 termination conditions are proved automatically, these simplification
  3751 rules are added to the simpset, as in primrec; rules may be named
  3752 individually as well, resulting in a separate list of theorems for
  3753 each equation;
  3754 
  3755 * HOL/While is a new theory that provides a while-combinator. It
  3756 permits the definition of tail-recursive functions without the
  3757 provision of a termination measure. The latter is necessary once the
  3758 invariant proof rule for while is applied.
  3759 
  3760 * HOL: new (overloaded) notation for the set of elements below/above
  3761 some element: {..u}, {..u(}, {l..}, {)l..}. See theory SetInterval.
  3762 
  3763 * HOL: theorems impI, allI, ballI bound as "strip";
  3764 
  3765 * HOL: new tactic induct_thm_tac: thm -> string -> int -> tactic
  3766 induct_tac th "x1 ... xn" expects th to have a conclusion of the form
  3767 P v1 ... vn and abbreviates res_inst_tac [("v1","x1"),...,("vn","xn")] th;
  3768 
  3769 * HOL/Real: "rabs" replaced by overloaded "abs" function;
  3770 
  3771 * HOL: theory Sexp now in HOL/Induct examples (it used to be part of
  3772 main HOL, but was unused);
  3773 
  3774 * HOL: fewer consts declared as global (e.g. have to refer to
  3775 "Lfp.lfp" instead of "lfp" internally; affects ML packages only);
  3776 
  3777 * HOL: tuned AST representation of nested pairs, avoiding bogus output
  3778 in case of overlap with user translations (e.g. judgements over
  3779 tuples); (note that the underlying logical represenation is still
  3780 bogus);
  3781 
  3782 
  3783 *** ZF ***
  3784 
  3785 * ZF: simplification automatically cancels common terms in arithmetic
  3786 expressions over nat and int;
  3787 
  3788 * ZF: new treatment of nat to minimize type-checking: all operators
  3789 coerce their operands to a natural number using the function natify,
  3790 making the algebraic laws unconditional;
  3791 
  3792 * ZF: as above, for int: operators coerce their operands to an integer
  3793 using the function intify;
  3794 
  3795 * ZF: the integer library now contains many of the usual laws for the
  3796 orderings, including $<=, and monotonicity laws for $+ and $*;
  3797 
  3798 * ZF: new example ZF/ex/NatSum to demonstrate integer arithmetic
  3799 simplification;
  3800 
  3801 * FOL and ZF: AddIffs now available, giving theorems of the form P<->Q
  3802 to the simplifier and classical reasoner simultaneously;
  3803 
  3804 
  3805 *** General ***
  3806 
  3807 * Provers: blast_tac now handles actual object-logic rules as
  3808 assumptions; note that auto_tac uses blast_tac internally as well;
  3809 
  3810 * Provers: new functions rulify/rulify_no_asm: thm -> thm for turning
  3811 outer -->/All/Ball into ==>/!!; qed_spec_mp now uses rulify_no_asm;
  3812 
  3813 * Provers: delrules now handles destruct rules as well (no longer need
  3814 explicit make_elim);
  3815 
  3816 * Provers: Blast_tac now warns of and ignores "weak elimination rules" e.g.
  3817   [| inj ?f;          ?f ?x = ?f ?y; ?x = ?y ==> ?W |] ==> ?W
  3818 use instead the strong form,
  3819   [| inj ?f; ~ ?W ==> ?f ?x = ?f ?y; ?x = ?y ==> ?W |] ==> ?W
  3820 in HOL, FOL and ZF the function cla_make_elim will create such rules
  3821 from destruct-rules;
  3822 
  3823 * Provers: Simplifier.easy_setup provides a fast path to basic
  3824 Simplifier setup for new object-logics;
  3825 
  3826 * Pure: AST translation rules no longer require constant head on LHS;
  3827 
  3828 * Pure: improved name spaces: ambiguous output is qualified; support
  3829 for hiding of names;
  3830 
  3831 * system: smart setup of canonical ML_HOME, ISABELLE_INTERFACE, and
  3832 XSYMBOL_HOME; no longer need to do manual configuration in most
  3833 situations;
  3834 
  3835 * system: compression of ML heaps images may now be controlled via -c
  3836 option of isabelle and isatool usedir (currently only observed by
  3837 Poly/ML);
  3838 
  3839 * system: isatool installfonts may handle X-Symbol fonts as well (very
  3840 useful for remote X11);
  3841 
  3842 * system: provide TAGS file for Isabelle sources;
  3843 
  3844 * ML: infix 'OF' is a version of 'MRS' with more appropriate argument
  3845 order;
  3846 
  3847 * ML: renamed flags Syntax.trace_norm_ast to Syntax.trace_ast; global
  3848 timing flag supersedes proof_timing and Toplevel.trace;
  3849 
  3850 * ML: new combinators |>> and |>>> for incremental transformations
  3851 with secondary results (e.g. certain theory extensions):
  3852 
  3853 * ML: PureThy.add_defs gets additional argument to indicate potential
  3854 overloading (usually false);
  3855 
  3856 * ML: PureThy.add_thms/add_axioms/add_defs now return theorems as
  3857 results;
  3858 
  3859 
  3860 
  3861 New in Isabelle99 (October 1999)
  3862 --------------------------------
  3863 
  3864 *** Overview of INCOMPATIBILITIES (see below for more details) ***
  3865 
  3866 * HOL: The THEN and ELSE parts of conditional expressions (if P then x else y)
  3867 are no longer simplified.  (This allows the simplifier to unfold recursive
  3868 functional programs.)  To restore the old behaviour, declare
  3869 
  3870     Delcongs [if_weak_cong];
  3871 
  3872 * HOL: Removed the obsolete syntax "Compl A"; use -A for set
  3873 complement;
  3874 
  3875 * HOL: the predicate "inj" is now defined by translation to "inj_on";
  3876 
  3877 * HOL/datatype: mutual_induct_tac no longer exists --
  3878   use induct_tac "x_1 ... x_n" instead of mutual_induct_tac ["x_1", ..., "x_n"]
  3879 
  3880 * HOL/typedef: fixed type inference for representing set; type
  3881 arguments now have to occur explicitly on the rhs as type constraints;
  3882 
  3883 * ZF: The con_defs part of an inductive definition may no longer refer
  3884 to constants declared in the same theory;
  3885 
  3886 * HOL, ZF: the function mk_cases, generated by the inductive
  3887 definition package, has lost an argument.  To simplify its result, it
  3888 uses the default simpset instead of a supplied list of theorems.
  3889 
  3890 * HOL/List: the constructors of type list are now Nil and Cons;
  3891 
  3892 * Simplifier: the type of the infix ML functions
  3893         setSSolver addSSolver setSolver addSolver
  3894 is now  simpset * solver -> simpset  where `solver' is a new abstract type
  3895 for packaging solvers. A solver is created via
  3896         mk_solver: string -> (thm list -> int -> tactic) -> solver
  3897 where the string argument is only a comment.
  3898 
  3899 
  3900 *** Proof tools ***
  3901 
  3902 * Provers/Arith/fast_lin_arith.ML contains a functor for creating a
  3903 decision procedure for linear arithmetic. Currently it is used for
  3904 types `nat', `int', and `real' in HOL (see below); it can, should and
  3905 will be instantiated for other types and logics as well.
  3906 
  3907 * The simplifier now accepts rewrite rules with flexible heads, eg
  3908      hom ?f ==> ?f(?x+?y) = ?f ?x + ?f ?y
  3909   They are applied like any rule with a non-pattern lhs, i.e. by first-order
  3910   matching.
  3911 
  3912 
  3913 *** General ***
  3914 
  3915 * New Isabelle/Isar subsystem provides an alternative to traditional
  3916 tactical theorem proving; together with the ProofGeneral/isar user
  3917 interface it offers an interactive environment for developing human
  3918 readable proof documents (Isar == Intelligible semi-automated
  3919 reasoning); for further information see isatool doc isar-ref,
  3920 src/HOL/Isar_examples and http://isabelle.in.tum.de/Isar/
  3921 
  3922 * improved and simplified presentation of theories: better HTML markup
  3923 (including colors), graph views in several sizes; isatool usedir now
  3924 provides a proper interface for user theories (via -P option); actual
  3925 document preparation based on (PDF)LaTeX is available as well (for
  3926 new-style theories only); see isatool doc system for more information;
  3927 
  3928 * native support for Proof General, both for classic Isabelle and
  3929 Isabelle/Isar;
  3930 
  3931 * ML function thm_deps visualizes dependencies of theorems and lemmas,
  3932 using the graph browser tool;
  3933 
  3934 * Isabelle manuals now also available as PDF;
  3935 
  3936 * theory loader rewritten from scratch (may not be fully
  3937 bug-compatible); old loadpath variable has been replaced by show_path,
  3938 add_path, del_path, reset_path functions; new operations such as
  3939 update_thy, touch_thy, remove_thy, use/update_thy_only (see also
  3940 isatool doc ref);
  3941 
  3942 * improved isatool install: option -k creates KDE application icon,
  3943 option -p DIR installs standalone binaries;
  3944 
  3945 * added ML_PLATFORM setting (useful for cross-platform installations);
  3946 more robust handling of platform specific ML images for SML/NJ;
  3947 
  3948 * the settings environment is now statically scoped, i.e. it is never
  3949 created again in sub-processes invoked from isabelle, isatool, or
  3950 Isabelle;
  3951 
  3952 * path element specification '~~' refers to '$ISABELLE_HOME';
  3953 
  3954 * in locales, the "assumes" and "defines" parts may be omitted if
  3955 empty;
  3956 
  3957 * new print_mode "xsymbols" for extended symbol support (e.g. genuine
  3958 long arrows);
  3959 
  3960 * new print_mode "HTML";
  3961 
  3962 * new flag show_tags controls display of tags of theorems (which are
  3963 basically just comments that may be attached by some tools);
  3964 
  3965 * Isamode 2.6 requires patch to accomodate change of Isabelle font
  3966 mode and goal output format:
  3967 
  3968 diff -r Isamode-2.6/elisp/isa-load.el Isamode/elisp/isa-load.el
  3969 244c244
  3970 <       (list (isa-getenv "ISABELLE") "-msymbols" logic-name)
  3971 ---
  3972 >       (list (isa-getenv "ISABELLE") "-misabelle_font" "-msymbols" logic-name)
  3973 diff -r Isabelle-2.6/elisp/isa-proofstate.el Isamode/elisp/isa-proofstate.el
  3974 181c181
  3975 < (defconst proofstate-proofstart-regexp "^Level [0-9]+$"
  3976 ---
  3977 > (defconst proofstate-proofstart-regexp "^Level [0-9]+"
  3978 
  3979 * function bind_thms stores lists of theorems (cf. bind_thm);
  3980 
  3981 * new shorthand tactics ftac, eatac, datac, fatac;
  3982 
  3983 * qed (and friends) now accept "" as result name; in that case the
  3984 theorem is not stored, but proper checks and presentation of the
  3985 result still apply;
  3986 
  3987 * theorem database now also indexes constants "Trueprop", "all",
  3988 "==>", "=="; thus thms_containing, findI etc. may retrieve more rules;
  3989 
  3990 
  3991 *** HOL ***
  3992 
  3993 ** HOL arithmetic **
  3994 
  3995 * There are now decision procedures for linear arithmetic over nat and
  3996 int:
  3997 
  3998 1. arith_tac copes with arbitrary formulae involving `=', `<', `<=',
  3999 `+', `-', `Suc', `min', `max' and numerical constants; other subterms
  4000 are treated as atomic; subformulae not involving type `nat' or `int'
  4001 are ignored; quantified subformulae are ignored unless they are
  4002 positive universal or negative existential. The tactic has to be
  4003 invoked by hand and can be a little bit slow. In particular, the
  4004 running time is exponential in the number of occurrences of `min' and
  4005 `max', and `-' on `nat'.
  4006 
  4007 2. fast_arith_tac is a cut-down version of arith_tac: it only takes
  4008 (negated) (in)equalities among the premises and the conclusion into
  4009 account (i.e. no compound formulae) and does not know about `min' and
  4010 `max', and `-' on `nat'. It is fast and is used automatically by the
  4011 simplifier.
  4012 
  4013 NB: At the moment, these decision procedures do not cope with mixed
  4014 nat/int formulae where the two parts interact, such as `m < n ==>
  4015 int(m) < int(n)'.
  4016 
  4017 * HOL/Numeral provides a generic theory of numerals (encoded
  4018 efficiently as bit strings); setup for types nat/int/real is in place;
  4019 INCOMPATIBILITY: since numeral syntax is now polymorphic, rather than
  4020 int, existing theories and proof scripts may require a few additional
  4021 type constraints;
  4022 
  4023 * integer division and remainder can now be performed on constant
  4024 arguments;
  4025 
  4026 * many properties of integer multiplication, division and remainder
  4027 are now available;
  4028 
  4029 * An interface to the Stanford Validity Checker (SVC) is available through the
  4030 tactic svc_tac.  Propositional tautologies and theorems of linear arithmetic
  4031 are proved automatically.  SVC must be installed separately, and its results
  4032 must be TAKEN ON TRUST (Isabelle does not check the proofs, but tags any
  4033 invocation of the underlying oracle).  For SVC see
  4034   http://verify.stanford.edu/SVC
  4035 
  4036 * IsaMakefile: the HOL-Real target now builds an actual image;
  4037 
  4038 
  4039 ** HOL misc **
  4040 
  4041 * HOL/Real/HahnBanach: the Hahn-Banach theorem for real vector spaces
  4042 (in Isabelle/Isar) -- by Gertrud Bauer;
  4043 
  4044 * HOL/BCV: generic model of bytecode verification, i.e. data-flow
  4045 analysis for assembly languages with subtypes;
  4046 
  4047 * HOL/TLA (Lamport's Temporal Logic of Actions): major reorganization
  4048 -- avoids syntactic ambiguities and treats state, transition, and
  4049 temporal levels more uniformly; introduces INCOMPATIBILITIES due to
  4050 changed syntax and (many) tactics;
  4051 
  4052 * HOL/inductive: Now also handles more general introduction rules such
  4053   as "ALL y. (y, x) : r --> y : acc r ==> x : acc r"; monotonicity
  4054   theorems are now maintained within the theory (maintained via the
  4055   "mono" attribute);
  4056 
  4057 * HOL/datatype: Now also handles arbitrarily branching datatypes
  4058   (using function types) such as
  4059 
  4060   datatype 'a tree = Atom 'a | Branch "nat => 'a tree"
  4061 
  4062 * HOL/record: record_simproc (part of the default simpset) takes care
  4063 of selectors applied to updated records; record_split_tac is no longer
  4064 part of the default claset; update_defs may now be removed from the
  4065 simpset in many cases; COMPATIBILITY: old behavior achieved by
  4066 
  4067   claset_ref () := claset() addSWrapper record_split_wrapper;
  4068   Delsimprocs [record_simproc]
  4069 
  4070 * HOL/typedef: fixed type inference for representing set; type
  4071 arguments now have to occur explicitly on the rhs as type constraints;
  4072 
  4073 * HOL/recdef (TFL): 'congs' syntax now expects comma separated list of theorem
  4074 names rather than an ML expression;
  4075 
  4076 * HOL/defer_recdef (TFL): like recdef but the well-founded relation can be
  4077 supplied later.  Program schemes can be defined, such as
  4078     "While B C s = (if B s then While B C (C s) else s)"
  4079 where the well-founded relation can be chosen after B and C have been given.
  4080 
  4081 * HOL/List: the constructors of type list are now Nil and Cons;
  4082 INCOMPATIBILITY: while [] and infix # syntax is still there, of
  4083 course, ML tools referring to List.list.op # etc. have to be adapted;
  4084 
  4085 * HOL_quantifiers flag superseded by "HOL" print mode, which is
  4086 disabled by default; run isabelle with option -m HOL to get back to
  4087 the original Gordon/HOL-style output;
  4088 
  4089 * HOL/Ord.thy: new bounded quantifier syntax (input only): ALL x<y. P,
  4090 ALL x<=y. P, EX x<y. P, EX x<=y. P;
  4091 
  4092 * HOL basic syntax simplified (more orthogonal): all variants of
  4093 All/Ex now support plain / symbolic / HOL notation; plain syntax for
  4094 Eps operator is provided as well: "SOME x. P[x]";
  4095 
  4096 * HOL/Sum.thy: sum_case has been moved to HOL/Datatype;
  4097 
  4098 * HOL/Univ.thy: infix syntax <*>, <+>, <**>, <+> eliminated and made
  4099 thus available for user theories;
  4100 
  4101 * HOLCF/IOA/Sequents: renamed 'Cons' to 'Consq' to avoid clash with
  4102 HOL/List; hardly an INCOMPATIBILITY since '>>' syntax is used all the
  4103 time;
  4104 
  4105 * HOL: new tactic smp_tac: int -> int -> tactic, which applies spec
  4106 several times and then mp;
  4107 
  4108 
  4109 *** LK ***
  4110 
  4111 * the notation <<...>> is now available as a notation for sequences of
  4112 formulas;
  4113 
  4114 * the simplifier is now installed
  4115 
  4116 * the axiom system has been generalized (thanks to Soren Heilmann)
  4117 
  4118 * the classical reasoner now has a default rule database
  4119 
  4120 
  4121 *** ZF ***
  4122 
  4123 * new primrec section allows primitive recursive functions to be given
  4124 directly (as in HOL) over datatypes and the natural numbers;
  4125 
  4126 * new tactics induct_tac and exhaust_tac for induction (or case
  4127 analysis) over datatypes and the natural numbers;
  4128 
  4129 * the datatype declaration of type T now defines the recursor T_rec;
  4130 
  4131 * simplification automatically does freeness reasoning for datatype
  4132 constructors;
  4133 
  4134 * automatic type-inference, with AddTCs command to insert new
  4135 type-checking rules;
  4136 
  4137 * datatype introduction rules are now added as Safe Introduction rules
  4138 to the claset;
  4139 
  4140 * the syntax "if P then x else y" is now available in addition to
  4141 if(P,x,y);
  4142 
  4143 
  4144 *** Internal programming interfaces ***
  4145 
  4146 * tuned simplifier trace output; new flag debug_simp;
  4147 
  4148 * structures Vartab / Termtab (instances of TableFun) offer efficient
  4149 tables indexed by indexname_ord / term_ord (compatible with aconv);
  4150 
  4151 * AxClass.axclass_tac lost the theory argument;
  4152 
  4153 * tuned current_goals_markers semantics: begin / end goal avoids
  4154 printing empty lines;
  4155 
  4156 * removed prs and prs_fn hook, which was broken because it did not
  4157 include \n in its semantics, forcing writeln to add one
  4158 uncoditionally; replaced prs_fn by writeln_fn; consider std_output:
  4159 string -> unit if you really want to output text without newline;
  4160 
  4161 * Symbol.output subject to print mode; INCOMPATIBILITY: defaults to
  4162 plain output, interface builders may have to enable 'isabelle_font'
  4163 mode to get Isabelle font glyphs as before;
  4164 
  4165 * refined token_translation interface; INCOMPATIBILITY: output length
  4166 now of type real instead of int;
  4167 
  4168 * theory loader actions may be traced via new ThyInfo.add_hook
  4169 interface (see src/Pure/Thy/thy_info.ML); example application: keep
  4170 your own database of information attached to *whole* theories -- as
  4171 opposed to intra-theory data slots offered via TheoryDataFun;
  4172 
  4173 * proper handling of dangling sort hypotheses (at last!);
  4174 Thm.strip_shyps and Drule.strip_shyps_warning take care of removing
  4175 extra sort hypotheses that can be witnessed from the type signature;
  4176 the force_strip_shyps flag is gone, any remaining shyps are simply
  4177 left in the theorem (with a warning issued by strip_shyps_warning);
  4178 
  4179 
  4180 
  4181 New in Isabelle98-1 (October 1998)
  4182 ----------------------------------
  4183 
  4184 *** Overview of INCOMPATIBILITIES (see below for more details) ***
  4185 
  4186 * several changes of automated proof tools;
  4187 
  4188 * HOL: major changes to the inductive and datatype packages, including
  4189 some minor incompatibilities of theory syntax;
  4190 
  4191 * HOL: renamed r^-1 to 'converse' from 'inverse'; 'inj_onto' is now
  4192 called `inj_on';
  4193 
  4194 * HOL: removed duplicate thms in Arith:
  4195   less_imp_add_less  should be replaced by  trans_less_add1
  4196   le_imp_add_le      should be replaced by  trans_le_add1
  4197 
  4198 * HOL: unary minus is now overloaded (new type constraints may be
  4199 required);
  4200 
  4201 * HOL and ZF: unary minus for integers is now #- instead of #~.  In
  4202 ZF, expressions such as n#-1 must be changed to n#- 1, since #-1 is
  4203 now taken as an integer constant.
  4204 
  4205 * Pure: ML function 'theory_of' renamed to 'theory';
  4206 
  4207 
  4208 *** Proof tools ***
  4209 
  4210 * Simplifier:
  4211   1. Asm_full_simp_tac is now more aggressive.
  4212      1. It will sometimes reorient premises if that increases their power to
  4213         simplify.
  4214      2. It does no longer proceed strictly from left to right but may also
  4215         rotate premises to achieve further simplification.
  4216      For compatibility reasons there is now Asm_lr_simp_tac which is like the
  4217      old Asm_full_simp_tac in that it does not rotate premises.
  4218   2. The simplifier now knows a little bit about nat-arithmetic.
  4219 
  4220 * Classical reasoner: wrapper mechanism for the classical reasoner now
  4221 allows for selected deletion of wrappers, by introduction of names for
  4222 wrapper functionals.  This implies that addbefore, addSbefore,
  4223 addaltern, and addSaltern now take a pair (name, tactic) as argument,
  4224 and that adding two tactics with the same name overwrites the first
  4225 one (emitting a warning).
  4226   type wrapper = (int -> tactic) -> (int -> tactic)
  4227   setWrapper, setSWrapper, compWrapper and compSWrapper are replaced by
  4228   addWrapper, addSWrapper: claset * (string * wrapper) -> claset
  4229   delWrapper, delSWrapper: claset *  string            -> claset
  4230   getWrapper is renamed to appWrappers, getSWrapper to appSWrappers;
  4231 
  4232 * Classical reasoner: addbefore/addSbefore now have APPEND/ORELSE
  4233 semantics; addbefore now affects only the unsafe part of step_tac
  4234 etc.; this affects addss/auto_tac/force_tac, so EXISTING PROOFS MAY
  4235 FAIL, but proofs should be fixable easily, e.g. by replacing Auto_tac
  4236 by Force_tac;
  4237 
  4238 * Classical reasoner: setwrapper to setWrapper and compwrapper to
  4239 compWrapper; added safe wrapper (and access functions for it);
  4240 
  4241 * HOL/split_all_tac is now much faster and fails if there is nothing
  4242 to split.  Some EXISTING PROOFS MAY REQUIRE ADAPTION because the order
  4243 and the names of the automatically generated variables have changed.
  4244 split_all_tac has moved within claset() from unsafe wrappers to safe
  4245 wrappers, which means that !!-bound variables are split much more
  4246 aggressively, and safe_tac and clarify_tac now split such variables.
  4247 If this splitting is not appropriate, use delSWrapper "split_all_tac".
  4248 Note: the same holds for record_split_tac, which does the job of
  4249 split_all_tac for record fields.
  4250 
  4251 * HOL/Simplifier: Rewrite rules for case distinctions can now be added
  4252 permanently to the default simpset using Addsplits just like
  4253 Addsimps. They can be removed via Delsplits just like
  4254 Delsimps. Lower-case versions are also available.
  4255 
  4256 * HOL/Simplifier: The rule split_if is now part of the default
  4257 simpset. This means that the simplifier will eliminate all occurrences
  4258 of if-then-else in the conclusion of a goal. To prevent this, you can
  4259 either remove split_if completely from the default simpset by
  4260 `Delsplits [split_if]' or remove it in a specific call of the
  4261 simplifier using `... delsplits [split_if]'.  You can also add/delete
  4262 other case splitting rules to/from the default simpset: every datatype
  4263 generates suitable rules `split_t_case' and `split_t_case_asm' (where
  4264 t is the name of the datatype).
  4265 
  4266 * Classical reasoner / Simplifier combination: new force_tac (and
  4267 derivatives Force_tac, force) combines rewriting and classical
  4268 reasoning (and whatever other tools) similarly to auto_tac, but is
  4269 aimed to solve the given subgoal completely.
  4270 
  4271 
  4272 *** General ***
  4273 
  4274 * new top-level commands `Goal' and `Goalw' that improve upon `goal'
  4275 and `goalw': the theory is no longer needed as an explicit argument -
  4276 the current theory context is used; assumptions are no longer returned
  4277 at the ML-level unless one of them starts with ==> or !!; it is
  4278 recommended to convert to these new commands using isatool fixgoal
  4279 (backup your sources first!);
  4280 
  4281 * new top-level commands 'thm' and 'thms' for retrieving theorems from
  4282 the current theory context, and 'theory' to lookup stored theories;
  4283 
  4284 * new theory section 'locale' for declaring constants, assumptions and
  4285 definitions that have local scope;
  4286 
  4287 * new theory section 'nonterminals' for purely syntactic types;
  4288 
  4289 * new theory section 'setup' for generic ML setup functions
  4290 (e.g. package initialization);
  4291 
  4292 * the distribution now includes Isabelle icons: see
  4293 lib/logo/isabelle-{small,tiny}.xpm;
  4294 
  4295 * isatool install - install binaries with absolute references to
  4296 ISABELLE_HOME/bin;
  4297 
  4298 * isatool logo -- create instances of the Isabelle logo (as EPS);
  4299 
  4300 * print mode 'emacs' reserved for Isamode;
  4301 
  4302 * support multiple print (ast) translations per constant name;
  4303 
  4304 * theorems involving oracles are now printed with a suffixed [!];
  4305 
  4306 
  4307 *** HOL ***
  4308 
  4309 * there is now a tutorial on Isabelle/HOL (do 'isatool doc tutorial');
  4310 
  4311 * HOL/inductive package reorganized and improved: now supports mutual
  4312 definitions such as
  4313 
  4314   inductive EVEN ODD
  4315     intrs
  4316       null "0 : EVEN"
  4317       oddI "n : EVEN ==> Suc n : ODD"
  4318       evenI "n : ODD ==> Suc n : EVEN"
  4319 
  4320 new theorem list "elims" contains an elimination rule for each of the
  4321 recursive sets; inductive definitions now handle disjunctive premises
  4322 correctly (also ZF);
  4323 
  4324 INCOMPATIBILITIES: requires Inductive as an ancestor; component
  4325 "mutual_induct" no longer exists - the induction rule is always
  4326 contained in "induct";
  4327 
  4328 
  4329 * HOL/datatype package re-implemented and greatly improved: now
  4330 supports mutually recursive datatypes such as
  4331 
  4332   datatype
  4333     'a aexp = IF_THEN_ELSE ('a bexp) ('a aexp) ('a aexp)
  4334             | SUM ('a aexp) ('a aexp)
  4335             | DIFF ('a aexp) ('a aexp)
  4336             | NUM 'a
  4337   and
  4338     'a bexp = LESS ('a aexp) ('a aexp)
  4339             | AND ('a bexp) ('a bexp)
  4340             | OR ('a bexp) ('a bexp)
  4341 
  4342 as well as indirectly recursive datatypes such as
  4343 
  4344   datatype
  4345     ('a, 'b) term = Var 'a
  4346                   | App 'b ((('a, 'b) term) list)
  4347 
  4348 The new tactic  mutual_induct_tac [<var_1>, ..., <var_n>] i  performs
  4349 induction on mutually / indirectly recursive datatypes.
  4350 
  4351 Primrec equations are now stored in theory and can be accessed via
  4352 <function_name>.simps.
  4353 
  4354 INCOMPATIBILITIES:
  4355 
  4356   - Theories using datatypes must now have theory Datatype as an
  4357     ancestor.
  4358   - The specific <typename>.induct_tac no longer exists - use the
  4359     generic induct_tac instead.
  4360   - natE has been renamed to nat.exhaust - use exhaust_tac
  4361     instead of res_inst_tac ... natE. Note that the variable
  4362     names in nat.exhaust differ from the names in natE, this
  4363     may cause some "fragile" proofs to fail.
  4364   - The theorems split_<typename>_case and split_<typename>_case_asm
  4365     have been renamed to <typename>.split and <typename>.split_asm.
  4366   - Since default sorts of type variables are now handled correctly,
  4367     some datatype definitions may have to be annotated with explicit
  4368     sort constraints.
  4369   - Primrec definitions no longer require function name and type
  4370     of recursive argument.
  4371 
  4372 Consider using isatool fixdatatype to adapt your theories and proof
  4373 scripts to the new package (backup your sources first!).
  4374 
  4375 
  4376 * HOL/record package: considerably improved implementation; now
  4377 includes concrete syntax for record types, terms, updates; theorems
  4378 for surjective pairing and splitting !!-bound record variables; proof
  4379 support is as follows:
  4380 
  4381   1) standard conversions (selectors or updates applied to record
  4382 constructor terms) are part of the standard simpset;
  4383 
  4384   2) inject equations of the form ((x, y) = (x', y')) == x=x' & y=y' are
  4385 made part of standard simpset and claset via addIffs;
  4386 
  4387   3) a tactic for record field splitting (record_split_tac) is part of
  4388 the standard claset (addSWrapper);
  4389 
  4390 To get a better idea about these rules you may retrieve them via
  4391 something like 'thms "foo.simps"' or 'thms "foo.iffs"', where "foo" is
  4392 the name of your record type.
  4393 
  4394 The split tactic 3) conceptually simplifies by the following rule:
  4395 
  4396   "(!!x. PROP ?P x) == (!!a b. PROP ?P (a, b))"
  4397 
  4398 Thus any record variable that is bound by meta-all will automatically
  4399 blow up into some record constructor term, consequently the
  4400 simplifications of 1), 2) apply.  Thus force_tac, auto_tac etc. shall
  4401 solve record problems automatically.
  4402 
  4403 
  4404 * reorganized the main HOL image: HOL/Integ and String loaded by
  4405 default; theory Main includes everything;
  4406 
  4407 * automatic simplification of integer sums and comparisons, using cancellation;
  4408 
  4409 * added option_map_eq_Some and not_Some_eq to the default simpset and claset;
  4410 
  4411 * added disj_not1 = "(~P | Q) = (P --> Q)" to the default simpset;
  4412 
  4413 * many new identities for unions, intersections, set difference, etc.;
  4414 
  4415 * expand_if, expand_split, expand_sum_case and expand_nat_case are now
  4416 called split_if, split_split, split_sum_case and split_nat_case (to go
  4417 with add/delsplits);
  4418 
  4419 * HOL/Prod introduces simplification procedure unit_eq_proc rewriting
  4420 (?x::unit) = (); this is made part of the default simpset, which COULD
  4421 MAKE EXISTING PROOFS FAIL under rare circumstances (consider
  4422 'Delsimprocs [unit_eq_proc];' as last resort); also note that
  4423 unit_abs_eta_conv is added in order to counter the effect of
  4424 unit_eq_proc on (%u::unit. f u), replacing it by f rather than by
  4425 %u.f();
  4426 
  4427 * HOL/Fun INCOMPATIBILITY: `inj_onto' is now called `inj_on' (which
  4428 makes more sense);
  4429 
  4430 * HOL/Set INCOMPATIBILITY: rule `equals0D' is now a well-formed destruct rule;
  4431   It and 'sym RS equals0D' are now in the default  claset, giving automatic
  4432   disjointness reasoning but breaking a few old proofs.
  4433 
  4434 * HOL/Relation INCOMPATIBILITY: renamed the relational operator r^-1
  4435 to 'converse' from 'inverse' (for compatibility with ZF and some
  4436 literature);
  4437 
  4438 * HOL/recdef can now declare non-recursive functions, with {} supplied as
  4439 the well-founded relation;
  4440 
  4441 * HOL/Set INCOMPATIBILITY: the complement of set A is now written -A instead of
  4442     Compl A.  The "Compl" syntax remains available as input syntax for this
  4443     release ONLY.
  4444 
  4445 * HOL/Update: new theory of function updates:
  4446     f(a:=b) == %x. if x=a then b else f x
  4447 may also be iterated as in f(a:=b,c:=d,...);
  4448 
  4449 * HOL/Vimage: new theory for inverse image of a function, syntax f-``B;
  4450 
  4451 * HOL/List:
  4452   - new function list_update written xs[i:=v] that updates the i-th
  4453     list position. May also be iterated as in xs[i:=a,j:=b,...].
  4454   - new function `upt' written [i..j(] which generates the list
  4455     [i,i+1,...,j-1], i.e. the upper bound is excluded. To include the upper
  4456     bound write [i..j], which is a shorthand for [i..j+1(].
  4457   - new lexicographic orderings and corresponding wellfoundedness theorems.
  4458 
  4459 * HOL/Arith:
  4460   - removed 'pred' (predecessor) function;
  4461   - generalized some theorems about n-1;
  4462   - many new laws about "div" and "mod";
  4463   - new laws about greatest common divisors (see theory ex/Primes);
  4464 
  4465 * HOL/Relation: renamed the relational operator r^-1 "converse"
  4466 instead of "inverse";
  4467 
  4468 * HOL/Induct/Multiset: a theory of multisets, including the wellfoundedness
  4469   of the multiset ordering;
  4470 
  4471 * directory HOL/Real: a construction of the reals using Dedekind cuts
  4472   (not included by default);
  4473 
  4474 * directory HOL/UNITY: Chandy and Misra's UNITY formalism;
  4475 
  4476 * directory HOL/Hoare: a new version of Hoare logic which permits many-sorted
  4477   programs, i.e. different program variables may have different types.
  4478 
  4479 * calling (stac rew i) now fails if "rew" has no effect on the goal
  4480   [previously, this check worked only if the rewrite rule was unconditional]
  4481   Now rew can involve either definitions or equalities (either == or =).
  4482 
  4483 
  4484 *** ZF ***
  4485 
  4486 * theory Main includes everything; INCOMPATIBILITY: theory ZF.thy contains
  4487   only the theorems proved on ZF.ML;
  4488 
  4489 * ZF INCOMPATIBILITY: rule `equals0D' is now a well-formed destruct rule;
  4490   It and 'sym RS equals0D' are now in the default  claset, giving automatic
  4491   disjointness reasoning but breaking a few old proofs.
  4492 
  4493 * ZF/Update: new theory of function updates
  4494     with default rewrite rule  f(x:=y) ` z = if(z=x, y, f`z)
  4495   may also be iterated as in f(a:=b,c:=d,...);
  4496 
  4497 * in  let x=t in u(x), neither t nor u(x) has to be an FOL term.
  4498 
  4499 * calling (stac rew i) now fails if "rew" has no effect on the goal
  4500   [previously, this check worked only if the rewrite rule was unconditional]
  4501   Now rew can involve either definitions or equalities (either == or =).
  4502 
  4503 * case_tac provided for compatibility with HOL
  4504     (like the old excluded_middle_tac, but with subgoals swapped)
  4505 
  4506 
  4507 *** Internal programming interfaces ***
  4508 
  4509 * Pure: several new basic modules made available for general use, see
  4510 also src/Pure/README;
  4511 
  4512 * improved the theory data mechanism to support encapsulation (data
  4513 kind name replaced by private Object.kind, acting as authorization
  4514 key); new type-safe user interface via functor TheoryDataFun; generic
  4515 print_data function becomes basically useless;
  4516 
  4517 * removed global_names compatibility flag -- all theory declarations
  4518 are qualified by default;
  4519 
  4520 * module Pure/Syntax now offers quote / antiquote translation
  4521 functions (useful for Hoare logic etc. with implicit dependencies);
  4522 see HOL/ex/Antiquote for an example use;
  4523 
  4524 * Simplifier now offers conversions (asm_)(full_)rewrite: simpset ->
  4525 cterm -> thm;
  4526 
  4527 * new tactical CHANGED_GOAL for checking that a tactic modifies a
  4528 subgoal;
  4529 
  4530 * Display.print_goals function moved to Locale.print_goals;
  4531 
  4532 * standard print function for goals supports current_goals_markers
  4533 variable for marking begin of proof, end of proof, start of goal; the
  4534 default is ("", "", ""); setting current_goals_markers := ("<proof>",
  4535 "</proof>", "<goal>") causes SGML like tagged proof state printing,
  4536 for example;
  4537 
  4538 
  4539 
  4540 New in Isabelle98 (January 1998)
  4541 --------------------------------
  4542 
  4543 *** Overview of INCOMPATIBILITIES (see below for more details) ***
  4544 
  4545 * changed lexical syntax of terms / types: dots made part of long
  4546 identifiers, e.g. "%x.x" no longer possible, should be "%x. x";
  4547 
  4548 * simpset (and claset) reference variable replaced by functions
  4549 simpset / simpset_ref;
  4550 
  4551 * no longer supports theory aliases (via merge) and non-trivial
  4552 implicit merge of thms' signatures;
  4553 
  4554 * most internal names of constants changed due to qualified names;
  4555 
  4556 * changed Pure/Sequence interface (see Pure/seq.ML);
  4557 
  4558 
  4559 *** General Changes ***
  4560 
  4561 * hierachically structured name spaces (for consts, types, axms, thms
  4562 etc.); new lexical class 'longid' (e.g. Foo.bar.x) may render much of
  4563 old input syntactically incorrect (e.g. "%x.x"); COMPATIBILITY:
  4564 isatool fixdots ensures space after dots (e.g. "%x. x"); set
  4565 long_names for fully qualified output names; NOTE: ML programs
  4566 (special tactics, packages etc.) referring to internal names may have
  4567 to be adapted to cope with fully qualified names; in case of severe
  4568 backward campatibility problems try setting 'global_names' at compile
  4569 time to have enrything declared within a flat name space; one may also
  4570 fine tune name declarations in theories via the 'global' and 'local'
  4571 section;
  4572 
  4573 * reimplemented the implicit simpset and claset using the new anytype
  4574 data filed in signatures; references simpset:simpset ref etc. are
  4575 replaced by functions simpset:unit->simpset and
  4576 simpset_ref:unit->simpset ref; COMPATIBILITY: use isatool fixclasimp
  4577 to patch your ML files accordingly;
  4578 
  4579 * HTML output now includes theory graph data for display with Java
  4580 applet or isatool browser; data generated automatically via isatool
  4581 usedir (see -i option, ISABELLE_USEDIR_OPTIONS);
  4582 
  4583 * defs may now be conditional; improved rewrite_goals_tac to handle
  4584 conditional equations;
  4585 
  4586 * defs now admits additional type arguments, using TYPE('a) syntax;
  4587 
  4588 * theory aliases via merge (e.g. M=A+B+C) no longer supported, always
  4589 creates a new theory node; implicit merge of thms' signatures is
  4590 restricted to 'trivial' ones; COMPATIBILITY: one may have to use
  4591 transfer:theory->thm->thm in (rare) cases;
  4592 
  4593 * improved handling of draft signatures / theories; draft thms (and
  4594 ctyps, cterms) are automatically promoted to real ones;
  4595 
  4596 * slightly changed interfaces for oracles: admit many per theory, named
  4597 (e.g. oracle foo = mlfun), additional name argument for invoke_oracle;
  4598 
  4599 * print_goals: optional output of const types (set show_consts and
  4600 show_types);
  4601 
  4602 * improved output of warnings (###) and errors (***);
  4603 
  4604 * subgoal_tac displays a warning if the new subgoal has type variables;
  4605 
  4606 * removed old README and Makefiles;
  4607 
  4608 * replaced print_goals_ref hook by print_current_goals_fn and result_error_fn;
  4609 
  4610 * removed obsolete init_pps and init_database;
  4611 
  4612 * deleted the obsolete tactical STATE, which was declared by
  4613     fun STATE tacfun st = tacfun st st;
  4614 
  4615 * cd and use now support path variables, e.g. $ISABELLE_HOME, or ~
  4616 (which abbreviates $HOME);
  4617 
  4618 * changed Pure/Sequence interface (see Pure/seq.ML); COMPATIBILITY:
  4619 use isatool fixseq to adapt your ML programs (this works for fully
  4620 qualified references to the Sequence structure only!);
  4621 
  4622 * use_thy no longer requires writable current directory; it always
  4623 reloads .ML *and* .thy file, if either one is out of date;
  4624 
  4625 
  4626 *** Classical Reasoner ***
  4627 
  4628 * Clarify_tac, clarify_tac, clarify_step_tac, Clarify_step_tac: new
  4629 tactics that use classical reasoning to simplify a subgoal without
  4630 splitting it into several subgoals;
  4631 
  4632 * Safe_tac: like safe_tac but uses the default claset;
  4633 
  4634 
  4635 *** Simplifier ***
  4636 
  4637 * added simplification meta rules:
  4638     (asm_)(full_)simplify: simpset -> thm -> thm;
  4639 
  4640 * simplifier.ML no longer part of Pure -- has to be loaded by object
  4641 logics (again);
  4642 
  4643 * added prems argument to simplification procedures;
  4644 
  4645 * HOL, FOL, ZF: added infix function `addsplits':
  4646   instead of `<simpset> setloop (split_tac <thms>)'
  4647   you can simply write `<simpset> addsplits <thms>'
  4648 
  4649 
  4650 *** Syntax ***
  4651 
  4652 * TYPE('a) syntax for type reflection terms;
  4653 
  4654 * no longer handles consts with name "" -- declare as 'syntax' instead;
  4655 
  4656 * pretty printer: changed order of mixfix annotation preference (again!);
  4657 
  4658 * Pure: fixed idt/idts vs. pttrn/pttrns syntactic categories;
  4659 
  4660 
  4661 *** HOL ***
  4662 
  4663 * HOL: there is a new splitter `split_asm_tac' that can be used e.g.
  4664   with `addloop' of the simplifier to faciliate case splitting in premises.
  4665 
  4666 * HOL/TLA: Stephan Merz's formalization of Lamport's Temporal Logic of Actions;
  4667 
  4668 * HOL/Auth: new protocol proofs including some for the Internet
  4669   protocol TLS;
  4670 
  4671 * HOL/Map: new theory of `maps' a la VDM;
  4672 
  4673 * HOL/simplifier: simplification procedures nat_cancel_sums for
  4674 cancelling out common nat summands from =, <, <= (in)equalities, or
  4675 differences; simplification procedures nat_cancel_factor for
  4676 cancelling common factor from =, <, <= (in)equalities over natural
  4677 sums; nat_cancel contains both kinds of procedures, it is installed by
  4678 default in Arith.thy -- this COULD MAKE EXISTING PROOFS FAIL;
  4679 
  4680 * HOL/simplifier: terms of the form
  4681   `? x. P1(x) & ... & Pn(x) & x=t & Q1(x) & ... Qn(x)'  (or t=x)
  4682   are rewritten to
  4683   `P1(t) & ... & Pn(t) & Q1(t) & ... Qn(t)',
  4684   and those of the form
  4685   `! x. P1(x) & ... & Pn(x) & x=t & Q1(x) & ... Qn(x) --> R(x)'  (or t=x)
  4686   are rewritten to
  4687   `P1(t) & ... & Pn(t) & Q1(t) & ... Qn(t) --> R(t)',
  4688 
  4689 * HOL/datatype
  4690   Each datatype `t' now comes with a theorem `split_t_case' of the form
  4691 
  4692   P(t_case f1 ... fn x) =
  4693      ( (!y1 ... ym1. x = C1 y1 ... ym1 --> P(f1 y1 ... ym1)) &
  4694         ...
  4695        (!y1 ... ymn. x = Cn y1 ... ymn --> P(f1 y1 ... ymn))
  4696      )
  4697 
  4698   and a theorem `split_t_case_asm' of the form
  4699 
  4700   P(t_case f1 ... fn x) =
  4701     ~( (? y1 ... ym1. x = C1 y1 ... ym1 & ~P(f1 y1 ... ym1)) |
  4702         ...
  4703        (? y1 ... ymn. x = Cn y1 ... ymn & ~P(f1 y1 ... ymn))
  4704      )
  4705   which can be added to a simpset via `addsplits'. The existing theorems
  4706   expand_list_case and expand_option_case have been renamed to
  4707   split_list_case and split_option_case.
  4708 
  4709 * HOL/Arithmetic:
  4710   - `pred n' is automatically converted to `n-1'.
  4711     Users are strongly encouraged not to use `pred' any longer,
  4712     because it will disappear altogether at some point.
  4713   - Users are strongly encouraged to write "0 < n" rather than
  4714     "n ~= 0". Theorems and proof tools have been modified towards this
  4715     `standard'.
  4716 
  4717 * HOL/Lists:
  4718   the function "set_of_list" has been renamed "set" (and its theorems too);
  4719   the function "nth" now takes its arguments in the reverse order and
  4720   has acquired the infix notation "!" as in "xs!n".
  4721 
  4722 * HOL/Set: UNIV is now a constant and is no longer translated to Compl{};
  4723 
  4724 * HOL/Set: The operator (UN x.B x) now abbreviates (UN x:UNIV. B x) and its
  4725   specialist theorems (like UN1_I) are gone.  Similarly for (INT x.B x);
  4726 
  4727 * HOL/record: extensible records with schematic structural subtyping
  4728 (single inheritance); EXPERIMENTAL version demonstrating the encoding,
  4729 still lacks various theorems and concrete record syntax;
  4730 
  4731 
  4732 *** HOLCF ***
  4733 
  4734 * removed "axioms" and "generated by" sections;
  4735 
  4736 * replaced "ops" section by extended "consts" section, which is capable of
  4737   handling the continuous function space "->" directly;
  4738 
  4739 * domain package:
  4740   . proves theorems immediately and stores them in the theory,
  4741   . creates hierachical name space,
  4742   . now uses normal mixfix annotations (instead of cinfix...),
  4743   . minor changes to some names and values (for consistency),
  4744   . e.g. cases -> casedist, dists_eq -> dist_eqs, [take_lemma] -> take_lemmas,
  4745   . separator between mutual domain defs: changed "," to "and",
  4746   . improved handling of sort constraints;  now they have to
  4747     appear on the left-hand side of the equations only;
  4748 
  4749 * fixed LAM <x,y,zs>.b syntax;
  4750 
  4751 * added extended adm_tac to simplifier in HOLCF -- can now discharge
  4752 adm (%x. P (t x)), where P is chainfinite and t continuous;
  4753 
  4754 
  4755 *** FOL and ZF ***
  4756 
  4757 * FOL: there is a new splitter `split_asm_tac' that can be used e.g.
  4758   with `addloop' of the simplifier to faciliate case splitting in premises.
  4759 
  4760 * qed_spec_mp, qed_goal_spec_mp, qed_goalw_spec_mp are available, as
  4761 in HOL, they strip ALL and --> from proved theorems;
  4762 
  4763 
  4764 
  4765 New in Isabelle94-8 (May 1997)
  4766 ------------------------------
  4767 
  4768 *** General Changes ***
  4769 
  4770 * new utilities to build / run / maintain Isabelle etc. (in parts
  4771 still somewhat experimental); old Makefiles etc. still functional;
  4772 
  4773 * new 'Isabelle System Manual';
  4774 
  4775 * INSTALL text, together with ./configure and ./build scripts;
  4776 
  4777 * reimplemented type inference for greater efficiency, better error
  4778 messages and clean internal interface;
  4779 
  4780 * prlim command for dealing with lots of subgoals (an easier way of
  4781 setting goals_limit);
  4782 
  4783 
  4784 *** Syntax ***
  4785 
  4786 * supports alternative (named) syntax tables (parser and pretty
  4787 printer); internal interface is provided by add_modesyntax(_i);
  4788 
  4789 * Pure, FOL, ZF, HOL, HOLCF now support symbolic input and output; to
  4790 be used in conjunction with the Isabelle symbol font; uses the
  4791 "symbols" syntax table;
  4792 
  4793 * added token_translation interface (may translate name tokens in
  4794 arbitrary ways, dependent on their type (free, bound, tfree, ...) and
  4795 the current print_mode); IMPORTANT: user print translation functions
  4796 are responsible for marking newly introduced bounds
  4797 (Syntax.mark_boundT);
  4798 
  4799 * token translations for modes "xterm" and "xterm_color" that display
  4800 names in bold, underline etc. or colors (which requires a color
  4801 version of xterm);
  4802 
  4803 * infixes may now be declared with names independent of their syntax;
  4804 
  4805 * added typed_print_translation (like print_translation, but may
  4806 access type of constant);
  4807 
  4808 
  4809 *** Classical Reasoner ***
  4810 
  4811 Blast_tac: a new tactic!  It is often more powerful than fast_tac, but has
  4812 some limitations.  Blast_tac...
  4813   + ignores addss, addbefore, addafter; this restriction is intrinsic
  4814   + ignores elimination rules that don't have the correct format
  4815         (the conclusion MUST be a formula variable)
  4816   + ignores types, which can make HOL proofs fail
  4817   + rules must not require higher-order unification, e.g. apply_type in ZF
  4818     [message "Function Var's argument not a bound variable" relates to this]
  4819   + its proof strategy is more general but can actually be slower
  4820 
  4821 * substitution with equality assumptions no longer permutes other
  4822 assumptions;
  4823 
  4824 * minor changes in semantics of addafter (now called addaltern); renamed
  4825 setwrapper to setWrapper and compwrapper to compWrapper; added safe wrapper
  4826 (and access functions for it);
  4827 
  4828 * improved combination of classical reasoner and simplifier:
  4829   + functions for handling clasimpsets
  4830   + improvement of addss: now the simplifier is called _after_ the
  4831     safe steps.
  4832   + safe variant of addss called addSss: uses safe simplifications
  4833     _during_ the safe steps. It is more complete as it allows multiple
  4834     instantiations of unknowns (e.g. with slow_tac).
  4835 
  4836 *** Simplifier ***
  4837 
  4838 * added interface for simplification procedures (functions that
  4839 produce *proven* rewrite rules on the fly, depending on current
  4840 redex);
  4841 
  4842 * ordering on terms as parameter (used for ordered rewriting);
  4843 
  4844 * new functions delcongs, deleqcongs, and Delcongs. richer rep_ss;
  4845 
  4846 * the solver is now split into a safe and an unsafe part.
  4847 This should be invisible for the normal user, except that the
  4848 functions setsolver and addsolver have been renamed to setSolver and
  4849 addSolver; added safe_asm_full_simp_tac;
  4850 
  4851 
  4852 *** HOL ***
  4853 
  4854 * a generic induction tactic `induct_tac' which works for all datatypes and
  4855 also for type `nat';
  4856 
  4857 * a generic case distinction tactic `exhaust_tac' which works for all
  4858 datatypes and also for type `nat';
  4859 
  4860 * each datatype comes with a function `size';
  4861 
  4862 * patterns in case expressions allow tuple patterns as arguments to
  4863 constructors, for example `case x of [] => ... | (x,y,z)#ps => ...';
  4864 
  4865 * primrec now also works with type nat;
  4866 
  4867 * recdef: a new declaration form, allows general recursive functions to be
  4868 defined in theory files.  See HOL/ex/Fib, HOL/ex/Primes, HOL/Subst/Unify.
  4869 
  4870 * the constant for negation has been renamed from "not" to "Not" to
  4871 harmonize with FOL, ZF, LK, etc.;
  4872 
  4873 * HOL/ex/LFilter theory of a corecursive "filter" functional for
  4874 infinite lists;
  4875 
  4876 * HOL/Modelcheck demonstrates invocation of model checker oracle;
  4877 
  4878 * HOL/ex/Ring.thy declares cring_simp, which solves equational
  4879 problems in commutative rings, using axiomatic type classes for + and *;
  4880 
  4881 * more examples in HOL/MiniML and HOL/Auth;
  4882 
  4883 * more default rewrite rules for quantifiers, union/intersection;
  4884 
  4885 * a new constant `arbitrary == @x.False';
  4886 
  4887 * HOLCF/IOA replaces old HOL/IOA;
  4888 
  4889 * HOLCF changes: derived all rules and arities
  4890   + axiomatic type classes instead of classes
  4891   + typedef instead of faking type definitions
  4892   + eliminated the internal constants less_fun, less_cfun, UU_fun, UU_cfun etc.
  4893   + new axclasses cpo, chfin, flat with flat < chfin < pcpo < cpo < po
  4894   + eliminated the types void, one, tr
  4895   + use unit lift and bool lift (with translations) instead of one and tr
  4896   + eliminated blift from Lift3.thy (use Def instead of blift)
  4897   all eliminated rules are derived as theorems --> no visible changes ;
  4898 
  4899 
  4900 *** ZF ***
  4901 
  4902 * ZF now has Fast_tac, Simp_tac and Auto_tac.  Union_iff is a now a default
  4903 rewrite rule; this may affect some proofs.  eq_cs is gone but can be put back
  4904 as ZF_cs addSIs [equalityI];
  4905 
  4906 
  4907 
  4908 New in Isabelle94-7 (November 96)
  4909 ---------------------------------
  4910 
  4911 * allowing negative levels (as offsets) in prlev and choplev;
  4912 
  4913 * super-linear speedup for large simplifications;
  4914 
  4915 * FOL, ZF and HOL now use miniscoping: rewriting pushes
  4916 quantifications in as far as possible (COULD MAKE EXISTING PROOFS
  4917 FAIL); can suppress it using the command Delsimps (ex_simps @
  4918 all_simps); De Morgan laws are also now included, by default;
  4919 
  4920 * improved printing of ==>  :  ~:
  4921 
  4922 * new object-logic "Sequents" adds linear logic, while replacing LK
  4923 and Modal (thanks to Sara Kalvala);
  4924 
  4925 * HOL/Auth: correctness proofs for authentication protocols;
  4926 
  4927 * HOL: new auto_tac combines rewriting and classical reasoning (many
  4928 examples on HOL/Auth);
  4929 
  4930 * HOL: new command AddIffs for declaring theorems of the form P=Q to
  4931 the rewriter and classical reasoner simultaneously;
  4932 
  4933 * function uresult no longer returns theorems in "standard" format;
  4934 regain previous version by: val uresult = standard o uresult;
  4935 
  4936 
  4937 
  4938 New in Isabelle94-6
  4939 -------------------
  4940 
  4941 * oracles -- these establish an interface between Isabelle and trusted
  4942 external reasoners, which may deliver results as theorems;
  4943 
  4944 * proof objects (in particular record all uses of oracles);
  4945 
  4946 * Simp_tac, Fast_tac, etc. that refer to implicit simpset / claset;
  4947 
  4948 * "constdefs" section in theory files;
  4949 
  4950 * "primrec" section (HOL) no longer requires names;
  4951 
  4952 * internal type "tactic" now simply "thm -> thm Sequence.seq";
  4953 
  4954 
  4955 
  4956 New in Isabelle94-5
  4957 -------------------
  4958 
  4959 * reduced space requirements;
  4960 
  4961 * automatic HTML generation from theories;
  4962 
  4963 * theory files no longer require "..." (quotes) around most types;
  4964 
  4965 * new examples, including two proofs of the Church-Rosser theorem;
  4966 
  4967 * non-curried (1994) version of HOL is no longer distributed;
  4968 
  4969 
  4970 
  4971 New in Isabelle94-4
  4972 -------------------
  4973 
  4974 * greatly reduced space requirements;
  4975 
  4976 * theory files (.thy) no longer require \...\ escapes at line breaks;
  4977 
  4978 * searchable theorem database (see the section "Retrieving theorems" on
  4979 page 8 of the Reference Manual);
  4980 
  4981 * new examples, including Grabczewski's monumental case study of the
  4982 Axiom of Choice;
  4983 
  4984 * The previous version of HOL renamed to Old_HOL;
  4985 
  4986 * The new version of HOL (previously called CHOL) uses a curried syntax
  4987 for functions.  Application looks like f a b instead of f(a,b);
  4988 
  4989 * Mutually recursive inductive definitions finally work in HOL;
  4990 
  4991 * In ZF, pattern-matching on tuples is now available in all abstractions and
  4992 translates to the operator "split";
  4993 
  4994 
  4995 
  4996 New in Isabelle94-3
  4997 -------------------
  4998 
  4999 * new infix operator, addss, allowing the classical reasoner to
  5000 perform simplification at each step of its search.  Example:
  5001         fast_tac (cs addss ss)
  5002 
  5003 * a new logic, CHOL, the same as HOL, but with a curried syntax
  5004 for functions.  Application looks like f a b instead of f(a,b).  Also pairs
  5005 look like (a,b) instead of <a,b>;
  5006 
  5007 * PLEASE NOTE: CHOL will eventually replace HOL!
  5008 
  5009 * In CHOL, pattern-matching on tuples is now available in all abstractions.
  5010 It translates to the operator "split".  A new theory of integers is available;
  5011 
  5012 * In ZF, integer numerals now denote two's-complement binary integers.
  5013 Arithmetic operations can be performed by rewriting.  See ZF/ex/Bin.ML;
  5014 
  5015 * Many new examples: I/O automata, Church-Rosser theorem, equivalents
  5016 of the Axiom of Choice;
  5017 
  5018 
  5019 
  5020 New in Isabelle94-2
  5021 -------------------
  5022 
  5023 * Significantly faster resolution;
  5024 
  5025 * the different sections in a .thy file can now be mixed and repeated
  5026 freely;
  5027 
  5028 * Database of theorems for FOL, HOL and ZF.  New
  5029 commands including qed, qed_goal and bind_thm store theorems in the database.
  5030 
  5031 * Simple database queries: return a named theorem (get_thm) or all theorems of
  5032 a given theory (thms_of), or find out what theory a theorem was proved in
  5033 (theory_of_thm);
  5034 
  5035 * Bugs fixed in the inductive definition and datatype packages;
  5036 
  5037 * The classical reasoner provides deepen_tac and depth_tac, making FOL_dup_cs
  5038 and HOL_dup_cs obsolete;
  5039 
  5040 * Syntactic ambiguities caused by the new treatment of syntax in Isabelle94-1
  5041 have been removed;
  5042 
  5043 * Simpler definition of function space in ZF;
  5044 
  5045 * new results about cardinal and ordinal arithmetic in ZF;
  5046 
  5047 * 'subtype' facility in HOL for introducing new types as subsets of existing
  5048 types;
  5049 
  5050 
  5051 $Id$