NEWS
author chaieb
Fri Jun 08 18:09:37 2007 +0200 (2007-06-08)
changeset 23295 86e225406859
parent 23251 471b576aad25
child 23300 b785068bd5dc
permissions -rw-r--r--
Method "algebra" solves polynomial equations over (semi)rings
     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).
    12 
    13 * Theory syntax: the header format ``theory A = B + C:'' has been
    14 discontinued in favour of ``theory A imports B C begin''.  Use isatool
    15 fixheaders to convert existing theory files.  INCOMPATIBILITY.
    16 
    17 * Theory syntax: the old non-Isar theory file format has been
    18 discontinued altogether.  Note that ML proof scripts may still be used
    19 with Isar theories; migration is usually quite simple with the ML
    20 function use_legacy_bindings.  INCOMPATIBILITY.
    21 
    22 * Theory syntax: some popular names (e.g. 'class', 'declaration',
    23 'fun', 'help', 'if') are now keywords.  INCOMPATIBILITY, use double
    24 quotes.
    25 
    26 * Legacy goal package: reduced interface to the bare minimum required
    27 to keep existing proof scripts running.  Most other user-level
    28 functions are now part of the OldGoals structure, which is *not* open
    29 by default (consider isatool expandshort before open OldGoals).
    30 Removed top_sg, prin, printyp, pprint_term/typ altogether, because
    31 these tend to cause confusion about the actual goal (!) context being
    32 used here, which is not necessarily the same as the_context().
    33 
    34 * Command 'find_theorems': support "*" wildcard in "name:" criterion.
    35 
    36 * Proof General interface: A search form for the "Find Theorems" command is
    37 now available via C-c C-a C-f.  The old minibuffer interface is available
    38 via C-c C-a C-m.  Variable proof-find-theorems-command (customizable via
    39 'Proof-General -> Advanced -> Internals -> Prover Config') controls the
    40 default behavior of 'ProofGeneral -> Find Theorems' (C-c C-f): set to
    41 isar-find-theorems-form or isar-find-theorems-minibuffer.
    42 
    43 * The ``prems limit'' option (cf. ProofContext.prems_limit) is now -1
    44 by default, which means that "prems" (and also "fixed variables") are
    45 suppressed from proof state output.  Note that the ProofGeneral
    46 settings mechanism allows to change and save options persistently, but
    47 older versions of Isabelle will fail to start up if a negative prems
    48 limit is imposed.
    49 
    50 * Local theory targets may be specified by non-nested blocks of
    51 ``context/locale/class ... begin'' followed by ``end''.  The body may
    52 contain definitions, theorems etc., including any derived mechanism
    53 that has been implemented on top of these primitives.  This concept
    54 generalizes the existing ``theorem (in ...)'' towards more versatility
    55 and scalability.
    56 
    57 * Proof General interface: proper undo of final 'end' command;
    58 discontinued Isabelle/classic mode (ML proof scripts).
    59 
    60 
    61 *** Document preparation ***
    62 
    63 * Added antiquotation @{theory name} which prints the given name,
    64 after checking that it refers to a valid ancestor theory in the
    65 current context.
    66 
    67 * Added antiquotations @{ML_type text} and @{ML_struct text} which
    68 check the given source text as ML type/structure, printing verbatim.
    69 
    70 * Added antiquotation @{abbrev "c args"} which prints the abbreviation
    71 "c args == rhs" given in the current context.  (Any number of
    72 arguments may be given on the LHS.)
    73 
    74 
    75 
    76 *** Pure ***
    77 
    78 * code generator: consts in 'consts_code' Isar commands are now referred
    79   to by usual term syntax (including optional type annotations).
    80 
    81 * code generator: 
    82   - Isar 'definition's, 'constdef's and primitive instance definitions are added
    83     explicitly to the table of defining equations
    84   - primitive definitions are not used as defining equations by default any longer
    85   - defining equations are now definitly restricted to meta "==" and object
    86         equality "="
    87   - HOL theories have been adopted accordingly
    88 
    89 * class_package.ML offers a combination of axclasses and locales to
    90 achieve Haskell-like type classes in Isabelle.  See
    91 HOL/ex/Classpackage.thy for examples.
    92 
    93 * Yet another code generator framework allows to generate executable
    94 code for ML and Haskell (including "class"es).  A short usage sketch:
    95 
    96     internal compilation:
    97         code_gen <list of constants (term syntax)> (SML #)
    98     writing SML code to a file:
    99         code_gen <list of constants (term syntax)> (SML <filename>)
   100     writing OCaml code to a file:
   101         code_gen <list of constants (term syntax)> (OCaml <filename>)
   102     writing Haskell code to a bunch of files:
   103         code_gen <list of constants (term syntax)> (Haskell <filename>)
   104 
   105 Reasonable default setup of framework in HOL/Main.
   106 
   107 Theorem attributs for selecting and transforming function equations theorems:
   108 
   109     [code fun]:        select a theorem as function equation for a specific constant
   110     [code fun del]:    deselect a theorem as function equation for a specific constant
   111     [code inline]:     select an equation theorem for unfolding (inlining) in place
   112     [code inline del]: deselect an equation theorem for unfolding (inlining) in place
   113 
   114 User-defined serializations (target in {SML, OCaml, Haskell}):
   115 
   116     code_const <and-list of constants (term syntax)>
   117       {(target) <and-list of const target syntax>}+
   118 
   119     code_type <and-list of type constructors>
   120       {(target) <and-list of type target syntax>}+
   121 
   122     code_instance <and-list of instances>
   123       {(target)}+
   124         where instance ::= <type constructor> :: <class>
   125 
   126     code_class <and_list of classes>
   127       {(target) <and-list of class target syntax>}+
   128         where class target syntax ::= <class name> {where {<classop> == <target syntax>}+}?
   129 
   130 code_instance and code_class only apply to target Haskell.
   131 
   132 See HOL theories and HOL/ex/Codegenerator*.thy for usage examples.
   133 Doc/Isar/Advanced/Codegen/ provides a tutorial.
   134 
   135 * Command 'no_translations' removes translation rules from theory
   136 syntax.
   137 
   138 * Overloaded definitions are now actually checked for acyclic
   139 dependencies.  The overloading scheme is slightly more general than
   140 that of Haskell98, although Isabelle does not demand an exact
   141 correspondence to type class and instance declarations.
   142 INCOMPATIBILITY, use ``defs (unchecked overloaded)'' to admit more
   143 exotic versions of overloading -- at the discretion of the user!
   144 
   145 Polymorphic constants are represented via type arguments, i.e. the
   146 instantiation that matches an instance against the most general
   147 declaration given in the signature.  For example, with the declaration
   148 c :: 'a => 'a => 'a, an instance c :: nat => nat => nat is represented
   149 as c(nat).  Overloading is essentially simultaneous structural
   150 recursion over such type arguments.  Incomplete specification patterns
   151 impose global constraints on all occurrences, e.g. c('a * 'a) on the
   152 LHS means that more general c('a * 'b) will be disallowed on any RHS.
   153 Command 'print_theory' outputs the normalized system of recursive
   154 equations, see section "definitions".
   155 
   156 * Isar: improper proof element 'guess' is like 'obtain', but derives
   157 the obtained context from the course of reasoning!  For example:
   158 
   159   assume "EX x y. A x & B y"   -- "any previous fact"
   160   then guess x and y by clarify
   161 
   162 This technique is potentially adventurous, depending on the facts and
   163 proof tools being involved here.
   164 
   165 * Isar: known facts from the proof context may be specified as literal
   166 propositions, using ASCII back-quote syntax.  This works wherever
   167 named facts used to be allowed so far, in proof commands, proof
   168 methods, attributes etc.  Literal facts are retrieved from the context
   169 according to unification of type and term parameters.  For example,
   170 provided that "A" and "A ==> B" and "!!x. P x ==> Q x" are known
   171 theorems in the current context, then these are valid literal facts:
   172 `A` and `A ==> B` and `!!x. P x ==> Q x" as well as `P a ==> Q a` etc.
   173 
   174 There is also a proof method "fact" which does the same composition
   175 for explicit goal states, e.g. the following proof texts coincide with
   176 certain special cases of literal facts:
   177 
   178   have "A" by fact                 ==  note `A`
   179   have "A ==> B" by fact           ==  note `A ==> B`
   180   have "!!x. P x ==> Q x" by fact  ==  note `!!x. P x ==> Q x`
   181   have "P a ==> Q a" by fact       ==  note `P a ==> Q a`
   182 
   183 * Isar: ":" (colon) is no longer a symbolic identifier character in
   184 outer syntax.  Thus symbolic identifiers may be used without
   185 additional white space in declarations like this: ``assume *: A''.
   186 
   187 * Isar: 'print_facts' prints all local facts of the current context,
   188 both named and unnamed ones.
   189 
   190 * Isar: 'def' now admits simultaneous definitions, e.g.:
   191 
   192   def x == "t" and y == "u"
   193 
   194 * Isar: added command 'unfolding', which is structurally similar to
   195 'using', but affects both the goal state and facts by unfolding given
   196 rewrite rules.  Thus many occurrences of the 'unfold' method or
   197 'unfolded' attribute may be replaced by first-class proof text.
   198 
   199 * Isar: methods 'unfold' / 'fold', attributes 'unfolded' / 'folded',
   200 and command 'unfolding' now all support object-level equalities
   201 (potentially conditional).  The underlying notion of rewrite rule is
   202 analogous to the 'rule_format' attribute, but *not* that of the
   203 Simplifier (which is usually more generous).
   204 
   205 * Isar: the goal restriction operator [N] (default N = 1) evaluates a
   206 method expression within a sandbox consisting of the first N
   207 sub-goals, which need to exist.  For example, ``simp_all [3]''
   208 simplifies the first three sub-goals, while (rule foo, simp_all)[]
   209 simplifies all new goals that emerge from applying rule foo to the
   210 originally first one.
   211 
   212 * Isar: schematic goals are no longer restricted to higher-order
   213 patterns; e.g. ``lemma "?P(?x)" by (rule TrueI)'' now works as
   214 expected.
   215 
   216 * Isar: the conclusion of a long theorem statement is now either
   217 'shows' (a simultaneous conjunction, as before), or 'obtains'
   218 (essentially a disjunction of cases with local parameters and
   219 assumptions).  The latter allows to express general elimination rules
   220 adequately; in this notation common elimination rules look like this:
   221 
   222   lemma exE:    -- "EX x. P x ==> (!!x. P x ==> thesis) ==> thesis"
   223     assumes "EX x. P x"
   224     obtains x where "P x"
   225 
   226   lemma conjE:  -- "A & B ==> (A ==> B ==> thesis) ==> thesis"
   227     assumes "A & B"
   228     obtains A and B
   229 
   230   lemma disjE:  -- "A | B ==> (A ==> thesis) ==> (B ==> thesis) ==> thesis"
   231     assumes "A | B"
   232     obtains
   233       A
   234     | B
   235 
   236 The subsequent classical rules even refer to the formal "thesis"
   237 explicitly:
   238 
   239   lemma classical:     -- "(~ thesis ==> thesis) ==> thesis"
   240     obtains "~ thesis"
   241 
   242   lemma Peirce's_Law:  -- "((thesis ==> something) ==> thesis) ==> thesis"
   243     obtains "thesis ==> something"
   244 
   245 The actual proof of an 'obtains' statement is analogous to that of the
   246 Isar proof element 'obtain', only that there may be several cases.
   247 Optional case names may be specified in parentheses; these will be
   248 available both in the present proof and as annotations in the
   249 resulting rule, for later use with the 'cases' method (cf. attribute
   250 case_names).
   251 
   252 * Isar: the assumptions of a long theorem statement are available as
   253 "assms" fact in the proof context.  This is more appropriate than the
   254 (historical) "prems", which refers to all assumptions of the current
   255 context, including those from the target locale, proof body etc.
   256 
   257 * Isar: 'print_statement' prints theorems from the current theory or
   258 proof context in long statement form, according to the syntax of a
   259 top-level lemma.
   260 
   261 * Isar: 'obtain' takes an optional case name for the local context
   262 introduction rule (default "that").
   263 
   264 * Isar: removed obsolete 'concl is' patterns.  INCOMPATIBILITY, use
   265 explicit (is "_ ==> ?foo") in the rare cases where this still happens
   266 to occur.
   267 
   268 * Pure: syntax "CONST name" produces a fully internalized constant
   269 according to the current context.  This is particularly useful for
   270 syntax translations that should refer to internal constant
   271 representations independently of name spaces.
   272 
   273 * Pure: syntax constant for foo (binder "FOO ") is called "foo_binder"
   274 instead of "FOO ". This allows multiple binder declarations to coexist
   275 in the same context.  INCOMPATIBILITY.
   276 
   277 * Isar/locales: 'notation' provides a robust interface to the 'syntax'
   278 primitive that also works in a locale context (both for constants and
   279 fixed variables).  Type declaration and internal syntactic
   280 representation of given constants retrieved from the context.
   281 
   282 * Isar/locales: new derived specification elements 'axiomatization',
   283 'definition', 'abbreviation', which support type-inference, admit
   284 object-level specifications (equality, equivalence).  See also the
   285 isar-ref manual.  Examples:
   286 
   287   axiomatization
   288     eq  (infix "===" 50) where
   289     eq_refl: "x === x" and eq_subst: "x === y ==> P x ==> P y"
   290 
   291   definition "f x y = x + y + 1"
   292   definition g where "g x = f x x"
   293 
   294   abbreviation
   295     neq  (infix "=!=" 50) where
   296     "x =!= y == ~ (x === y)"
   297 
   298 These specifications may be also used in a locale context.  Then the
   299 constants being introduced depend on certain fixed parameters, and the
   300 constant name is qualified by the locale base name.  An internal
   301 abbreviation takes care for convenient input and output, making the
   302 parameters implicit and using the original short name.  See also
   303 HOL/ex/Abstract_NAT.thy for an example of deriving polymorphic
   304 entities from a monomorphic theory.
   305 
   306 Presently, abbreviations are only available 'in' a target locale, but
   307 not inherited by general import expressions.  Also note that
   308 'abbreviation' may be used as a type-safe replacement for 'syntax' +
   309 'translations' in common applications.
   310 
   311 Concrete syntax is attached to specified constants in internal form,
   312 independently of name spaces.  The parse tree representation is
   313 slightly different -- use 'notation' instead of raw 'syntax', and
   314 'translations' with explicit "CONST" markup to accommodate this.
   315 
   316 * Pure: command 'print_abbrevs' prints all constant abbreviations of
   317 the current context.  Print mode "no_abbrevs" prevents inversion of
   318 abbreviations on output.
   319 
   320 * Isar/locales: improved parameter handling:
   321 - use of locales "var" and "struct" no longer necessary;
   322 - parameter renamings are no longer required to be injective.
   323   This enables, for example, to define a locale for endomorphisms thus:
   324   locale endom = homom mult mult h.
   325 
   326 * Isar/locales: changed the way locales with predicates are defined.
   327 Instead of accumulating the specification, the imported expression is
   328 now an interpretation.  INCOMPATIBILITY: different normal form of
   329 locale expressions.  In particular, in interpretations of locales with
   330 predicates, goals repesenting already interpreted fragments are not
   331 removed automatically.  Use methods `intro_locales' and
   332 `unfold_locales'; see below.
   333 
   334 * Isar/locales: new methods `intro_locales' and `unfold_locales'
   335 provide backward reasoning on locales predicates.  The methods are
   336 aware of interpretations and discharge corresponding goals.
   337 `intro_locales' is less aggressive then `unfold_locales' and does not
   338 unfold predicates to assumptions.
   339 
   340 * Isar/locales: the order in which locale fragments are accumulated
   341 has changed.  This enables to override declarations from fragments due
   342 to interpretations -- for example, unwanted simp rules.
   343 
   344 * Provers/induct: improved internal context management to support
   345 local fixes and defines on-the-fly.  Thus explicit meta-level
   346 connectives !! and ==> are rarely required anymore in inductive goals
   347 (using object-logic connectives for this purpose has been long
   348 obsolete anyway).  The subsequent proof patterns illustrate advanced
   349 techniques of natural induction; general datatypes and inductive sets
   350 work analogously (see also src/HOL/Lambda for realistic examples).
   351 
   352 (1) This is how to ``strengthen'' an inductive goal wrt. certain
   353 parameters:
   354 
   355   lemma
   356     fixes n :: nat and x :: 'a
   357     assumes a: "A n x"
   358     shows "P n x"
   359     using a                     -- {* make induct insert fact a *}
   360   proof (induct n arbitrary: x) -- {* generalize goal to "!!x. A n x ==> P n x" *}
   361     case 0
   362     show ?case sorry
   363   next
   364     case (Suc n)
   365     note `!!x. A n x ==> P n x` -- {* induction hypothesis, according to induction rule *}
   366     note `A (Suc n) x`          -- {* induction premise, stemming from fact a *}
   367     show ?case sorry
   368   qed
   369 
   370 (2) This is how to perform induction over ``expressions of a certain
   371 form'', using a locally defined inductive parameter n == "a x"
   372 together with strengthening (the latter is usually required to get
   373 sufficiently flexible induction hypotheses):
   374 
   375   lemma
   376     fixes a :: "'a => nat"
   377     assumes a: "A (a x)"
   378     shows "P (a x)"
   379     using a
   380   proof (induct n == "a x" arbitrary: x)
   381     ...
   382 
   383 See also HOL/Isar_examples/Puzzle.thy for an application of the this
   384 particular technique.
   385 
   386 (3) This is how to perform existential reasoning ('obtains' or
   387 'obtain') by induction, while avoiding explicit object-logic
   388 encodings:
   389 
   390   lemma
   391     fixes n :: nat
   392     obtains x :: 'a where "P n x" and "Q n x"
   393   proof (induct n arbitrary: thesis)
   394     case 0
   395     obtain x where "P 0 x" and "Q 0 x" sorry
   396     then show thesis by (rule 0)
   397   next
   398     case (Suc n)
   399     obtain x where "P n x" and "Q n x" by (rule Suc.hyps)
   400     obtain x where "P (Suc n) x" and "Q (Suc n) x" sorry
   401     then show thesis by (rule Suc.prems)
   402   qed
   403 
   404 Here the 'arbitrary: thesis' specification essentially modifies the
   405 scope of the formal thesis parameter, in order to the get the whole
   406 existence statement through the induction as expected.
   407 
   408 * Provers/induct: mutual induction rules are now specified as a list
   409 of rule sharing the same induction cases.  HOL packages usually
   410 provide foo_bar.inducts for mutually defined items foo and bar
   411 (e.g. inductive sets or datatypes).  INCOMPATIBILITY, users need to
   412 specify mutual induction rules differently, i.e. like this:
   413 
   414   (induct rule: foo_bar.inducts)
   415   (induct set: foo bar)
   416   (induct type: foo bar)
   417 
   418 The ML function ProjectRule.projections turns old-style rules into the
   419 new format.
   420 
   421 * Provers/induct: improved handling of simultaneous goals.  Instead of
   422 introducing object-level conjunction, the statement is now split into
   423 several conclusions, while the corresponding symbolic cases are
   424 nested accordingly.  INCOMPATIBILITY, proofs need to be structured
   425 explicitly.  For example:
   426 
   427   lemma
   428     fixes n :: nat
   429     shows "P n" and "Q n"
   430   proof (induct n)
   431     case 0 case 1
   432     show "P 0" sorry
   433   next
   434     case 0 case 2
   435     show "Q 0" sorry
   436   next
   437     case (Suc n) case 1
   438     note `P n` and `Q n`
   439     show "P (Suc n)" sorry
   440   next
   441     case (Suc n) case 2
   442     note `P n` and `Q n`
   443     show "Q (Suc n)" sorry
   444   qed
   445 
   446 The split into subcases may be deferred as follows -- this is
   447 particularly relevant for goal statements with local premises.
   448 
   449   lemma
   450     fixes n :: nat
   451     shows "A n ==> P n" and "B n ==> Q n"
   452   proof (induct n)
   453     case 0
   454     {
   455       case 1
   456       note `A 0`
   457       show "P 0" sorry
   458     next
   459       case 2
   460       note `B 0`
   461       show "Q 0" sorry
   462     }
   463   next
   464     case (Suc n)
   465     note `A n ==> P n` and `B n ==> Q n`
   466     {
   467       case 1
   468       note `A (Suc n)`
   469       show "P (Suc n)" sorry
   470     next
   471       case 2
   472       note `B (Suc n)`
   473       show "Q (Suc n)" sorry
   474     }
   475   qed
   476 
   477 If simultaneous goals are to be used with mutual rules, the statement
   478 needs to be structured carefully as a two-level conjunction, using
   479 lists of propositions separated by 'and':
   480 
   481   lemma
   482     shows "a : A ==> P1 a"
   483           "a : A ==> P2 a"
   484       and "b : B ==> Q1 b"
   485           "b : B ==> Q2 b"
   486           "b : B ==> Q3 b"
   487   proof (induct set: A B)
   488 
   489 * Provers/induct: support coinduction as well.  See
   490 src/HOL/Library/Coinductive_List.thy for various examples.
   491 
   492 * Attribute "symmetric" produces result with standardized schematic
   493 variables (index 0).  Potential INCOMPATIBILITY.
   494 
   495 * Simplifier: by default the simplifier trace only shows top level
   496 rewrites now. That is, trace_simp_depth_limit is set to 1 by
   497 default. Thus there is less danger of being flooded by the trace. The
   498 trace indicates where parts have been suppressed.
   499   
   500 * Provers/classical: removed obsolete classical version of elim_format
   501 attribute; classical elim/dest rules are now treated uniformly when
   502 manipulating the claset.
   503 
   504 * Provers/classical: stricter checks to ensure that supplied intro,
   505 dest and elim rules are well-formed; dest and elim rules must have at
   506 least one premise.
   507 
   508 * Provers/classical: attributes dest/elim/intro take an optional
   509 weight argument for the rule (just as the Pure versions).  Weights are
   510 ignored by automated tools, but determine the search order of single
   511 rule steps.
   512 
   513 * Syntax: input syntax now supports dummy variable binding "%_. b",
   514 where the body does not mention the bound variable.  Note that dummy
   515 patterns implicitly depend on their context of bounds, which makes
   516 "{_. _}" match any set comprehension as expected.  Potential
   517 INCOMPATIBILITY -- parse translations need to cope with syntactic
   518 constant "_idtdummy" in the binding position.
   519 
   520 * Syntax: removed obsolete syntactic constant "_K" and its associated
   521 parse translation.  INCOMPATIBILITY -- use dummy abstraction instead,
   522 for example "A -> B" => "Pi A (%_. B)".
   523 
   524 * Pure: 'class_deps' command visualizes the subclass relation, using
   525 the graph browser tool.
   526 
   527 * Pure: 'print_theory' now suppresses entities with internal name
   528 (trailing "_") by default; use '!' option for full details.
   529 
   530 
   531 *** HOL ***
   532 
   533 * Method "algebra" solves polynomial equations over (semi)rings using
   534   Groebner bases. The (semi)ring structure is defined by locales and
   535   the tool setup depends on that generic context. Installing the
   536   method for a specific type involves instantiating the locale and
   537   possibly adding declarations for computation on the coefficients.
   538   The method is already instantiated for natural numbers and for the
   539   axiomatic class of idoms with numerals.  See also the paper by
   540   Chaieb and Wenzel at CALCULEMUS 2007 for the general principles
   541   underlying this architecture of context-aware proof-tools.
   542 
   543 * constant "List.op @" now named "List.append".  Use ML antiquotations
   544 @{const_name List.append} or @{term " ... @ ... "} to circumvent
   545 possible incompatibilities when working on ML level.
   546 
   547 * Constant renames due to introduction of canonical name prefixing for
   548   class package:
   549 
   550     HOL.abs ~> HOL.minus_class.abs
   551     HOL.divide ~> HOL.divide_class.divide
   552     Nat.power ~> Nat.power_class.power
   553     Nat.size ~> Nat.size_class.size
   554     Numeral.number_of ~> Numeral.number_class.number_of
   555     FixedPoint.Inf ~> FixedPoint.complete_lattice_class.Inf
   556 
   557 * Rudimentary class target mechanism involves constant renames:
   558 
   559     Orderings.min ~> Orderings.ord_class.min
   560     Orderings.max ~> Orderings.ord_class.max
   561     FixedPoint.Sup ~> FixedPoint.complete_lattice_class.Sup
   562 
   563 * case expressions and primrec: missing cases now mapped to "undefined"
   564 instead of "arbitrary"
   565 
   566 * new constant "undefined" with axiom "undefined x = undefined"
   567 
   568 * new class "default" with associated constant "default"
   569 
   570 * new function listsum :: 'a list => 'a for arbitrary monoids.
   571   Special syntax: "SUM x <- xs. f x" (and latex variants)
   572 
   573 * new (input only) syntax for Haskell-like list comprehension, eg
   574   [(x,y). x <- xs, y <- ys, x ~= y]
   575   For details see List.thy.
   576 
   577 * Library/Pretty_Int.thy: maps HOL numerals on target language integer literals
   578     when generating code.
   579 
   580 * Library/Pretty_Char.thy: maps HOL characters on target language character literals
   581     when generating code.
   582 
   583 * Library/Commutative_Ring.thy: switched from recdef to function package;
   584     constants add, mul, pow now curried.  Infix syntax for algebraic operations.
   585 
   586 * Some steps towards more uniform lattice theory development in HOL.
   587 
   588     constants "meet" and "join" now named "inf" and "sup"
   589     constant "Meet" now named "Inf"
   590 
   591     classes "meet_semilorder" and "join_semilorder" now named
   592       "lower_semilattice" and "upper_semilattice"
   593     class "lorder" now named "lattice"
   594     class "comp_lat" now named "complete_lattice"
   595 
   596     Instantiation of lattice classes allows explicit definitions
   597     for "inf" and "sup" operations.
   598 
   599   INCOMPATIBILITY.  Theorem renames:
   600 
   601     meet_left_le            ~> inf_le1
   602     meet_right_le           ~> inf_le2
   603     join_left_le            ~> sup_ge1
   604     join_right_le           ~> sup_ge2
   605     meet_join_le            ~> inf_sup_ord
   606     le_meetI                ~> le_infI
   607     join_leI                ~> le_supI
   608     le_meet                 ~> le_inf_iff
   609     le_join                 ~> ge_sup_conv
   610     meet_idempotent         ~> inf_idem
   611     join_idempotent         ~> sup_idem
   612     meet_comm               ~> inf_commute
   613     join_comm               ~> sup_commute
   614     meet_leI1               ~> le_infI1
   615     meet_leI2               ~> le_infI2
   616     le_joinI1               ~> le_supI1
   617     le_joinI2               ~> le_supI2
   618     meet_assoc              ~> inf_assoc
   619     join_assoc              ~> sup_assoc
   620     meet_left_comm          ~> inf_left_commute
   621     meet_left_idempotent    ~> inf_left_idem
   622     join_left_comm          ~> sup_left_commute
   623     join_left_idempotent    ~> sup_left_idem
   624     meet_aci                ~> inf_aci
   625     join_aci                ~> sup_aci
   626     le_def_meet             ~> le_iff_inf
   627     le_def_join             ~> le_iff_sup
   628     join_absorp2            ~> sup_absorb2
   629     join_absorp1            ~> sup_absorb1
   630     meet_absorp1            ~> inf_absorb1
   631     meet_absorp2            ~> inf_absorb2
   632     meet_join_absorp        ~> inf_sup_absorb
   633     join_meet_absorp        ~> sup_inf_absorb
   634     distrib_join_le         ~> distrib_sup_le
   635     distrib_meet_le         ~> distrib_inf_le
   636 
   637     add_meet_distrib_left   ~> add_inf_distrib_left
   638     add_join_distrib_left   ~> add_sup_distrib_left
   639     is_join_neg_meet        ~> is_join_neg_inf
   640     is_meet_neg_join        ~> is_meet_neg_sup
   641     add_meet_distrib_right  ~> add_inf_distrib_right
   642     add_join_distrib_right  ~> add_sup_distrib_right
   643     add_meet_join_distribs  ~> add_sup_inf_distribs
   644     join_eq_neg_meet        ~> sup_eq_neg_inf
   645     meet_eq_neg_join        ~> inf_eq_neg_sup
   646     add_eq_meet_join        ~> add_eq_inf_sup
   647     meet_0_imp_0            ~> inf_0_imp_0
   648     join_0_imp_0            ~> sup_0_imp_0
   649     meet_0_eq_0             ~> inf_0_eq_0
   650     join_0_eq_0             ~> sup_0_eq_0
   651     neg_meet_eq_join        ~> neg_inf_eq_sup
   652     neg_join_eq_meet        ~> neg_sup_eq_inf
   653     join_eq_if              ~> sup_eq_if
   654 
   655     mono_meet               ~> mono_inf
   656     mono_join               ~> mono_sup
   657     meet_bool_eq            ~> inf_bool_eq
   658     join_bool_eq            ~> sup_bool_eq
   659     meet_fun_eq             ~> inf_fun_eq
   660     join_fun_eq             ~> sup_fun_eq
   661     meet_set_eq             ~> inf_set_eq
   662     join_set_eq             ~> sup_set_eq
   663     meet1_iff               ~> inf1_iff
   664     meet2_iff               ~> inf2_iff
   665     meet1I                  ~> inf1I
   666     meet2I                  ~> inf2I
   667     meet1D1                 ~> inf1D1
   668     meet2D1                 ~> inf2D1
   669     meet1D2                 ~> inf1D2
   670     meet2D2                 ~> inf2D2
   671     meet1E                  ~> inf1E
   672     meet2E                  ~> inf2E
   673     join1_iff               ~> sup1_iff
   674     join2_iff               ~> sup2_iff
   675     join1I1                 ~> sup1I1
   676     join2I1                 ~> sup2I1
   677     join1I1                 ~> sup1I1
   678     join2I2                 ~> sup1I2
   679     join1CI                 ~> sup1CI
   680     join2CI                 ~> sup2CI
   681     join1E                  ~> sup1E
   682     join2E                  ~> sup2E
   683 
   684     is_meet_Meet            ~> is_meet_Inf
   685     Meet_bool_def           ~> Inf_bool_def
   686     Meet_fun_def            ~> Inf_fun_def
   687     Meet_greatest           ~> Inf_greatest
   688     Meet_lower              ~> Inf_lower
   689     Meet_set_def            ~> Inf_set_def
   690 
   691     listsp_meetI            ~> listsp_infI
   692     listsp_meet_eq          ~> listsp_inf_eq
   693 
   694     meet_min                ~> inf_min
   695     join_max                ~> sup_max
   696 
   697 * Classes "order" and "linorder": facts "refl", "trans" and
   698 "cases" renamed ro "order_refl", "order_trans" and "linorder_cases", to
   699 avoid clashes with HOL "refl" and "trans". INCOMPATIBILITY.
   700 
   701 * Classes "order" and "linorder": 
   702 potential INCOMPATIBILITY: order of proof goals in order/linorder instance
   703 proofs changed.
   704 
   705 * Dropped lemma duplicate def_imp_eq in favor of meta_eq_to_obj_eq.
   706 INCOMPATIBILITY.
   707 
   708 * Dropped lemma duplicate if_def2 in favor of if_bool_eq_conj.
   709 INCOMPATIBILITY.
   710 
   711 * Added syntactic class "size"; overloaded constant "size" now has
   712 type "'a::size ==> bool"
   713 
   714 * Renamed constants "Divides.op div", "Divides.op mod" and "Divides.op
   715 dvd" to "Divides.div_class.div", "Divides.div_class.mod" and "Divides.dvd". INCOMPATIBILITY.
   716 
   717 * Added method "lexicographic_order" automatically synthesizes
   718 termination relations as lexicographic combinations of size measures
   719 -- 'function' package.
   720 
   721 * HOL/records: generalised field-update to take a function on the
   722 field rather than the new value: r(|A := x|) is translated to A_update
   723 (K x) r The K-combinator that is internally used is called K_record.
   724 INCOMPATIBILITY: Usage of the plain update functions has to be
   725 adapted.
   726  
   727 * axclass "semiring_0" now contains annihilation axioms x * 0 = 0 and
   728 0 * x = 0, which are required for a semiring.  Richer structures do
   729 not inherit from semiring_0 anymore, because this property is a
   730 theorem there, not an axiom.  INCOMPATIBILITY: In instances of
   731 semiring_0, there is more to prove, but this is mostly trivial.
   732 
   733 * axclass "recpower" was generalized to arbitrary monoids, not just
   734 commutative semirings.  INCOMPATIBILITY: If you use recpower and need
   735 commutativity or a semiring property, add the corresponding classes.
   736 
   737 * Unified locale partial_order with class definition (cf. theory
   738 Orderings), added parameter ``less''.  INCOMPATIBILITY.
   739 
   740 * Constant "List.list_all2" in List.thy now uses authentic syntax.
   741 INCOMPATIBILITY: translations containing list_all2 may go wrong.  On
   742 Isar level, use abbreviations instead.
   743 
   744 * Renamed constant "List.op mem" to "List.memberl" INCOMPATIBILITY:
   745 rarely occuring name references (e.g. ``List.op mem.simps'') require
   746 renaming (e.g. ``List.memberl.simps'').
   747 
   748 * Renamed constants "0" to "HOL.zero_class.zero" and "1" to "HOL.one_class.one".
   749 INCOMPATIBILITY.
   750 
   751 * Added class "HOL.eq", allowing for code generation with polymorphic equality.
   752 
   753 * Numeral syntax: type 'bin' which was a mere type copy of 'int' has
   754 been abandoned in favour of plain 'int'. INCOMPATIBILITY --
   755 significant changes for setting up numeral syntax for types:
   756 
   757   - new constants Numeral.pred and Numeral.succ instead
   758       of former Numeral.bin_pred and Numeral.bin_succ.
   759   - Use integer operations instead of bin_add, bin_mult and so on.
   760   - Numeral simplification theorems named Numeral.numeral_simps instead of Bin_simps.
   761   - ML structure Bin_Simprocs now named Int_Numeral_Base_Simprocs.
   762 
   763 See HOL/Integ/IntArith.thy for an example setup.
   764 
   765 * New top level command 'normal_form' computes the normal form of a
   766 term that may contain free variables. For example ``normal_form
   767 "rev[a,b,c]"'' produces ``[b,c,a]'' (without proof).  This command is
   768 suitable for heavy-duty computations because the functions are
   769 compiled to ML first.
   770 
   771 * Alternative iff syntax "A <-> B" for equality on bool (with priority
   772 25 like -->); output depends on the "iff" print_mode, the default is
   773 "A = B" (with priority 50).
   774 
   775 * Renamed constants in HOL.thy and Orderings.thy:
   776     op +   ~> HOL.plus_class.plus
   777     op -   ~> HOL.minus_class.minus
   778     uminus ~> HOL.minus_class.uminus
   779     op *   ~> HOL.times_class.times
   780     op <   ~> Orderings.ord_class.less
   781     op <=  ~> Orderings.ord_class.less_eq
   782 
   783 Adaptions may be required in the following cases:
   784 
   785 a) User-defined constants using any of the names "plus", "minus", "times",
   786 "less" or "less_eq". The standard syntax translations for "+", "-" and "*"
   787 may go wrong.
   788 INCOMPATIBILITY: use more specific names.
   789 
   790 b) Variables named "plus", "minus", "times", "less", "less_eq"
   791 INCOMPATIBILITY: use more specific names.
   792 
   793 c) Permutative equations (e.g. "a + b = b + a")
   794 Since the change of names also changes the order of terms, permutative
   795 rewrite rules may get applied in a different order. Experience shows that
   796 this is rarely the case (only two adaptions in the whole Isabelle
   797 distribution).
   798 INCOMPATIBILITY: rewrite proofs
   799 
   800 d) ML code directly refering to constant names
   801 This in general only affects hand-written proof tactics, simprocs and so on.
   802 INCOMPATIBILITY: grep your sourcecode and replace names.  Consider use
   803 of const_name ML antiquotations.
   804 
   805 * Relations less (<) and less_eq (<=) are also available on type bool.
   806 Modified syntax to disallow nesting without explicit parentheses,
   807 e.g. "(x < y) < z" or "x < (y < z)", but NOT "x < y < z".
   808 
   809 * "LEAST x:A. P" expands to "LEAST x. x:A & P" (input only).
   810 
   811 * Relation composition operator "op O" now has precedence 75 and binds
   812 stronger than union and intersection. INCOMPATIBILITY.
   813 
   814 * The old set interval syntax "{m..n(}" (and relatives) has been
   815 removed.  Use "{m..<n}" (and relatives) instead.
   816 
   817 * In the context of the assumption "~(s = t)" the Simplifier rewrites
   818 "t = s" to False (by simproc "neq_simproc").  For backward
   819 compatibility this can be disabled by ML "reset use_neq_simproc".
   820 
   821 * "m dvd n" where m and n are numbers is evaluated to True/False by
   822 simp.
   823 
   824 * Theorem Cons_eq_map_conv no longer declared as ``simp''.
   825 
   826 * Theorem setsum_mult renamed to setsum_right_distrib.
   827 
   828 * Prefer ex1I over ex_ex1I in single-step reasoning, e.g. by the
   829 ``rule'' method.
   830 
   831 * Reimplemented methods ``sat'' and ``satx'', with several
   832 improvements: goals no longer need to be stated as "<prems> ==>
   833 False", equivalences (i.e. "=" on type bool) are handled, variable
   834 names of the form "lit_<n>" are no longer reserved, significant
   835 speedup.
   836 
   837 * Methods ``sat'' and ``satx'' can now replay MiniSat proof traces.
   838 zChaff is still supported as well.
   839 
   840 * 'inductive' and 'datatype': provide projections of mutual rules,
   841 bundled as foo_bar.inducts;
   842 
   843 * Library: moved theories Parity, GCD, Binomial, Infinite_Set to
   844 Library.
   845 
   846 * Library: moved theory Accessible_Part to main HOL.
   847 
   848 * Library: added theory Coinductive_List of potentially infinite lists
   849 as greatest fixed-point.
   850 
   851 * Library: added theory AssocList which implements (finite) maps as
   852 association lists.
   853 
   854 * Added proof method ``evaluation'' for efficiently solving a goal
   855 (i.e. a boolean expression) by compiling it to ML. The goal is
   856 "proved" (via an oracle) if it evaluates to True.
   857 
   858 * Linear arithmetic now splits certain operators (e.g. min, max, abs)
   859 also when invoked by the simplifier.  This results in the simplifier
   860 being more powerful on arithmetic goals.  INCOMPATIBILITY.  Set
   861 fast_arith_split_limit to 0 to obtain the old behavior.
   862 
   863 * Support for hex (0x20) and binary (0b1001) numerals.
   864 
   865 * New method: reify eqs (t), where eqs are equations for an
   866 interpretation I :: 'a list => 'b => 'c and t::'c is an optional
   867 parameter, computes a term s::'b and a list xs::'a list and proves the
   868 theorem I xs s = t. This is also known as reification or quoting. The
   869 resulting theorem is applied to the subgoal to substitute t with I xs
   870 s.  If t is omitted, the subgoal itself is reified.
   871 
   872 * New method: reflection corr_thm eqs (t). The parameters eqs and (t)
   873 are as explained above. corr_thm is a theorem for I vs (f t) = I vs t,
   874 where f is supposed to be a computable function (in the sense of code
   875 generattion). The method uses reify to compute s and xs as above then
   876 applies corr_thm and uses normalization by evaluation to "prove" f s =
   877 r and finally gets the theorem t = r, which is again applied to the
   878 subgoal. An Example is available in HOL/ex/ReflectionEx.thy.
   879 
   880 * Reflection: Automatic refification now handels binding, an example
   881 is available in HOL/ex/ReflectionEx.thy
   882 
   883 
   884 *** HOL-Algebra ***
   885 
   886 * Formalisation of ideals and the quotient construction over rings.
   887 
   888 * Order and lattice theory no longer based on records.
   889 INCOMPATIBILITY.
   890 
   891 * Renamed lemmas least_carrier -> least_closed and greatest_carrier ->
   892 greatest_closed.  INCOMPATIBILITY.
   893 
   894 * Method algebra is now set up via an attribute.  For examples see
   895 Ring.thy.  INCOMPATIBILITY: the method is now weaker on combinations
   896 of algebraic structures.
   897 
   898 * Renamed theory CRing to Ring.
   899 
   900 
   901 *** HOL-Complex ***
   902 
   903 * Theory Real: new method ferrack implements quantifier elimination
   904 for linear arithmetic over the reals. The quantifier elimination
   905 feature is used only for decision, for compatibility with arith. This
   906 means a goal is either solved or left unchanged, no simplification.
   907 
   908 * Hyperreal: Functions root and sqrt are now defined on negative real
   909 inputs so that root n (- x) = - root n x and sqrt (- x) = - sqrt x.
   910 Nonnegativity side conditions have been removed from many lemmas, so
   911 that more subgoals may now be solved by simplification; potential
   912 INCOMPATIBILITY.
   913 
   914 * Real: New axiomatic classes formalize real normed vector spaces and
   915 algebras, using new overloaded constants scaleR :: real => 'a => 'a
   916 and norm :: 'a => real.
   917 
   918 * Real: New constant of_real :: real => 'a::real_algebra_1 injects
   919 from reals into other types. The overloaded constant Reals :: 'a set
   920 is now defined as range of_real; potential INCOMPATIBILITY.
   921 
   922 * Real: ML code generation is supported now and hence also quickcheck.
   923 Reals are implemented as arbitrary precision rationals.
   924 
   925 * Hyperreal: Several constants that previously worked only for the
   926 reals have been generalized, so they now work over arbitrary vector
   927 spaces. Type annotations may need to be added in some cases; potential
   928 INCOMPATIBILITY.
   929 
   930   Infinitesimal  :: ('a::real_normed_vector) star set
   931   HFinite        :: ('a::real_normed_vector) star set
   932   HInfinite      :: ('a::real_normed_vector) star set
   933   approx         :: ('a::real_normed_vector) star => 'a star => bool
   934   monad          :: ('a::real_normed_vector) star => 'a star set
   935   galaxy         :: ('a::real_normed_vector) star => 'a star set
   936   (NS)LIMSEQ     :: [nat => 'a::real_normed_vector, 'a] => bool
   937   (NS)convergent :: (nat => 'a::real_normed_vector) => bool
   938   (NS)Bseq       :: (nat => 'a::real_normed_vector) => bool
   939   (NS)Cauchy     :: (nat => 'a::real_normed_vector) => bool
   940   (NS)LIM        :: ['a::real_normed_vector => 'b::real_normed_vector, 'a, 'b] => bool
   941   is(NS)Cont     :: ['a::real_normed_vector => 'b::real_normed_vector, 'a] => bool
   942   deriv          :: ['a::real_normed_field => 'a, 'a, 'a] => bool
   943   sgn            :: 'a::real_normed_vector => 'a
   944   exp            :: 'a::{recpower,real_normed_field,banach} => 'a
   945 
   946 * Complex: Some complex-specific constants are now abbreviations for
   947 overloaded ones: complex_of_real = of_real, cmod = norm, hcmod =
   948 hnorm.  Other constants have been entirely removed in favor of the
   949 polymorphic versions (INCOMPATIBILITY):
   950 
   951   approx        <-- capprox
   952   HFinite       <-- CFinite
   953   HInfinite     <-- CInfinite
   954   Infinitesimal <-- CInfinitesimal
   955   monad         <-- cmonad
   956   galaxy        <-- cgalaxy
   957   (NS)LIM       <-- (NS)CLIM, (NS)CRLIM
   958   is(NS)Cont    <-- is(NS)Contc, is(NS)contCR
   959   (ns)deriv     <-- (ns)cderiv
   960 
   961 
   962 *** ML ***
   963 
   964 * Generic arithmetic modules: Tools/integer.ML, Tools/rat.ML, Tools/float.ML
   965 
   966 * Context data interfaces (Theory/Proof/GenericDataFun): removed
   967 name/print, uninitialized data defaults to ad-hoc copy of empty value,
   968 init only required for impure data.  INCOMPATIBILITY: empty really
   969 need to be empty (no dependencies on theory content!)
   970 
   971 * ML within Isar: antiquotations allow to embed statically-checked
   972 formal entities in the source, referring to the context available at
   973 compile-time.  For example:
   974 
   975 ML {* @{typ "'a => 'b"} *}
   976 ML {* @{term "%x. x"} *}
   977 ML {* @{prop "x == y"} *}
   978 ML {* @{ctyp "'a => 'b"} *}
   979 ML {* @{cterm "%x. x"} *}
   980 ML {* @{cprop "x == y"} *}
   981 ML {* @{thm asm_rl} *}
   982 ML {* @{thms asm_rl} *}
   983 ML {* @{const_name c} *}
   984 ML {* @{const_syntax c} *}
   985 ML {* @{context} *}
   986 ML {* @{theory} *}
   987 ML {* @{theory Pure} *}
   988 ML {* @{simpset} *}
   989 ML {* @{claset} *}
   990 ML {* @{clasimpset} *}
   991 
   992 The same works for sources being ``used'' within an Isar context.
   993 
   994 * ML in Isar: improved error reporting; extra verbosity with
   995 Toplevel.debug enabled.
   996 
   997 * Pure/library:
   998 
   999   val burrow: ('a list -> 'b list) -> 'a list list -> 'b list list
  1000   val fold_burrow: ('a list -> 'c -> 'b list * 'd) -> 'a list list -> 'c -> 'b list list * 'd
  1001 
  1002 The semantics of "burrow" is: "take a function with *simulatanously*
  1003 transforms a list of value, and apply it *simulatanously* to a list of
  1004 list of values of the appropriate type". Compare this with "map" which
  1005 would *not* apply its argument function simulatanously but in
  1006 sequence; "fold_burrow" has an additional context.
  1007 
  1008 * Pure/library: functions map2 and fold2 with curried syntax for
  1009 simultanous mapping and folding:
  1010 
  1011     val map2: ('a -> 'b -> 'c) -> 'a list -> 'b list -> 'c list
  1012     val fold2: ('a -> 'b -> 'c -> 'c) -> 'a list -> 'b list -> 'c -> 'c
  1013 
  1014 * Pure/library: indexed lists - some functions in the Isabelle library
  1015 treating lists over 'a as finite mappings from [0...n] to 'a have been
  1016 given more convenient names and signatures reminiscent of similar
  1017 functions for alists, tables, etc:
  1018 
  1019   val nth: 'a list -> int -> 'a 
  1020   val nth_map: int -> ('a -> 'a) -> 'a list -> 'a list
  1021   val fold_index: (int * 'a -> 'b -> 'b) -> 'a list -> 'b -> 'b
  1022 
  1023 Note that fold_index starts counting at index 0, not 1 like foldln
  1024 used to.
  1025 
  1026 * Pure/library: added general ``divide_and_conquer'' combinator on
  1027 lists.
  1028 
  1029 * Pure/General/table.ML: the join operations now works via exceptions
  1030 DUP/SAME instead of type option.  This is simpler in simple cases, and
  1031 admits slightly more efficient complex applications.
  1032 
  1033 * Pure: datatype Context.generic joins theory/Proof.context and
  1034 provides some facilities for code that works in either kind of
  1035 context, notably GenericDataFun for uniform theory and proof data.
  1036 
  1037 * Pure: 'advanced' translation functions (parse_translation etc.) now
  1038 use Context.generic instead of just theory.
  1039 
  1040 * Pure: simplified internal attribute type, which is now always
  1041 Context.generic * thm -> Context.generic * thm.  Global (theory)
  1042 vs. local (Proof.context) attributes have been discontinued, while
  1043 minimizing code duplication.  Thm.rule_attribute and
  1044 Thm.declaration_attribute build canonical attributes; see also
  1045 structure Context for further operations on Context.generic, notably
  1046 GenericDataFun.  INCOMPATIBILITY, need to adapt attribute type
  1047 declarations and definitions.
  1048 
  1049 * Pure/kernel: consts certification ignores sort constraints given in
  1050 signature declarations.  (This information is not relevant to the
  1051 logic, but only for type inference.)  IMPORTANT INTERNAL CHANGE,
  1052 potential INCOMPATIBILITY.
  1053 
  1054 * Pure: axiomatic type classes are now purely definitional, with
  1055 explicit proofs of class axioms and super class relations performed
  1056 internally.  See Pure/axclass.ML for the main internal interfaces --
  1057 notably AxClass.define_class supercedes AxClass.add_axclass, and
  1058 AxClass.axiomatize_class/classrel/arity supercede
  1059 Sign.add_classes/classrel/arities.
  1060 
  1061 * Pure/Isar: Args/Attrib parsers operate on Context.generic --
  1062 global/local versions on theory vs. Proof.context have been
  1063 discontinued; Attrib.syntax and Method.syntax have been adapted
  1064 accordingly.  INCOMPATIBILITY, need to adapt parser expressions for
  1065 attributes, methods, etc.
  1066 
  1067 * Pure: several functions of signature "... -> theory -> theory * ..."
  1068 have been reoriented to "... -> theory -> ... * theory" in order to
  1069 allow natural usage in combination with the ||>, ||>>, |-> and
  1070 fold_map combinators.
  1071 
  1072 * Pure: official theorem names (closed derivations) and additional
  1073 comments (tags) are now strictly separate.  Name hints -- which are
  1074 maintained as tags -- may be attached any time without affecting the
  1075 derivation.
  1076 
  1077 * Pure: primitive rule lift_rule now takes goal cterm instead of an
  1078 actual goal state (thm).  Use Thm.lift_rule (Thm.cprem_of st i) to
  1079 achieve the old behaviour.
  1080 
  1081 * Pure: the "Goal" constant is now called "prop", supporting a
  1082 slightly more general idea of ``protecting'' meta-level rule
  1083 statements.
  1084 
  1085 * Pure: Logic.(un)varify only works in a global context, which is now
  1086 enforced instead of silently assumed.  INCOMPATIBILITY, may use
  1087 Logic.legacy_(un)varify as temporary workaround.
  1088 
  1089 * Pure: structure Name provides scalable operations for generating
  1090 internal variable names, notably Name.variants etc.  This replaces
  1091 some popular functions from term.ML:
  1092 
  1093   Term.variant		->  Name.variant
  1094   Term.variantlist	->  Name.variant_list  (*canonical argument order*)
  1095   Term.invent_names	->  Name.invent_list
  1096 
  1097 Note that low-level renaming rarely occurs in new code -- operations
  1098 from structure Variable are used instead (see below).
  1099 
  1100 * Pure: structure Variable provides fundamental operations for proper
  1101 treatment of fixed/schematic variables in a context.  For example,
  1102 Variable.import introduces fixes for schematics of given facts and
  1103 Variable.export reverses the effect (up to renaming) -- this replaces
  1104 various freeze_thaw operations.
  1105 
  1106 * Pure: structure Goal provides simple interfaces for
  1107 init/conclude/finish and tactical prove operations (replacing former
  1108 Tactic.prove).  Goal.prove is the canonical way to prove results
  1109 within a given context; Goal.prove_global is a degraded version for
  1110 theory level goals, including a global Drule.standard.  Note that
  1111 OldGoals.prove_goalw_cterm has long been obsolete, since it is
  1112 ill-behaved in a local proof context (e.g. with local fixes/assumes or
  1113 in a locale context).
  1114 
  1115 * Isar: simplified treatment of user-level errors, using exception
  1116 ERROR of string uniformly.  Function error now merely raises ERROR,
  1117 without any side effect on output channels.  The Isar toplevel takes
  1118 care of proper display of ERROR exceptions.  ML code may use plain
  1119 handle/can/try; cat_error may be used to concatenate errors like this:
  1120 
  1121   ... handle ERROR msg => cat_error msg "..."
  1122 
  1123 Toplevel ML code (run directly or through the Isar toplevel) may be
  1124 embedded into the Isar toplevel with exception display/debug like
  1125 this:
  1126 
  1127   Isar.toplevel (fn () => ...)
  1128 
  1129 INCOMPATIBILITY, removed special transform_error facilities, removed
  1130 obsolete variants of user-level exceptions (ERROR_MESSAGE,
  1131 Context.PROOF, ProofContext.CONTEXT, Proof.STATE, ProofHistory.FAIL)
  1132 -- use plain ERROR instead.
  1133 
  1134 * Isar: theory setup now has type (theory -> theory), instead of a
  1135 list.  INCOMPATIBILITY, may use #> to compose setup functions.
  1136 
  1137 * Isar: installed ML toplevel pretty printer for type Proof.context,
  1138 subject to ProofContext.debug/verbose flags.
  1139 
  1140 * Isar: Toplevel.theory_to_proof admits transactions that modify the
  1141 theory before entering a proof state.  Transactions now always see a
  1142 quasi-functional intermediate checkpoint, both in interactive and
  1143 batch mode.
  1144 
  1145 * Simplifier: the simpset of a running simplification process now
  1146 contains a proof context (cf. Simplifier.the_context), which is the
  1147 very context that the initial simpset has been retrieved from (by
  1148 simpset_of/local_simpset_of).  Consequently, all plug-in components
  1149 (solver, looper etc.) may depend on arbitrary proof data.
  1150 
  1151 * Simplifier.inherit_context inherits the proof context (plus the
  1152 local bounds) of the current simplification process; any simproc
  1153 etc. that calls the Simplifier recursively should do this!  Removed
  1154 former Simplifier.inherit_bounds, which is already included here --
  1155 INCOMPATIBILITY.  Tools based on low-level rewriting may even have to
  1156 specify an explicit context using Simplifier.context/theory_context.
  1157 
  1158 * Simplifier/Classical Reasoner: more abstract interfaces
  1159 change_simpset/claset for modifying the simpset/claset reference of a
  1160 theory; raw versions simpset/claset_ref etc. have been discontinued --
  1161 INCOMPATIBILITY.
  1162 
  1163 * Provers: more generic wrt. syntax of object-logics, avoid hardwired
  1164 "Trueprop" etc.
  1165 
  1166 
  1167 *** System ***
  1168 
  1169 * settings: ML_IDENTIFIER -- which is appended to user specific heap
  1170 locations -- now includes the Isabelle version identifier as well.
  1171 This simplifies use of multiple Isabelle installations.
  1172 
  1173 * isabelle-process: option -S (secure mode) disables some critical
  1174 operations, notably runtime compilation and evaluation of ML source
  1175 code.
  1176 
  1177 
  1178 New in Isabelle2005 (October 2005)
  1179 ----------------------------------
  1180 
  1181 *** General ***
  1182 
  1183 * Theory headers: the new header syntax for Isar theories is
  1184 
  1185   theory <name>
  1186   imports <theory1> ... <theoryN>
  1187   uses <file1> ... <fileM>
  1188   begin
  1189 
  1190 where the 'uses' part is optional.  The previous syntax
  1191 
  1192   theory <name> = <theory1> + ... + <theoryN>:
  1193 
  1194 will disappear in the next release.  Use isatool fixheaders to convert
  1195 existing theory files.  Note that there is no change in ancient
  1196 non-Isar theories now, but these will disappear soon.
  1197 
  1198 * Theory loader: parent theories can now also be referred to via
  1199 relative and absolute paths.
  1200 
  1201 * Command 'find_theorems' searches for a list of criteria instead of a
  1202 list of constants. Known criteria are: intro, elim, dest, name:string,
  1203 simp:term, and any term. Criteria can be preceded by '-' to select
  1204 theorems that do not match. Intro, elim, dest select theorems that
  1205 match the current goal, name:s selects theorems whose fully qualified
  1206 name contain s, and simp:term selects all simplification rules whose
  1207 lhs match term.  Any other term is interpreted as pattern and selects
  1208 all theorems matching the pattern. Available in ProofGeneral under
  1209 'ProofGeneral -> Find Theorems' or C-c C-f.  Example:
  1210 
  1211   C-c C-f (100) "(_::nat) + _ + _" intro -name: "HOL."
  1212 
  1213 prints the last 100 theorems matching the pattern "(_::nat) + _ + _",
  1214 matching the current goal as introduction rule and not having "HOL."
  1215 in their name (i.e. not being defined in theory HOL).
  1216 
  1217 * Command 'thms_containing' has been discontinued in favour of
  1218 'find_theorems'; INCOMPATIBILITY.
  1219 
  1220 * Communication with Proof General is now 8bit clean, which means that
  1221 Unicode text in UTF-8 encoding may be used within theory texts (both
  1222 formal and informal parts).  Cf. option -U of the Isabelle Proof
  1223 General interface.  Here are some simple examples (cf. src/HOL/ex):
  1224 
  1225   http://isabelle.in.tum.de/library/HOL/ex/Hebrew.html
  1226   http://isabelle.in.tum.de/library/HOL/ex/Chinese.html
  1227 
  1228 * Improved efficiency of the Simplifier and, to a lesser degree, the
  1229 Classical Reasoner.  Typical big applications run around 2 times
  1230 faster.
  1231 
  1232 
  1233 *** Document preparation ***
  1234 
  1235 * Commands 'display_drafts' and 'print_drafts' perform simple output
  1236 of raw sources.  Only those symbols that do not require additional
  1237 LaTeX packages (depending on comments in isabellesym.sty) are
  1238 displayed properly, everything else is left verbatim.  isatool display
  1239 and isatool print are used as front ends (these are subject to the
  1240 DVI/PDF_VIEWER and PRINT_COMMAND settings, respectively).
  1241 
  1242 * Command tags control specific markup of certain regions of text,
  1243 notably folding and hiding.  Predefined tags include "theory" (for
  1244 theory begin and end), "proof" for proof commands, and "ML" for
  1245 commands involving ML code; the additional tags "visible" and
  1246 "invisible" are unused by default.  Users may give explicit tag
  1247 specifications in the text, e.g. ''by %invisible (auto)''.  The
  1248 interpretation of tags is determined by the LaTeX job during document
  1249 preparation: see option -V of isatool usedir, or options -n and -t of
  1250 isatool document, or even the LaTeX macros \isakeeptag, \isafoldtag,
  1251 \isadroptag.
  1252 
  1253 Several document versions may be produced at the same time via isatool
  1254 usedir (the generated index.html will link all of them).  Typical
  1255 specifications include ''-V document=theory,proof,ML'' to present
  1256 theory/proof/ML parts faithfully, ''-V outline=/proof,/ML'' to fold
  1257 proof and ML commands, and ''-V mutilated=-theory,-proof,-ML'' to omit
  1258 these parts without any formal replacement text.  The Isabelle site
  1259 default settings produce ''document'' and ''outline'' versions as
  1260 specified above.
  1261 
  1262 * Several new antiquotations:
  1263 
  1264   @{term_type term} prints a term with its type annotated;
  1265 
  1266   @{typeof term} prints the type of a term;
  1267 
  1268   @{const const} is the same as @{term const}, but checks that the
  1269   argument is a known logical constant;
  1270 
  1271   @{term_style style term} and @{thm_style style thm} print a term or
  1272   theorem applying a "style" to it
  1273 
  1274   @{ML text}
  1275 
  1276 Predefined styles are 'lhs' and 'rhs' printing the lhs/rhs of
  1277 definitions, equations, inequations etc., 'concl' printing only the
  1278 conclusion of a meta-logical statement theorem, and 'prem1' .. 'prem19'
  1279 to print the specified premise.  TermStyle.add_style provides an ML
  1280 interface for introducing further styles.  See also the "LaTeX Sugar"
  1281 document practical applications.  The ML antiquotation prints
  1282 type-checked ML expressions verbatim.
  1283 
  1284 * Markup commands 'chapter', 'section', 'subsection', 'subsubsection',
  1285 and 'text' support optional locale specification '(in loc)', which
  1286 specifies the default context for interpreting antiquotations.  For
  1287 example: 'text (in lattice) {* @{thm inf_assoc}*}'.
  1288 
  1289 * Option 'locale=NAME' of antiquotations specifies an alternative
  1290 context interpreting the subsequent argument.  For example: @{thm
  1291 [locale=lattice] inf_assoc}.
  1292 
  1293 * Proper output of proof terms (@{prf ...} and @{full_prf ...}) within
  1294 a proof context.
  1295 
  1296 * Proper output of antiquotations for theory commands involving a
  1297 proof context (such as 'locale' or 'theorem (in loc) ...').
  1298 
  1299 * Delimiters of outer tokens (string etc.) now produce separate LaTeX
  1300 macros (\isachardoublequoteopen, isachardoublequoteclose etc.).
  1301 
  1302 * isatool usedir: new option -C (default true) controls whether option
  1303 -D should include a copy of the original document directory; -C false
  1304 prevents unwanted effects such as copying of administrative CVS data.
  1305 
  1306 
  1307 *** Pure ***
  1308 
  1309 * Considerably improved version of 'constdefs' command.  Now performs
  1310 automatic type-inference of declared constants; additional support for
  1311 local structure declarations (cf. locales and HOL records), see also
  1312 isar-ref manual.  Potential INCOMPATIBILITY: need to observe strictly
  1313 sequential dependencies of definitions within a single 'constdefs'
  1314 section; moreover, the declared name needs to be an identifier.  If
  1315 all fails, consider to fall back on 'consts' and 'defs' separately.
  1316 
  1317 * Improved indexed syntax and implicit structures.  First of all,
  1318 indexed syntax provides a notational device for subscripted
  1319 application, using the new syntax \<^bsub>term\<^esub> for arbitrary
  1320 expressions.  Secondly, in a local context with structure
  1321 declarations, number indexes \<^sub>n or the empty index (default
  1322 number 1) refer to a certain fixed variable implicitly; option
  1323 show_structs controls printing of implicit structures.  Typical
  1324 applications of these concepts involve record types and locales.
  1325 
  1326 * New command 'no_syntax' removes grammar declarations (and
  1327 translations) resulting from the given syntax specification, which is
  1328 interpreted in the same manner as for the 'syntax' command.
  1329 
  1330 * 'Advanced' translation functions (parse_translation etc.) may depend
  1331 on the signature of the theory context being presently used for
  1332 parsing/printing, see also isar-ref manual.
  1333 
  1334 * Improved 'oracle' command provides a type-safe interface to turn an
  1335 ML expression of type theory -> T -> term into a primitive rule of
  1336 type theory -> T -> thm (i.e. the functionality of Thm.invoke_oracle
  1337 is already included here); see also FOL/ex/IffExample.thy;
  1338 INCOMPATIBILITY.
  1339 
  1340 * axclass: name space prefix for class "c" is now "c_class" (was "c"
  1341 before); "cI" is no longer bound, use "c.intro" instead.
  1342 INCOMPATIBILITY.  This change avoids clashes of fact bindings for
  1343 axclasses vs. locales.
  1344 
  1345 * Improved internal renaming of symbolic identifiers -- attach primes
  1346 instead of base 26 numbers.
  1347 
  1348 * New flag show_question_marks controls printing of leading question
  1349 marks in schematic variable names.
  1350 
  1351 * In schematic variable names, *any* symbol following \<^isub> or
  1352 \<^isup> is now treated as part of the base name.  For example, the
  1353 following works without printing of awkward ".0" indexes:
  1354 
  1355   lemma "x\<^isub>1 = x\<^isub>2 ==> x\<^isub>2 = x\<^isub>1"
  1356     by simp
  1357 
  1358 * Inner syntax includes (*(*nested*) comments*).
  1359 
  1360 * Pretty printer now supports unbreakable blocks, specified in mixfix
  1361 annotations as "(00...)".
  1362 
  1363 * Clear separation of logical types and nonterminals, where the latter
  1364 may only occur in 'syntax' specifications or type abbreviations.
  1365 Before that distinction was only partially implemented via type class
  1366 "logic" vs. "{}".  Potential INCOMPATIBILITY in rare cases of improper
  1367 use of 'types'/'consts' instead of 'nonterminals'/'syntax'.  Some very
  1368 exotic syntax specifications may require further adaption
  1369 (e.g. Cube/Cube.thy).
  1370 
  1371 * Removed obsolete type class "logic", use the top sort {} instead.
  1372 Note that non-logical types should be declared as 'nonterminals'
  1373 rather than 'types'.  INCOMPATIBILITY for new object-logic
  1374 specifications.
  1375 
  1376 * Attributes 'induct' and 'cases': type or set names may now be
  1377 locally fixed variables as well.
  1378 
  1379 * Simplifier: can now control the depth to which conditional rewriting
  1380 is traced via the PG menu Isabelle -> Settings -> Trace Simp Depth
  1381 Limit.
  1382 
  1383 * Simplifier: simplification procedures may now take the current
  1384 simpset into account (cf. Simplifier.simproc(_i) / mk_simproc
  1385 interface), which is very useful for calling the Simplifier
  1386 recursively.  Minor INCOMPATIBILITY: the 'prems' argument of simprocs
  1387 is gone -- use prems_of_ss on the simpset instead.  Moreover, the
  1388 low-level mk_simproc no longer applies Logic.varify internally, to
  1389 allow for use in a context of fixed variables.
  1390 
  1391 * thin_tac now works even if the assumption being deleted contains !!
  1392 or ==>.  More generally, erule now works even if the major premise of
  1393 the elimination rule contains !! or ==>.
  1394 
  1395 * Method 'rules' has been renamed to 'iprover'. INCOMPATIBILITY.
  1396 
  1397 * Reorganized bootstrapping of the Pure theories; CPure is now derived
  1398 from Pure, which contains all common declarations already.  Both
  1399 theories are defined via plain Isabelle/Isar .thy files.
  1400 INCOMPATIBILITY: elements of CPure (such as the CPure.intro /
  1401 CPure.elim / CPure.dest attributes) now appear in the Pure name space;
  1402 use isatool fixcpure to adapt your theory and ML sources.
  1403 
  1404 * New syntax 'name(i-j, i-, i, ...)' for referring to specific
  1405 selections of theorems in named facts via index ranges.
  1406 
  1407 * 'print_theorems': in theory mode, really print the difference
  1408 wrt. the last state (works for interactive theory development only),
  1409 in proof mode print all local facts (cf. 'print_facts');
  1410 
  1411 * 'hide': option '(open)' hides only base names.
  1412 
  1413 * More efficient treatment of intermediate checkpoints in interactive
  1414 theory development.
  1415 
  1416 * Code generator is now invoked via code_module (incremental code
  1417 generation) and code_library (modular code generation, ML structures
  1418 for each theory).  INCOMPATIBILITY: new keywords 'file' and 'contains'
  1419 must be quoted when used as identifiers.
  1420 
  1421 * New 'value' command for reading, evaluating and printing terms using
  1422 the code generator.  INCOMPATIBILITY: command keyword 'value' must be
  1423 quoted when used as identifier.
  1424 
  1425 
  1426 *** Locales ***
  1427 
  1428 * New commands for the interpretation of locale expressions in
  1429 theories (1), locales (2) and proof contexts (3).  These generate
  1430 proof obligations from the expression specification.  After the
  1431 obligations have been discharged, theorems of the expression are added
  1432 to the theory, target locale or proof context.  The synopsis of the
  1433 commands is a follows:
  1434 
  1435   (1) interpretation expr inst
  1436   (2) interpretation target < expr
  1437   (3) interpret expr inst
  1438 
  1439 Interpretation in theories and proof contexts require a parameter
  1440 instantiation of terms from the current context.  This is applied to
  1441 specifications and theorems of the interpreted expression.
  1442 Interpretation in locales only permits parameter renaming through the
  1443 locale expression.  Interpretation is smart in that interpretations
  1444 that are active already do not occur in proof obligations, neither are
  1445 instantiated theorems stored in duplicate.  Use 'print_interps' to
  1446 inspect active interpretations of a particular locale.  For details,
  1447 see the Isar Reference manual.  Examples can be found in
  1448 HOL/Finite_Set.thy and HOL/Algebra/UnivPoly.thy.
  1449 
  1450 INCOMPATIBILITY: former 'instantiate' has been withdrawn, use
  1451 'interpret' instead.
  1452 
  1453 * New context element 'constrains' for adding type constraints to
  1454 parameters.
  1455 
  1456 * Context expressions: renaming of parameters with syntax
  1457 redeclaration.
  1458 
  1459 * Locale declaration: 'includes' disallowed.
  1460 
  1461 * Proper static binding of attribute syntax -- i.e. types / terms /
  1462 facts mentioned as arguments are always those of the locale definition
  1463 context, independently of the context of later invocations.  Moreover,
  1464 locale operations (renaming and type / term instantiation) are applied
  1465 to attribute arguments as expected.
  1466 
  1467 INCOMPATIBILITY of the ML interface: always pass Attrib.src instead of
  1468 actual attributes; rare situations may require Attrib.attribute to
  1469 embed those attributes into Attrib.src that lack concrete syntax.
  1470 Attribute implementations need to cooperate properly with the static
  1471 binding mechanism.  Basic parsers Args.XXX_typ/term/prop and
  1472 Attrib.XXX_thm etc. already do the right thing without further
  1473 intervention.  Only unusual applications -- such as "where" or "of"
  1474 (cf. src/Pure/Isar/attrib.ML), which process arguments depending both
  1475 on the context and the facts involved -- may have to assign parsed
  1476 values to argument tokens explicitly.
  1477 
  1478 * Changed parameter management in theorem generation for long goal
  1479 statements with 'includes'.  INCOMPATIBILITY: produces a different
  1480 theorem statement in rare situations.
  1481 
  1482 * Locale inspection command 'print_locale' omits notes elements.  Use
  1483 'print_locale!' to have them included in the output.
  1484 
  1485 
  1486 *** Provers ***
  1487 
  1488 * Provers/hypsubst.ML: improved version of the subst method, for
  1489 single-step rewriting: it now works in bound variable contexts. New is
  1490 'subst (asm)', for rewriting an assumption.  INCOMPATIBILITY: may
  1491 rewrite a different subterm than the original subst method, which is
  1492 still available as 'simplesubst'.
  1493 
  1494 * Provers/quasi.ML: new transitivity reasoners for transitivity only
  1495 and quasi orders.
  1496 
  1497 * Provers/trancl.ML: new transitivity reasoner for transitive and
  1498 reflexive-transitive closure of relations.
  1499 
  1500 * Provers/blast.ML: new reference depth_limit to make blast's depth
  1501 limit (previously hard-coded with a value of 20) user-definable.
  1502 
  1503 * Provers/simplifier.ML has been moved to Pure, where Simplifier.setup
  1504 is peformed already.  Object-logics merely need to finish their
  1505 initial simpset configuration as before.  INCOMPATIBILITY.
  1506 
  1507 
  1508 *** HOL ***
  1509 
  1510 * Symbolic syntax of Hilbert Choice Operator is now as follows:
  1511 
  1512   syntax (epsilon)
  1513     "_Eps" :: "[pttrn, bool] => 'a"    ("(3\<some>_./ _)" [0, 10] 10)
  1514 
  1515 The symbol \<some> is displayed as the alternative epsilon of LaTeX
  1516 and x-symbol; use option '-m epsilon' to get it actually printed.
  1517 Moreover, the mathematically important symbolic identifier \<epsilon>
  1518 becomes available as variable, constant etc.  INCOMPATIBILITY,
  1519 
  1520 * "x > y" abbreviates "y < x" and "x >= y" abbreviates "y <= x".
  1521 Similarly for all quantifiers: "ALL x > y" etc.  The x-symbol for >=
  1522 is \<ge>. New transitivity rules have been added to HOL/Orderings.thy to
  1523 support corresponding Isar calculations.
  1524 
  1525 * "{x:A. P}" abbreviates "{x. x:A & P}", and similarly for "\<in>"
  1526 instead of ":".
  1527 
  1528 * theory SetInterval: changed the syntax for open intervals:
  1529 
  1530   Old       New
  1531   {..n(}    {..<n}
  1532   {)n..}    {n<..}
  1533   {m..n(}   {m..<n}
  1534   {)m..n}   {m<..n}
  1535   {)m..n(}  {m<..<n}
  1536 
  1537 The old syntax is still supported but will disappear in the next
  1538 release.  For conversion use the following Emacs search and replace
  1539 patterns (these are not perfect but work quite well):
  1540 
  1541   {)\([^\.]*\)\.\.  ->  {\1<\.\.}
  1542   \.\.\([^(}]*\)(}  ->  \.\.<\1}
  1543 
  1544 * Theory Commutative_Ring (in Library): method comm_ring for proving
  1545 equalities in commutative rings; method 'algebra' provides a generic
  1546 interface.
  1547 
  1548 * Theory Finite_Set: changed the syntax for 'setsum', summation over
  1549 finite sets: "setsum (%x. e) A", which used to be "\<Sum>x:A. e", is
  1550 now either "SUM x:A. e" or "\<Sum>x \<in> A. e". The bound variable can
  1551 be a tuple pattern.
  1552 
  1553 Some new syntax forms are available:
  1554 
  1555   "\<Sum>x | P. e"      for     "setsum (%x. e) {x. P}"
  1556   "\<Sum>x = a..b. e"   for     "setsum (%x. e) {a..b}"
  1557   "\<Sum>x = a..<b. e"  for     "setsum (%x. e) {a..<b}"
  1558   "\<Sum>x < k. e"      for     "setsum (%x. e) {..<k}"
  1559 
  1560 The latter form "\<Sum>x < k. e" used to be based on a separate
  1561 function "Summation", which has been discontinued.
  1562 
  1563 * theory Finite_Set: in structured induction proofs, the insert case
  1564 is now 'case (insert x F)' instead of the old counterintuitive 'case
  1565 (insert F x)'.
  1566 
  1567 * The 'refute' command has been extended to support a much larger
  1568 fragment of HOL, including axiomatic type classes, constdefs and
  1569 typedefs, inductive datatypes and recursion.
  1570 
  1571 * New tactics 'sat' and 'satx' to prove propositional tautologies.
  1572 Requires zChaff with proof generation to be installed.  See
  1573 HOL/ex/SAT_Examples.thy for examples.
  1574 
  1575 * Datatype induction via method 'induct' now preserves the name of the
  1576 induction variable. For example, when proving P(xs::'a list) by
  1577 induction on xs, the induction step is now P(xs) ==> P(a#xs) rather
  1578 than P(list) ==> P(a#list) as previously.  Potential INCOMPATIBILITY
  1579 in unstructured proof scripts.
  1580 
  1581 * Reworked implementation of records.  Improved scalability for
  1582 records with many fields, avoiding performance problems for type
  1583 inference. Records are no longer composed of nested field types, but
  1584 of nested extension types. Therefore the record type only grows linear
  1585 in the number of extensions and not in the number of fields.  The
  1586 top-level (users) view on records is preserved.  Potential
  1587 INCOMPATIBILITY only in strange cases, where the theory depends on the
  1588 old record representation. The type generated for a record is called
  1589 <record_name>_ext_type.
  1590 
  1591 Flag record_quick_and_dirty_sensitive can be enabled to skip the
  1592 proofs triggered by a record definition or a simproc (if
  1593 quick_and_dirty is enabled).  Definitions of large records can take
  1594 quite long.
  1595 
  1596 New simproc record_upd_simproc for simplification of multiple record
  1597 updates enabled by default.  Moreover, trivial updates are also
  1598 removed: r(|x := x r|) = r.  INCOMPATIBILITY: old proofs break
  1599 occasionally, since simplification is more powerful by default.
  1600 
  1601 * typedef: proper support for polymorphic sets, which contain extra
  1602 type-variables in the term.
  1603 
  1604 * Simplifier: automatically reasons about transitivity chains
  1605 involving "trancl" (r^+) and "rtrancl" (r^*) by setting up tactics
  1606 provided by Provers/trancl.ML as additional solvers.  INCOMPATIBILITY:
  1607 old proofs break occasionally as simplification may now solve more
  1608 goals than previously.
  1609 
  1610 * Simplifier: converts x <= y into x = y if assumption y <= x is
  1611 present.  Works for all partial orders (class "order"), in particular
  1612 numbers and sets.  For linear orders (e.g. numbers) it treats ~ x < y
  1613 just like y <= x.
  1614 
  1615 * Simplifier: new simproc for "let x = a in f x".  If a is a free or
  1616 bound variable or a constant then the let is unfolded.  Otherwise
  1617 first a is simplified to b, and then f b is simplified to g. If
  1618 possible we abstract b from g arriving at "let x = b in h x",
  1619 otherwise we unfold the let and arrive at g.  The simproc can be
  1620 enabled/disabled by the reference use_let_simproc.  Potential
  1621 INCOMPATIBILITY since simplification is more powerful by default.
  1622 
  1623 * Classical reasoning: the meson method now accepts theorems as arguments.
  1624 
  1625 * Prover support: pre-release of the Isabelle-ATP linkup, which runs background
  1626 jobs to provide advice on the provability of subgoals.
  1627 
  1628 * Theory OrderedGroup and Ring_and_Field: various additions and
  1629 improvements to faciliate calculations involving equalities and
  1630 inequalities.
  1631 
  1632 The following theorems have been eliminated or modified
  1633 (INCOMPATIBILITY):
  1634 
  1635   abs_eq             now named abs_of_nonneg
  1636   abs_of_ge_0        now named abs_of_nonneg
  1637   abs_minus_eq       now named abs_of_nonpos
  1638   imp_abs_id         now named abs_of_nonneg
  1639   imp_abs_neg_id     now named abs_of_nonpos
  1640   mult_pos           now named mult_pos_pos
  1641   mult_pos_le        now named mult_nonneg_nonneg
  1642   mult_pos_neg_le    now named mult_nonneg_nonpos
  1643   mult_pos_neg2_le   now named mult_nonneg_nonpos2
  1644   mult_neg           now named mult_neg_neg
  1645   mult_neg_le        now named mult_nonpos_nonpos
  1646 
  1647 * Theory Parity: added rules for simplifying exponents.
  1648 
  1649 * Theory List:
  1650 
  1651 The following theorems have been eliminated or modified
  1652 (INCOMPATIBILITY):
  1653 
  1654   list_all_Nil       now named list_all.simps(1)
  1655   list_all_Cons      now named list_all.simps(2)
  1656   list_all_conv      now named list_all_iff
  1657   set_mem_eq         now named mem_iff
  1658 
  1659 * Theories SetsAndFunctions and BigO (see HOL/Library) support
  1660 asymptotic "big O" calculations.  See the notes in BigO.thy.
  1661 
  1662 
  1663 *** HOL-Complex ***
  1664 
  1665 * Theory RealDef: better support for embedding natural numbers and
  1666 integers in the reals.
  1667 
  1668 The following theorems have been eliminated or modified
  1669 (INCOMPATIBILITY):
  1670 
  1671   exp_ge_add_one_self  now requires no hypotheses
  1672   real_of_int_add      reversed direction of equality (use [symmetric])
  1673   real_of_int_minus    reversed direction of equality (use [symmetric])
  1674   real_of_int_diff     reversed direction of equality (use [symmetric])
  1675   real_of_int_mult     reversed direction of equality (use [symmetric])
  1676 
  1677 * Theory RComplete: expanded support for floor and ceiling functions.
  1678 
  1679 * Theory Ln is new, with properties of the natural logarithm
  1680 
  1681 * Hyperreal: There is a new type constructor "star" for making
  1682 nonstandard types.  The old type names are now type synonyms:
  1683 
  1684   hypreal = real star
  1685   hypnat = nat star
  1686   hcomplex = complex star
  1687 
  1688 * Hyperreal: Many groups of similarly-defined constants have been
  1689 replaced by polymorphic versions (INCOMPATIBILITY):
  1690 
  1691   star_of <-- hypreal_of_real, hypnat_of_nat, hcomplex_of_complex
  1692 
  1693   starset      <-- starsetNat, starsetC
  1694   *s*          <-- *sNat*, *sc*
  1695   starset_n    <-- starsetNat_n, starsetC_n
  1696   *sn*         <-- *sNatn*, *scn*
  1697   InternalSets <-- InternalNatSets, InternalCSets
  1698 
  1699   starfun      <-- starfun{Nat,Nat2,C,RC,CR}
  1700   *f*          <-- *fNat*, *fNat2*, *fc*, *fRc*, *fcR*
  1701   starfun_n    <-- starfun{Nat,Nat2,C,RC,CR}_n
  1702   *fn*         <-- *fNatn*, *fNat2n*, *fcn*, *fRcn*, *fcRn*
  1703   InternalFuns <-- InternalNatFuns, InternalNatFuns2, Internal{C,RC,CR}Funs
  1704 
  1705 * Hyperreal: Many type-specific theorems have been removed in favor of
  1706 theorems specific to various axiomatic type classes (INCOMPATIBILITY):
  1707 
  1708   add_commute <-- {hypreal,hypnat,hcomplex}_add_commute
  1709   add_assoc   <-- {hypreal,hypnat,hcomplex}_add_assocs
  1710   OrderedGroup.add_0 <-- {hypreal,hypnat,hcomplex}_add_zero_left
  1711   OrderedGroup.add_0_right <-- {hypreal,hcomplex}_add_zero_right
  1712   right_minus <-- hypreal_add_minus
  1713   left_minus <-- {hypreal,hcomplex}_add_minus_left
  1714   mult_commute <-- {hypreal,hypnat,hcomplex}_mult_commute
  1715   mult_assoc <-- {hypreal,hypnat,hcomplex}_mult_assoc
  1716   mult_1_left <-- {hypreal,hypnat}_mult_1, hcomplex_mult_one_left
  1717   mult_1_right <-- hcomplex_mult_one_right
  1718   mult_zero_left <-- hcomplex_mult_zero_left
  1719   left_distrib <-- {hypreal,hypnat,hcomplex}_add_mult_distrib
  1720   right_distrib <-- hypnat_add_mult_distrib2
  1721   zero_neq_one <-- {hypreal,hypnat,hcomplex}_zero_not_eq_one
  1722   right_inverse <-- hypreal_mult_inverse
  1723   left_inverse <-- hypreal_mult_inverse_left, hcomplex_mult_inv_left
  1724   order_refl <-- {hypreal,hypnat}_le_refl
  1725   order_trans <-- {hypreal,hypnat}_le_trans
  1726   order_antisym <-- {hypreal,hypnat}_le_anti_sym
  1727   order_less_le <-- {hypreal,hypnat}_less_le
  1728   linorder_linear <-- {hypreal,hypnat}_le_linear
  1729   add_left_mono <-- {hypreal,hypnat}_add_left_mono
  1730   mult_strict_left_mono <-- {hypreal,hypnat}_mult_less_mono2
  1731   add_nonneg_nonneg <-- hypreal_le_add_order
  1732 
  1733 * Hyperreal: Separate theorems having to do with type-specific
  1734 versions of constants have been merged into theorems that apply to the
  1735 new polymorphic constants (INCOMPATIBILITY):
  1736 
  1737   STAR_UNIV_set <-- {STAR_real,NatStar_real,STARC_complex}_set
  1738   STAR_empty_set <-- {STAR,NatStar,STARC}_empty_set
  1739   STAR_Un <-- {STAR,NatStar,STARC}_Un
  1740   STAR_Int <-- {STAR,NatStar,STARC}_Int
  1741   STAR_Compl <-- {STAR,NatStar,STARC}_Compl
  1742   STAR_subset <-- {STAR,NatStar,STARC}_subset
  1743   STAR_mem <-- {STAR,NatStar,STARC}_mem
  1744   STAR_mem_Compl <-- {STAR,STARC}_mem_Compl
  1745   STAR_diff <-- {STAR,STARC}_diff
  1746   STAR_star_of_image_subset <-- {STAR_hypreal_of_real, NatStar_hypreal_of_real,
  1747     STARC_hcomplex_of_complex}_image_subset
  1748   starset_n_Un <-- starset{Nat,C}_n_Un
  1749   starset_n_Int <-- starset{Nat,C}_n_Int
  1750   starset_n_Compl <-- starset{Nat,C}_n_Compl
  1751   starset_n_diff <-- starset{Nat,C}_n_diff
  1752   InternalSets_Un <-- Internal{Nat,C}Sets_Un
  1753   InternalSets_Int <-- Internal{Nat,C}Sets_Int
  1754   InternalSets_Compl <-- Internal{Nat,C}Sets_Compl
  1755   InternalSets_diff <-- Internal{Nat,C}Sets_diff
  1756   InternalSets_UNIV_diff <-- Internal{Nat,C}Sets_UNIV_diff
  1757   InternalSets_starset_n <-- Internal{Nat,C}Sets_starset{Nat,C}_n
  1758   starset_starset_n_eq <-- starset{Nat,C}_starset{Nat,C}_n_eq
  1759   starset_n_starset <-- starset{Nat,C}_n_starset{Nat,C}
  1760   starfun_n_starfun <-- starfun{Nat,Nat2,C,RC,CR}_n_starfun{Nat,Nat2,C,RC,CR}
  1761   starfun <-- starfun{Nat,Nat2,C,RC,CR}
  1762   starfun_mult <-- starfun{Nat,Nat2,C,RC,CR}_mult
  1763   starfun_add <-- starfun{Nat,Nat2,C,RC,CR}_add
  1764   starfun_minus <-- starfun{Nat,Nat2,C,RC,CR}_minus
  1765   starfun_diff <-- starfun{C,RC,CR}_diff
  1766   starfun_o <-- starfun{NatNat2,Nat2,_stafunNat,C,C_starfunRC,_starfunCR}_o
  1767   starfun_o2 <-- starfun{NatNat2,_stafunNat,C,C_starfunRC,_starfunCR}_o2
  1768   starfun_const_fun <-- starfun{Nat,Nat2,C,RC,CR}_const_fun
  1769   starfun_inverse <-- starfun{Nat,C,RC,CR}_inverse
  1770   starfun_eq <-- starfun{Nat,Nat2,C,RC,CR}_eq
  1771   starfun_eq_iff <-- starfun{C,RC,CR}_eq_iff
  1772   starfun_Id <-- starfunC_Id
  1773   starfun_approx <-- starfun{Nat,CR}_approx
  1774   starfun_capprox <-- starfun{C,RC}_capprox
  1775   starfun_abs <-- starfunNat_rabs
  1776   starfun_lambda_cancel <-- starfun{C,CR,RC}_lambda_cancel
  1777   starfun_lambda_cancel2 <-- starfun{C,CR,RC}_lambda_cancel2
  1778   starfun_mult_HFinite_approx <-- starfunCR_mult_HFinite_capprox
  1779   starfun_mult_CFinite_capprox <-- starfun{C,RC}_mult_CFinite_capprox
  1780   starfun_add_capprox <-- starfun{C,RC}_add_capprox
  1781   starfun_add_approx <-- starfunCR_add_approx
  1782   starfun_inverse_inverse <-- starfunC_inverse_inverse
  1783   starfun_divide <-- starfun{C,CR,RC}_divide
  1784   starfun_n <-- starfun{Nat,C}_n
  1785   starfun_n_mult <-- starfun{Nat,C}_n_mult
  1786   starfun_n_add <-- starfun{Nat,C}_n_add
  1787   starfun_n_add_minus <-- starfunNat_n_add_minus
  1788   starfun_n_const_fun <-- starfun{Nat,C}_n_const_fun
  1789   starfun_n_minus <-- starfun{Nat,C}_n_minus
  1790   starfun_n_eq <-- starfun{Nat,C}_n_eq
  1791 
  1792   star_n_add <-- {hypreal,hypnat,hcomplex}_add
  1793   star_n_minus <-- {hypreal,hcomplex}_minus
  1794   star_n_diff <-- {hypreal,hcomplex}_diff
  1795   star_n_mult <-- {hypreal,hcomplex}_mult
  1796   star_n_inverse <-- {hypreal,hcomplex}_inverse
  1797   star_n_le <-- {hypreal,hypnat}_le
  1798   star_n_less <-- {hypreal,hypnat}_less
  1799   star_n_zero_num <-- {hypreal,hypnat,hcomplex}_zero_num
  1800   star_n_one_num <-- {hypreal,hypnat,hcomplex}_one_num
  1801   star_n_abs <-- hypreal_hrabs
  1802   star_n_divide <-- hcomplex_divide
  1803 
  1804   star_of_add <-- {hypreal_of_real,hypnat_of_nat,hcomplex_of_complex}_add
  1805   star_of_minus <-- {hypreal_of_real,hcomplex_of_complex}_minus
  1806   star_of_diff <-- hypreal_of_real_diff
  1807   star_of_mult <-- {hypreal_of_real,hypnat_of_nat,hcomplex_of_complex}_mult
  1808   star_of_one <-- {hypreal_of_real,hcomplex_of_complex}_one
  1809   star_of_zero <-- {hypreal_of_real,hypnat_of_nat,hcomplex_of_complex}_zero
  1810   star_of_le <-- {hypreal_of_real,hypnat_of_nat}_le_iff
  1811   star_of_less <-- {hypreal_of_real,hypnat_of_nat}_less_iff
  1812   star_of_eq <-- {hypreal_of_real,hypnat_of_nat,hcomplex_of_complex}_eq_iff
  1813   star_of_inverse <-- {hypreal_of_real,hcomplex_of_complex}_inverse
  1814   star_of_divide <-- {hypreal_of_real,hcomplex_of_complex}_divide
  1815   star_of_of_nat <-- {hypreal_of_real,hcomplex_of_complex}_of_nat
  1816   star_of_of_int <-- {hypreal_of_real,hcomplex_of_complex}_of_int
  1817   star_of_number_of <-- {hypreal,hcomplex}_number_of
  1818   star_of_number_less <-- number_of_less_hypreal_of_real_iff
  1819   star_of_number_le <-- number_of_le_hypreal_of_real_iff
  1820   star_of_eq_number <-- hypreal_of_real_eq_number_of_iff
  1821   star_of_less_number <-- hypreal_of_real_less_number_of_iff
  1822   star_of_le_number <-- hypreal_of_real_le_number_of_iff
  1823   star_of_power <-- hypreal_of_real_power
  1824   star_of_eq_0 <-- hcomplex_of_complex_zero_iff
  1825 
  1826 * Hyperreal: new method "transfer" that implements the transfer
  1827 principle of nonstandard analysis. With a subgoal that mentions
  1828 nonstandard types like "'a star", the command "apply transfer"
  1829 replaces it with an equivalent one that mentions only standard types.
  1830 To be successful, all free variables must have standard types; non-
  1831 standard variables must have explicit universal quantifiers.
  1832 
  1833 * Hyperreal: A theory of Taylor series.
  1834 
  1835 
  1836 *** HOLCF ***
  1837 
  1838 * Discontinued special version of 'constdefs' (which used to support
  1839 continuous functions) in favor of the general Pure one with full
  1840 type-inference.
  1841 
  1842 * New simplification procedure for solving continuity conditions; it
  1843 is much faster on terms with many nested lambda abstractions (cubic
  1844 instead of exponential time).
  1845 
  1846 * New syntax for domain package: selector names are now optional.
  1847 Parentheses should be omitted unless argument is lazy, for example:
  1848 
  1849   domain 'a stream = cons "'a" (lazy "'a stream")
  1850 
  1851 * New command 'fixrec' for defining recursive functions with pattern
  1852 matching; defining multiple functions with mutual recursion is also
  1853 supported.  Patterns may include the constants cpair, spair, up, sinl,
  1854 sinr, or any data constructor defined by the domain package. The given
  1855 equations are proven as rewrite rules. See HOLCF/ex/Fixrec_ex.thy for
  1856 syntax and examples.
  1857 
  1858 * New commands 'cpodef' and 'pcpodef' for defining predicate subtypes
  1859 of cpo and pcpo types. Syntax is exactly like the 'typedef' command,
  1860 but the proof obligation additionally includes an admissibility
  1861 requirement. The packages generate instances of class cpo or pcpo,
  1862 with continuity and strictness theorems for Rep and Abs.
  1863 
  1864 * HOLCF: Many theorems have been renamed according to a more standard naming
  1865 scheme (INCOMPATIBILITY):
  1866 
  1867   foo_inject:  "foo$x = foo$y ==> x = y"
  1868   foo_eq:      "(foo$x = foo$y) = (x = y)"
  1869   foo_less:    "(foo$x << foo$y) = (x << y)"
  1870   foo_strict:  "foo$UU = UU"
  1871   foo_defined: "... ==> foo$x ~= UU"
  1872   foo_defined_iff: "(foo$x = UU) = (x = UU)"
  1873 
  1874 
  1875 *** ZF ***
  1876 
  1877 * ZF/ex: theories Group and Ring provide examples in abstract algebra,
  1878 including the First Isomorphism Theorem (on quotienting by the kernel
  1879 of a homomorphism).
  1880 
  1881 * ZF/Simplifier: install second copy of type solver that actually
  1882 makes use of TC rules declared to Isar proof contexts (or locales);
  1883 the old version is still required for ML proof scripts.
  1884 
  1885 
  1886 *** Cube ***
  1887 
  1888 * Converted to Isar theory format; use locales instead of axiomatic
  1889 theories.
  1890 
  1891 
  1892 *** ML ***
  1893 
  1894 * Pure/library.ML: added ##>, ##>>, #>> -- higher-order counterparts
  1895 for ||>, ||>>, |>>,
  1896 
  1897 * Pure/library.ML no longer defines its own option datatype, but uses
  1898 that of the SML basis, which has constructors NONE and SOME instead of
  1899 None and Some, as well as exception Option.Option instead of OPTION.
  1900 The functions the, if_none, is_some, is_none have been adapted
  1901 accordingly, while Option.map replaces apsome.
  1902 
  1903 * Pure/library.ML: the exception LIST has been given up in favour of
  1904 the standard exceptions Empty and Subscript, as well as
  1905 Library.UnequalLengths.  Function like Library.hd and Library.tl are
  1906 superceded by the standard hd and tl functions etc.
  1907 
  1908 A number of basic list functions are no longer exported to the ML
  1909 toplevel, as they are variants of predefined functions.  The following
  1910 suggests how one can translate existing code:
  1911 
  1912     rev_append xs ys = List.revAppend (xs, ys)
  1913     nth_elem (i, xs) = List.nth (xs, i)
  1914     last_elem xs = List.last xs
  1915     flat xss = List.concat xss
  1916     seq fs = List.app fs
  1917     partition P xs = List.partition P xs
  1918     mapfilter f xs = List.mapPartial f xs
  1919 
  1920 * Pure/library.ML: several combinators for linear functional
  1921 transformations, notably reverse application and composition:
  1922 
  1923   x |> f                f #> g
  1924   (x, y) |-> f          f #-> g
  1925 
  1926 * Pure/library.ML: introduced/changed precedence of infix operators:
  1927 
  1928   infix 1 |> |-> ||> ||>> |>> |>>> #> #->;
  1929   infix 2 ?;
  1930   infix 3 o oo ooo oooo;
  1931   infix 4 ~~ upto downto;
  1932 
  1933 Maybe INCOMPATIBILITY when any of those is used in conjunction with other
  1934 infix operators.
  1935 
  1936 * Pure/library.ML: natural list combinators fold, fold_rev, and
  1937 fold_map support linear functional transformations and nesting.  For
  1938 example:
  1939 
  1940   fold f [x1, ..., xN] y =
  1941     y |> f x1 |> ... |> f xN
  1942 
  1943   (fold o fold) f [xs1, ..., xsN] y =
  1944     y |> fold f xs1 |> ... |> fold f xsN
  1945 
  1946   fold f [x1, ..., xN] =
  1947     f x1 #> ... #> f xN
  1948 
  1949   (fold o fold) f [xs1, ..., xsN] =
  1950     fold f xs1 #> ... #> fold f xsN
  1951 
  1952 * Pure/library.ML: the following selectors on type 'a option are
  1953 available:
  1954 
  1955   the:               'a option -> 'a  (*partial*)
  1956   these:             'a option -> 'a  where 'a = 'b list
  1957   the_default: 'a -> 'a option -> 'a
  1958   the_list:          'a option -> 'a list
  1959 
  1960 * Pure/General: structure AList (cf. Pure/General/alist.ML) provides
  1961 basic operations for association lists, following natural argument
  1962 order; moreover the explicit equality predicate passed here avoids
  1963 potentially expensive polymorphic runtime equality checks.
  1964 The old functions may be expressed as follows:
  1965 
  1966   assoc = uncurry (AList.lookup (op =))
  1967   assocs = these oo AList.lookup (op =)
  1968   overwrite = uncurry (AList.update (op =)) o swap
  1969 
  1970 * Pure/General: structure AList (cf. Pure/General/alist.ML) provides
  1971 
  1972   val make: ('a -> 'b) -> 'a list -> ('a * 'b) list
  1973   val find: ('a * 'b -> bool) -> ('c * 'b) list -> 'a -> 'c list
  1974 
  1975 replacing make_keylist and keyfilter (occassionally used)
  1976 Naive rewrites:
  1977 
  1978   make_keylist = AList.make
  1979   keyfilter = AList.find (op =)
  1980 
  1981 * eq_fst and eq_snd now take explicit equality parameter, thus
  1982   avoiding eqtypes. Naive rewrites:
  1983 
  1984     eq_fst = eq_fst (op =)
  1985     eq_snd = eq_snd (op =)
  1986 
  1987 * Removed deprecated apl and apr (rarely used).
  1988   Naive rewrites:
  1989 
  1990     apl (n, op) =>>= curry op n
  1991     apr (op, m) =>>= fn n => op (n, m)
  1992 
  1993 * Pure/General: structure OrdList (cf. Pure/General/ord_list.ML)
  1994 provides a reasonably efficient light-weight implementation of sets as
  1995 lists.
  1996 
  1997 * Pure/General: generic tables (cf. Pure/General/table.ML) provide a
  1998 few new operations; existing lookup and update are now curried to
  1999 follow natural argument order (for use with fold etc.);
  2000 INCOMPATIBILITY, use (uncurry Symtab.lookup) etc. as last resort.
  2001 
  2002 * Pure/General: output via the Isabelle channels of
  2003 writeln/warning/error etc. is now passed through Output.output, with a
  2004 hook for arbitrary transformations depending on the print_mode
  2005 (cf. Output.add_mode -- the first active mode that provides a output
  2006 function wins).  Already formatted output may be embedded into further
  2007 text via Output.raw; the result of Pretty.string_of/str_of and derived
  2008 functions (string_of_term/cterm/thm etc.) is already marked raw to
  2009 accommodate easy composition of diagnostic messages etc.  Programmers
  2010 rarely need to care about Output.output or Output.raw at all, with
  2011 some notable exceptions: Output.output is required when bypassing the
  2012 standard channels (writeln etc.), or in token translations to produce
  2013 properly formatted results; Output.raw is required when capturing
  2014 already output material that will eventually be presented to the user
  2015 a second time.  For the default print mode, both Output.output and
  2016 Output.raw have no effect.
  2017 
  2018 * Pure/General: Output.time_accumulator NAME creates an operator ('a
  2019 -> 'b) -> 'a -> 'b to measure runtime and count invocations; the
  2020 cumulative results are displayed at the end of a batch session.
  2021 
  2022 * Pure/General: File.sysify_path and File.quote_sysify path have been
  2023 replaced by File.platform_path and File.shell_path (with appropriate
  2024 hooks).  This provides a clean interface for unusual systems where the
  2025 internal and external process view of file names are different.
  2026 
  2027 * Pure: more efficient orders for basic syntactic entities: added
  2028 fast_string_ord, fast_indexname_ord, fast_term_ord; changed sort_ord
  2029 and typ_ord to use fast_string_ord and fast_indexname_ord (term_ord is
  2030 NOT affected); structures Symtab, Vartab, Typtab, Termtab use the fast
  2031 orders now -- potential INCOMPATIBILITY for code that depends on a
  2032 particular order for Symtab.keys, Symtab.dest, etc. (consider using
  2033 Library.sort_strings on result).
  2034 
  2035 * Pure/term.ML: combinators fold_atyps, fold_aterms, fold_term_types,
  2036 fold_types traverse types/terms from left to right, observing natural
  2037 argument order.  Supercedes previous foldl_XXX versions, add_frees,
  2038 add_vars etc. have been adapted as well: INCOMPATIBILITY.
  2039 
  2040 * Pure: name spaces have been refined, with significant changes of the
  2041 internal interfaces -- INCOMPATIBILITY.  Renamed cond_extern(_table)
  2042 to extern(_table).  The plain name entry path is superceded by a
  2043 general 'naming' context, which also includes the 'policy' to produce
  2044 a fully qualified name and external accesses of a fully qualified
  2045 name; NameSpace.extend is superceded by context dependent
  2046 Sign.declare_name.  Several theory and proof context operations modify
  2047 the naming context.  Especially note Theory.restore_naming and
  2048 ProofContext.restore_naming to get back to a sane state; note that
  2049 Theory.add_path is no longer sufficient to recover from
  2050 Theory.absolute_path in particular.
  2051 
  2052 * Pure: new flags short_names (default false) and unique_names
  2053 (default true) for controlling output of qualified names.  If
  2054 short_names is set, names are printed unqualified.  If unique_names is
  2055 reset, the name prefix is reduced to the minimum required to achieve
  2056 the original result when interning again, even if there is an overlap
  2057 with earlier declarations.
  2058 
  2059 * Pure/TheoryDataFun: change of the argument structure; 'prep_ext' is
  2060 now 'extend', and 'merge' gets an additional Pretty.pp argument
  2061 (useful for printing error messages).  INCOMPATIBILITY.
  2062 
  2063 * Pure: major reorganization of the theory context.  Type Sign.sg and
  2064 Theory.theory are now identified, referring to the universal
  2065 Context.theory (see Pure/context.ML).  Actual signature and theory
  2066 content is managed as theory data.  The old code and interfaces were
  2067 spread over many files and structures; the new arrangement introduces
  2068 considerable INCOMPATIBILITY to gain more clarity:
  2069 
  2070   Context -- theory management operations (name, identity, inclusion,
  2071     parents, ancestors, merge, etc.), plus generic theory data;
  2072 
  2073   Sign -- logical signature and syntax operations (declaring consts,
  2074     types, etc.), plus certify/read for common entities;
  2075 
  2076   Theory -- logical theory operations (stating axioms, definitions,
  2077     oracles), plus a copy of logical signature operations (consts,
  2078     types, etc.); also a few basic management operations (Theory.copy,
  2079     Theory.merge, etc.)
  2080 
  2081 The most basic sign_of operations (Theory.sign_of, Thm.sign_of_thm
  2082 etc.) as well as the sign field in Thm.rep_thm etc. have been retained
  2083 for convenience -- they merely return the theory.
  2084 
  2085 * Pure: type Type.tsig is superceded by theory in most interfaces.
  2086 
  2087 * Pure: the Isar proof context type is already defined early in Pure
  2088 as Context.proof (note that ProofContext.context and Proof.context are
  2089 aliases, where the latter is the preferred name).  This enables other
  2090 Isabelle components to refer to that type even before Isar is present.
  2091 
  2092 * Pure/sign/theory: discontinued named name spaces (i.e. classK,
  2093 typeK, constK, axiomK, oracleK), but provide explicit operations for
  2094 any of these kinds.  For example, Sign.intern typeK is now
  2095 Sign.intern_type, Theory.hide_space Sign.typeK is now
  2096 Theory.hide_types.  Also note that former
  2097 Theory.hide_classes/types/consts are now
  2098 Theory.hide_classes_i/types_i/consts_i, while the non '_i' versions
  2099 internalize their arguments!  INCOMPATIBILITY.
  2100 
  2101 * Pure: get_thm interface (of PureThy and ProofContext) expects
  2102 datatype thmref (with constructors Name and NameSelection) instead of
  2103 plain string -- INCOMPATIBILITY;
  2104 
  2105 * Pure: cases produced by proof methods specify options, where NONE
  2106 means to remove case bindings -- INCOMPATIBILITY in
  2107 (RAW_)METHOD_CASES.
  2108 
  2109 * Pure: the following operations retrieve axioms or theorems from a
  2110 theory node or theory hierarchy, respectively:
  2111 
  2112   Theory.axioms_of: theory -> (string * term) list
  2113   Theory.all_axioms_of: theory -> (string * term) list
  2114   PureThy.thms_of: theory -> (string * thm) list
  2115   PureThy.all_thms_of: theory -> (string * thm) list
  2116 
  2117 * Pure: print_tac now outputs the goal through the trace channel.
  2118 
  2119 * Isar toplevel: improved diagnostics, mostly for Poly/ML only.
  2120 Reference Toplevel.debug (default false) controls detailed printing
  2121 and tracing of low-level exceptions; Toplevel.profiling (default 0)
  2122 controls execution profiling -- set to 1 for time and 2 for space
  2123 (both increase the runtime).
  2124 
  2125 * Isar session: The initial use of ROOT.ML is now always timed,
  2126 i.e. the log will show the actual process times, in contrast to the
  2127 elapsed wall-clock time that the outer shell wrapper produces.
  2128 
  2129 * Simplifier: improved handling of bound variables (nameless
  2130 representation, avoid allocating new strings).  Simprocs that invoke
  2131 the Simplifier recursively should use Simplifier.inherit_bounds to
  2132 avoid local name clashes.  Failure to do so produces warnings
  2133 "Simplifier: renamed bound variable ..."; set Simplifier.debug_bounds
  2134 for further details.
  2135 
  2136 * ML functions legacy_bindings and use_legacy_bindings produce ML fact
  2137 bindings for all theorems stored within a given theory; this may help
  2138 in porting non-Isar theories to Isar ones, while keeping ML proof
  2139 scripts for the time being.
  2140 
  2141 * ML operator HTML.with_charset specifies the charset begin used for
  2142 generated HTML files.  For example:
  2143 
  2144   HTML.with_charset "utf-8" use_thy "Hebrew";
  2145   HTML.with_charset "utf-8" use_thy "Chinese";
  2146 
  2147 
  2148 *** System ***
  2149 
  2150 * Allow symlinks to all proper Isabelle executables (Isabelle,
  2151 isabelle, isatool etc.).
  2152 
  2153 * ISABELLE_DOC_FORMAT setting specifies preferred document format (for
  2154 isatool doc, isatool mkdir, display_drafts etc.).
  2155 
  2156 * isatool usedir: option -f allows specification of the ML file to be
  2157 used by Isabelle; default is ROOT.ML.
  2158 
  2159 * New isatool version outputs the version identifier of the Isabelle
  2160 distribution being used.
  2161 
  2162 * HOL: new isatool dimacs2hol converts files in DIMACS CNF format
  2163 (containing Boolean satisfiability problems) into Isabelle/HOL
  2164 theories.
  2165 
  2166 
  2167 
  2168 New in Isabelle2004 (April 2004)
  2169 --------------------------------
  2170 
  2171 *** General ***
  2172 
  2173 * Provers/order.ML:  new efficient reasoner for partial and linear orders.
  2174   Replaces linorder.ML.
  2175 
  2176 * Pure: Greek letters (except small lambda, \<lambda>), as well as Gothic
  2177   (\<aa>...\<zz>\<AA>...\<ZZ>), calligraphic (\<A>...\<Z>), and Euler
  2178   (\<a>...\<z>), are now considered normal letters, and can therefore
  2179   be used anywhere where an ASCII letter (a...zA...Z) has until
  2180   now. COMPATIBILITY: This obviously changes the parsing of some
  2181   terms, especially where a symbol has been used as a binder, say
  2182   '\<Pi>x. ...', which is now a type error since \<Pi>x will be parsed
  2183   as an identifier.  Fix it by inserting a space around former
  2184   symbols.  Call 'isatool fixgreek' to try to fix parsing errors in
  2185   existing theory and ML files.
  2186 
  2187 * Pure: Macintosh and Windows line-breaks are now allowed in theory files.
  2188 
  2189 * Pure: single letter sub/superscripts (\<^isub> and \<^isup>) are now
  2190   allowed in identifiers. Similar to Greek letters \<^isub> is now considered
  2191   a normal (but invisible) letter. For multiple letter subscripts repeat
  2192   \<^isub> like this: x\<^isub>1\<^isub>2.
  2193 
  2194 * Pure: There are now sub-/superscripts that can span more than one
  2195   character. Text between \<^bsub> and \<^esub> is set in subscript in
  2196   ProofGeneral and LaTeX, text between \<^bsup> and \<^esup> in
  2197   superscript. The new control characters are not identifier parts.
  2198 
  2199 * Pure: Control-symbols of the form \<^raw:...> will literally print the
  2200   content of "..." to the latex file instead of \isacntrl... . The "..."
  2201   may consist of any printable characters excluding the end bracket >.
  2202 
  2203 * Pure: Using new Isar command "finalconsts" (or the ML functions
  2204   Theory.add_finals or Theory.add_finals_i) it is now possible to
  2205   declare constants "final", which prevents their being given a definition
  2206   later.  It is useful for constants whose behaviour is fixed axiomatically
  2207   rather than definitionally, such as the meta-logic connectives.
  2208 
  2209 * Pure: 'instance' now handles general arities with general sorts
  2210   (i.e. intersections of classes),
  2211 
  2212 * Presentation: generated HTML now uses a CSS style sheet to make layout
  2213   (somewhat) independent of content. It is copied from lib/html/isabelle.css.
  2214   It can be changed to alter the colors/layout of generated pages.
  2215 
  2216 
  2217 *** Isar ***
  2218 
  2219 * Tactic emulation methods rule_tac, erule_tac, drule_tac, frule_tac,
  2220   cut_tac, subgoal_tac and thin_tac:
  2221   - Now understand static (Isar) contexts.  As a consequence, users of Isar
  2222     locales are no longer forced to write Isar proof scripts.
  2223     For details see Isar Reference Manual, paragraph 4.3.2: Further tactic
  2224     emulations.
  2225   - INCOMPATIBILITY: names of variables to be instantiated may no
  2226     longer be enclosed in quotes.  Instead, precede variable name with `?'.
  2227     This is consistent with the instantiation attribute "where".
  2228 
  2229 * Attributes "where" and "of":
  2230   - Now take type variables of instantiated theorem into account when reading
  2231     the instantiation string.  This fixes a bug that caused instantiated
  2232     theorems to have too special types in some circumstances.
  2233   - "where" permits explicit instantiations of type variables.
  2234 
  2235 * Calculation commands "moreover" and "also" no longer interfere with
  2236   current facts ("this"), admitting arbitrary combinations with "then"
  2237   and derived forms.
  2238 
  2239 * Locales:
  2240   - Goal statements involving the context element "includes" no longer
  2241     generate theorems with internal delta predicates (those ending on
  2242     "_axioms") in the premise.
  2243     Resolve particular premise with <locale>.intro to obtain old form.
  2244   - Fixed bug in type inference ("unify_frozen") that prevented mix of target
  2245     specification and "includes" elements in goal statement.
  2246   - Rule sets <locale>.intro and <locale>.axioms no longer declared as
  2247     [intro?] and [elim?] (respectively) by default.
  2248   - Experimental command for instantiation of locales in proof contexts:
  2249         instantiate <label>[<attrs>]: <loc>
  2250     Instantiates locale <loc> and adds all its theorems to the current context
  2251     taking into account their attributes.  Label and attrs are optional
  2252     modifiers, like in theorem declarations.  If present, names of
  2253     instantiated theorems are qualified with <label>, and the attributes
  2254     <attrs> are applied after any attributes these theorems might have already.
  2255       If the locale has assumptions, a chained fact of the form
  2256     "<loc> t1 ... tn" is expected from which instantiations of the parameters
  2257     are derived.  The command does not support old-style locales declared
  2258     with "locale (open)".
  2259       A few (very simple) examples can be found in FOL/ex/LocaleInst.thy.
  2260 
  2261 * HOL: Tactic emulation methods induct_tac and case_tac understand static
  2262   (Isar) contexts.
  2263 
  2264 
  2265 *** HOL ***
  2266 
  2267 * Proof import: new image HOL4 contains the imported library from
  2268   the HOL4 system with about 2500 theorems. It is imported by
  2269   replaying proof terms produced by HOL4 in Isabelle. The HOL4 image
  2270   can be used like any other Isabelle image.  See
  2271   HOL/Import/HOL/README for more information.
  2272 
  2273 * Simplifier:
  2274   - Much improved handling of linear and partial orders.
  2275     Reasoners for linear and partial orders are set up for type classes
  2276     "linorder" and "order" respectively, and are added to the default simpset
  2277     as solvers.  This means that the simplifier can build transitivity chains
  2278     to solve goals from the assumptions.
  2279   - INCOMPATIBILITY: old proofs break occasionally.  Typically, applications
  2280     of blast or auto after simplification become unnecessary because the goal
  2281     is solved by simplification already.
  2282 
  2283 * Numerics: new theory Ring_and_Field contains over 250 basic numerical laws,
  2284     all proved in axiomatic type classes for semirings, rings and fields.
  2285 
  2286 * Numerics:
  2287   - Numeric types (nat, int, and in HOL-Complex rat, real, complex, etc.) are
  2288     now formalized using the Ring_and_Field theory mentioned above.
  2289   - INCOMPATIBILITY: simplification and arithmetic behaves somewhat differently
  2290     than before, because now they are set up once in a generic manner.
  2291   - INCOMPATIBILITY: many type-specific arithmetic laws have gone.
  2292     Look for the general versions in Ring_and_Field (and Power if they concern
  2293     exponentiation).
  2294 
  2295 * Type "rat" of the rational numbers is now available in HOL-Complex.
  2296 
  2297 * Records:
  2298   - Record types are now by default printed with their type abbreviation
  2299     instead of the list of all field types. This can be configured via
  2300     the reference "print_record_type_abbr".
  2301   - Simproc "record_upd_simproc" for simplification of multiple updates added
  2302     (not enabled by default).
  2303   - Simproc "record_ex_sel_eq_simproc" to simplify EX x. sel r = x resp.
  2304     EX x. x = sel r to True (not enabled by default).
  2305   - Tactic "record_split_simp_tac" to split and simplify records added.
  2306 
  2307 * 'specification' command added, allowing for definition by
  2308   specification.  There is also an 'ax_specification' command that
  2309   introduces the new constants axiomatically.
  2310 
  2311 * arith(_tac) is now able to generate counterexamples for reals as well.
  2312 
  2313 * HOL-Algebra: new locale "ring" for non-commutative rings.
  2314 
  2315 * HOL-ex: InductiveInvariant_examples illustrates advanced recursive function
  2316   definitions, thanks to Sava Krsti\'{c} and John Matthews.
  2317 
  2318 * HOL-Matrix: a first theory for matrices in HOL with an application of
  2319   matrix theory to linear programming.
  2320 
  2321 * Unions and Intersections:
  2322   The latex output syntax of UN and INT has been changed
  2323   from "\Union x \in A. B" to "\Union_{x \in A} B"
  2324   i.e. the index formulae has become a subscript.
  2325   Similarly for "\Union x. B", and for \Inter instead of \Union.
  2326 
  2327 * Unions and Intersections over Intervals:
  2328   There is new short syntax "UN i<=n. A" for "UN i:{0..n}. A". There is
  2329   also an x-symbol version with subscripts "\<Union>\<^bsub>i <= n\<^esub>. A"
  2330   like in normal math, and corresponding versions for < and for intersection.
  2331 
  2332 * HOL/List: Ordering "lexico" is renamed "lenlex" and the standard
  2333   lexicographic dictonary ordering has been added as "lexord".
  2334 
  2335 * ML: the legacy theory structures Int and List have been removed. They had
  2336   conflicted with ML Basis Library structures having the same names.
  2337 
  2338 * 'refute' command added to search for (finite) countermodels.  Only works
  2339   for a fragment of HOL.  The installation of an external SAT solver is
  2340   highly recommended.  See "HOL/Refute.thy" for details.
  2341 
  2342 * 'quickcheck' command: Allows to find counterexamples by evaluating
  2343   formulae under an assignment of free variables to random values.
  2344   In contrast to 'refute', it can deal with inductive datatypes,
  2345   but cannot handle quantifiers. See "HOL/ex/Quickcheck_Examples.thy"
  2346   for examples.
  2347 
  2348 
  2349 *** HOLCF ***
  2350 
  2351 * Streams now come with concatenation and are part of the HOLCF image
  2352 
  2353 
  2354 
  2355 New in Isabelle2003 (May 2003)
  2356 ------------------------------
  2357 
  2358 *** General ***
  2359 
  2360 * Provers/simplifier:
  2361 
  2362   - Completely reimplemented method simp (ML: Asm_full_simp_tac):
  2363     Assumptions are now subject to complete mutual simplification,
  2364     not just from left to right. The simplifier now preserves
  2365     the order of assumptions.
  2366 
  2367     Potential INCOMPATIBILITY:
  2368 
  2369     -- simp sometimes diverges where the old version did
  2370        not, e.g. invoking simp on the goal
  2371 
  2372         [| P (f x); y = x; f x = f y |] ==> Q
  2373 
  2374        now gives rise to the infinite reduction sequence
  2375 
  2376         P(f x) --(f x = f y)--> P(f y) --(y = x)--> P(f x) --(f x = f y)--> ...
  2377 
  2378        Using "simp (asm_lr)" (ML: Asm_lr_simp_tac) instead often solves this
  2379        kind of problem.
  2380 
  2381     -- Tactics combining classical reasoner and simplification (such as auto)
  2382        are also affected by this change, because many of them rely on
  2383        simp. They may sometimes diverge as well or yield a different numbers
  2384        of subgoals. Try to use e.g. force, fastsimp, or safe instead of auto
  2385        in case of problems. Sometimes subsequent calls to the classical
  2386        reasoner will fail because a preceeding call to the simplifier too
  2387        eagerly simplified the goal, e.g. deleted redundant premises.
  2388 
  2389   - The simplifier trace now shows the names of the applied rewrite rules
  2390 
  2391   - You can limit the number of recursive invocations of the simplifier
  2392     during conditional rewriting (where the simplifie tries to solve the
  2393     conditions before applying the rewrite rule):
  2394     ML "simp_depth_limit := n"
  2395     where n is an integer. Thus you can force termination where previously
  2396     the simplifier would diverge.
  2397 
  2398   - Accepts free variables as head terms in congruence rules.  Useful in Isar.
  2399 
  2400   - No longer aborts on failed congruence proof.  Instead, the
  2401     congruence is ignored.
  2402 
  2403 * Pure: New generic framework for extracting programs from constructive
  2404   proofs. See HOL/Extraction.thy for an example instantiation, as well
  2405   as HOL/Extraction for some case studies.
  2406 
  2407 * Pure: The main goal of the proof state is no longer shown by default, only
  2408 the subgoals. This behaviour is controlled by a new flag.
  2409    PG menu: Isabelle/Isar -> Settings -> Show Main Goal
  2410 (ML: Proof.show_main_goal).
  2411 
  2412 * Pure: You can find all matching introduction rules for subgoal 1, i.e. all
  2413 rules whose conclusion matches subgoal 1:
  2414       PG menu: Isabelle/Isar -> Show me -> matching rules
  2415 The rules are ordered by how closely they match the subgoal.
  2416 In particular, rules that solve a subgoal outright are displayed first
  2417 (or rather last, the way they are printed).
  2418 (ML: ProofGeneral.print_intros())
  2419 
  2420 * Pure: New flag trace_unify_fail causes unification to print
  2421 diagnostic information (PG: in trace buffer) when it fails. This is
  2422 useful for figuring out why single step proofs like rule, erule or
  2423 assumption failed.
  2424 
  2425 * Pure: Locale specifications now produce predicate definitions
  2426 according to the body of text (covering assumptions modulo local
  2427 definitions); predicate "loc_axioms" covers newly introduced text,
  2428 while "loc" is cumulative wrt. all included locale expressions; the
  2429 latter view is presented only on export into the global theory
  2430 context; potential INCOMPATIBILITY, use "(open)" option to fall back
  2431 on the old view without predicates;
  2432 
  2433 * Pure: predefined locales "var" and "struct" are useful for sharing
  2434 parameters (as in CASL, for example); just specify something like
  2435 ``var x + var y + struct M'' as import;
  2436 
  2437 * Pure: improved thms_containing: proper indexing of facts instead of
  2438 raw theorems; check validity of results wrt. current name space;
  2439 include local facts of proof configuration (also covers active
  2440 locales), cover fixed variables in index; may use "_" in term
  2441 specification; an optional limit for the number of printed facts may
  2442 be given (the default is 40);
  2443 
  2444 * Pure: disallow duplicate fact bindings within new-style theory files
  2445 (batch-mode only);
  2446 
  2447 * Provers: improved induct method: assumptions introduced by case
  2448 "foo" are split into "foo.hyps" (from the rule) and "foo.prems" (from
  2449 the goal statement); "foo" still refers to all facts collectively;
  2450 
  2451 * Provers: the function blast.overloaded has been removed: all constants
  2452 are regarded as potentially overloaded, which improves robustness in exchange
  2453 for slight decrease in efficiency;
  2454 
  2455 * Provers/linorder: New generic prover for transitivity reasoning over
  2456 linear orders.  Note: this prover is not efficient!
  2457 
  2458 * Isar: preview of problems to finish 'show' now produce an error
  2459 rather than just a warning (in interactive mode);
  2460 
  2461 
  2462 *** HOL ***
  2463 
  2464 * arith(_tac)
  2465 
  2466  - Produces a counter example if it cannot prove a goal.
  2467    Note that the counter example may be spurious if the goal is not a formula
  2468    of quantifier-free linear arithmetic.
  2469    In ProofGeneral the counter example appears in the trace buffer.
  2470 
  2471  - Knows about div k and mod k where k is a numeral of type nat or int.
  2472 
  2473  - Calls full Presburger arithmetic (by Amine Chaieb) if quantifier-free
  2474    linear arithmetic fails. This takes account of quantifiers and divisibility.
  2475    Presburger arithmetic can also be called explicitly via presburger(_tac).
  2476 
  2477 * simp's arithmetic capabilities have been enhanced a bit: it now
  2478 takes ~= in premises into account (by performing a case split);
  2479 
  2480 * simp reduces "m*(n div m) + n mod m" to n, even if the two summands
  2481 are distributed over a sum of terms;
  2482 
  2483 * New tactic "trans_tac" and method "trans" instantiate
  2484 Provers/linorder.ML for axclasses "order" and "linorder" (predicates
  2485 "<=", "<" and "=").
  2486 
  2487 * function INCOMPATIBILITIES: Pi-sets have been redefined and moved from main
  2488 HOL to Library/FuncSet; constant "Fun.op o" is now called "Fun.comp";
  2489 
  2490 * 'typedef' command has new option "open" to suppress the set
  2491 definition;
  2492 
  2493 * functions Min and Max on finite sets have been introduced (theory
  2494 Finite_Set);
  2495 
  2496 * attribute [symmetric] now works for relations as well; it turns
  2497 (x,y) : R^-1 into (y,x) : R, and vice versa;
  2498 
  2499 * induct over a !!-quantified statement (say !!x1..xn):
  2500   each "case" automatically performs "fix x1 .. xn" with exactly those names.
  2501 
  2502 * Map: `empty' is no longer a constant but a syntactic abbreviation for
  2503 %x. None. Warning: empty_def now refers to the previously hidden definition
  2504 of the empty set.
  2505 
  2506 * Algebra: formalization of classical algebra.  Intended as base for
  2507 any algebraic development in Isabelle.  Currently covers group theory
  2508 (up to Sylow's theorem) and ring theory (Universal Property of
  2509 Univariate Polynomials).  Contributions welcome;
  2510 
  2511 * GroupTheory: deleted, since its material has been moved to Algebra;
  2512 
  2513 * Complex: new directory of the complex numbers with numeric constants,
  2514 nonstandard complex numbers, and some complex analysis, standard and
  2515 nonstandard (Jacques Fleuriot);
  2516 
  2517 * HOL-Complex: new image for analysis, replacing HOL-Real and HOL-Hyperreal;
  2518 
  2519 * Hyperreal: introduced Gauge integration and hyperreal logarithms (Jacques
  2520 Fleuriot);
  2521 
  2522 * Real/HahnBanach: updated and adapted to locales;
  2523 
  2524 * NumberTheory: added Gauss's law of quadratic reciprocity (by Avigad,
  2525 Gray and Kramer);
  2526 
  2527 * UNITY: added the Meier-Sanders theory of progress sets;
  2528 
  2529 * MicroJava: bytecode verifier and lightweight bytecode verifier
  2530 as abstract algorithms, instantiated to the JVM;
  2531 
  2532 * Bali: Java source language formalization. Type system, operational
  2533 semantics, axiomatic semantics. Supported language features:
  2534 classes, interfaces, objects,virtual methods, static methods,
  2535 static/instance fields, arrays, access modifiers, definite
  2536 assignment, exceptions.
  2537 
  2538 
  2539 *** ZF ***
  2540 
  2541 * ZF/Constructible: consistency proof for AC (Gdel's constructible
  2542 universe, etc.);
  2543 
  2544 * Main ZF: virtually all theories converted to new-style format;
  2545 
  2546 
  2547 *** ML ***
  2548 
  2549 * Pure: Tactic.prove provides sane interface for internal proofs;
  2550 omits the infamous "standard" operation, so this is more appropriate
  2551 than prove_goalw_cterm in many situations (e.g. in simprocs);
  2552 
  2553 * Pure: improved error reporting of simprocs;
  2554 
  2555 * Provers: Simplifier.simproc(_i) provides sane interface for setting
  2556 up simprocs;
  2557 
  2558 
  2559 *** Document preparation ***
  2560 
  2561 * uses \par instead of \\ for line breaks in theory text. This may
  2562 shift some page breaks in large documents. To get the old behaviour
  2563 use \renewcommand{\isanewline}{\mbox{}\\\mbox{}} in root.tex.
  2564 
  2565 * minimized dependencies of isabelle.sty and isabellesym.sty on
  2566 other packages
  2567 
  2568 * \<euro> now needs package babel/greek instead of marvosym (which
  2569 broke \Rightarrow)
  2570 
  2571 * normal size for \<zero>...\<nine> (uses \mathbf instead of
  2572 textcomp package)
  2573 
  2574 
  2575 
  2576 New in Isabelle2002 (March 2002)
  2577 --------------------------------
  2578 
  2579 *** Document preparation ***
  2580 
  2581 * greatly simplified document preparation setup, including more
  2582 graceful interpretation of isatool usedir -i/-d/-D options, and more
  2583 instructive isatool mkdir; users should basically be able to get
  2584 started with "isatool mkdir HOL Test && isatool make"; alternatively,
  2585 users may run a separate document processing stage manually like this:
  2586 "isatool usedir -D output HOL Test && isatool document Test/output";
  2587 
  2588 * theory dependency graph may now be incorporated into documents;
  2589 isatool usedir -g true will produce session_graph.eps/.pdf for use
  2590 with \includegraphics of LaTeX;
  2591 
  2592 * proper spacing of consecutive markup elements, especially text
  2593 blocks after section headings;
  2594 
  2595 * support bold style (for single symbols only), input syntax is like
  2596 this: "\<^bold>\<alpha>" or "\<^bold>A";
  2597 
  2598 * \<bullet> is now output as bold \cdot by default, which looks much
  2599 better in printed text;
  2600 
  2601 * added default LaTeX bindings for \<tturnstile> and \<TTurnstile>;
  2602 note that these symbols are currently unavailable in Proof General /
  2603 X-Symbol; new symbols \<zero>, \<one>, ..., \<nine>, and \<euro>;
  2604 
  2605 * isatool latex no longer depends on changed TEXINPUTS, instead
  2606 isatool document copies the Isabelle style files to the target
  2607 location;
  2608 
  2609 
  2610 *** Isar ***
  2611 
  2612 * Pure/Provers: improved proof by cases and induction;
  2613   - 'case' command admits impromptu naming of parameters (such as
  2614     "case (Suc n)");
  2615   - 'induct' method divinates rule instantiation from the inductive
  2616     claim; no longer requires excessive ?P bindings for proper
  2617     instantiation of cases;
  2618   - 'induct' method properly enumerates all possibilities of set/type
  2619     rules; as a consequence facts may be also passed through *type*
  2620     rules without further ado;
  2621   - 'induct' method now derives symbolic cases from the *rulified*
  2622     rule (before it used to rulify cases stemming from the internal
  2623     atomized version); this means that the context of a non-atomic
  2624     statement becomes is included in the hypothesis, avoiding the
  2625     slightly cumbersome show "PROP ?case" form;
  2626   - 'induct' may now use elim-style induction rules without chaining
  2627     facts, using ``missing'' premises from the goal state; this allows
  2628     rules stemming from inductive sets to be applied in unstructured
  2629     scripts, while still benefitting from proper handling of non-atomic
  2630     statements; NB: major inductive premises need to be put first, all
  2631     the rest of the goal is passed through the induction;
  2632   - 'induct' proper support for mutual induction involving non-atomic
  2633     rule statements (uses the new concept of simultaneous goals, see
  2634     below);
  2635   - append all possible rule selections, but only use the first
  2636     success (no backtracking);
  2637   - removed obsolete "(simplified)" and "(stripped)" options of methods;
  2638   - undeclared rule case names default to numbers 1, 2, 3, ...;
  2639   - added 'print_induct_rules' (covered by help item in recent Proof
  2640     General versions);
  2641   - moved induct/cases attributes to Pure, methods to Provers;
  2642   - generic method setup instantiated for FOL and HOL;
  2643 
  2644 * Pure: support multiple simultaneous goal statements, for example
  2645 "have a: A and b: B" (same for 'theorem' etc.); being a pure
  2646 meta-level mechanism, this acts as if several individual goals had
  2647 been stated separately; in particular common proof methods need to be
  2648 repeated in order to cover all claims; note that a single elimination
  2649 step is *not* sufficient to establish the two conjunctions, so this
  2650 fails:
  2651 
  2652   assume "A & B" then have A and B ..   (*".." fails*)
  2653 
  2654 better use "obtain" in situations as above; alternative refer to
  2655 multi-step methods like 'auto', 'simp_all', 'blast+' etc.;
  2656 
  2657 * Pure: proper integration with ``locales''; unlike the original
  2658 version by Florian Kammller, Isar locales package high-level proof
  2659 contexts rather than raw logical ones (e.g. we admit to include
  2660 attributes everywhere); operations on locales include merge and
  2661 rename; support for implicit arguments (``structures''); simultaneous
  2662 type-inference over imports and text; see also HOL/ex/Locales.thy for
  2663 some examples;
  2664 
  2665 * Pure: the following commands have been ``localized'', supporting a
  2666 target locale specification "(in name)": 'lemma', 'theorem',
  2667 'corollary', 'lemmas', 'theorems', 'declare'; the results will be
  2668 stored both within the locale and at the theory level (exported and
  2669 qualified by the locale name);
  2670 
  2671 * Pure: theory goals may now be specified in ``long'' form, with
  2672 ad-hoc contexts consisting of arbitrary locale elements. for example
  2673 ``lemma foo: fixes x assumes "A x" shows "B x"'' (local syntax and
  2674 definitions may be given, too); the result is a meta-level rule with
  2675 the context elements being discharged in the obvious way;
  2676 
  2677 * Pure: new proof command 'using' allows to augment currently used
  2678 facts after a goal statement ('using' is syntactically analogous to
  2679 'apply', but acts on the goal's facts only); this allows chained facts
  2680 to be separated into parts given before and after a claim, as in
  2681 ``from a and b have C using d and e <proof>'';
  2682 
  2683 * Pure: renamed "antecedent" case to "rule_context";
  2684 
  2685 * Pure: new 'judgment' command records explicit information about the
  2686 object-logic embedding (used by several tools internally); no longer
  2687 use hard-wired "Trueprop";
  2688 
  2689 * Pure: added 'corollary' command;
  2690 
  2691 * Pure: fixed 'token_translation' command;
  2692 
  2693 * Pure: removed obsolete 'exported' attribute;
  2694 
  2695 * Pure: dummy pattern "_" in is/let is now automatically lifted over
  2696 bound variables: "ALL x. P x --> Q x" (is "ALL x. _ --> ?C x")
  2697 supersedes more cumbersome ... (is "ALL x. _ x --> ?C x");
  2698 
  2699 * Pure: method 'atomize' presents local goal premises as object-level
  2700 statements (atomic meta-level propositions); setup controlled via
  2701 rewrite rules declarations of 'atomize' attribute; example
  2702 application: 'induct' method with proper rule statements in improper
  2703 proof *scripts*;
  2704 
  2705 * Pure: emulation of instantiation tactics (rule_tac, cut_tac, etc.)
  2706 now consider the syntactic context of assumptions, giving a better
  2707 chance to get type-inference of the arguments right (this is
  2708 especially important for locales);
  2709 
  2710 * Pure: "sorry" no longer requires quick_and_dirty in interactive
  2711 mode;
  2712 
  2713 * Pure/obtain: the formal conclusion "thesis", being marked as
  2714 ``internal'', may no longer be reference directly in the text;
  2715 potential INCOMPATIBILITY, may need to use "?thesis" in rare
  2716 situations;
  2717 
  2718 * Pure: generic 'sym' attribute which declares a rule both as pure
  2719 'elim?' and for the 'symmetric' operation;
  2720 
  2721 * Pure: marginal comments ``--'' may now occur just anywhere in the
  2722 text; the fixed correlation with particular command syntax has been
  2723 discontinued;
  2724 
  2725 * Pure: new method 'rules' is particularly well-suited for proof
  2726 search in intuitionistic logic; a bit slower than 'blast' or 'fast',
  2727 but often produces more compact proof terms with less detours;
  2728 
  2729 * Pure/Provers/classical: simplified integration with pure rule
  2730 attributes and methods; the classical "intro?/elim?/dest?"
  2731 declarations coincide with the pure ones; the "rule" method no longer
  2732 includes classically swapped intros; "intro" and "elim" methods no
  2733 longer pick rules from the context; also got rid of ML declarations
  2734 AddXIs/AddXEs/AddXDs; all of this has some potential for
  2735 INCOMPATIBILITY;
  2736 
  2737 * Provers/classical: attribute 'swapped' produces classical inversions
  2738 of introduction rules;
  2739 
  2740 * Provers/simplifier: 'simplified' attribute may refer to explicit
  2741 rules instead of full simplifier context; 'iff' attribute handles
  2742 conditional rules;
  2743 
  2744 * HOL: 'typedef' now allows alternative names for Rep/Abs morphisms;
  2745 
  2746 * HOL: 'recdef' now fails on unfinished automated proofs, use
  2747 "(permissive)" option to recover old behavior;
  2748 
  2749 * HOL: 'inductive' no longer features separate (collective) attributes
  2750 for 'intros' (was found too confusing);
  2751 
  2752 * HOL: properly declared induction rules less_induct and
  2753 wf_induct_rule;
  2754 
  2755 
  2756 *** HOL ***
  2757 
  2758 * HOL: moved over to sane numeral syntax; the new policy is as
  2759 follows:
  2760 
  2761   - 0 and 1 are polymorphic constants, which are defined on any
  2762   numeric type (nat, int, real etc.);
  2763 
  2764   - 2, 3, 4, ... and -1, -2, -3, ... are polymorphic numerals, based
  2765   binary representation internally;
  2766 
  2767   - type nat has special constructor Suc, and generally prefers Suc 0
  2768   over 1::nat and Suc (Suc 0) over 2::nat;
  2769 
  2770 This change may cause significant problems of INCOMPATIBILITY; here
  2771 are some hints on converting existing sources:
  2772 
  2773   - due to the new "num" token, "-0" and "-1" etc. are now atomic
  2774   entities, so expressions involving "-" (unary or binary minus) need
  2775   to be spaced properly;
  2776 
  2777   - existing occurrences of "1" may need to be constraint "1::nat" or
  2778   even replaced by Suc 0; similar for old "2";
  2779 
  2780   - replace "#nnn" by "nnn", and "#-nnn" by "-nnn";
  2781 
  2782   - remove all special provisions on numerals in proofs;
  2783 
  2784 * HOL: simp rules nat_number expand numerals on nat to Suc/0
  2785 representation (depends on bin_arith_simps in the default context);
  2786 
  2787 * HOL: symbolic syntax for x^2 (numeral 2);
  2788 
  2789 * HOL: the class of all HOL types is now called "type" rather than
  2790 "term"; INCOMPATIBILITY, need to adapt references to this type class
  2791 in axclass/classes, instance/arities, and (usually rare) occurrences
  2792 in typings (of consts etc.); internally the class is called
  2793 "HOL.type", ML programs should refer to HOLogic.typeS;
  2794 
  2795 * HOL/record package improvements:
  2796   - new derived operations "fields" to build a partial record section,
  2797     "extend" to promote a fixed record to a record scheme, and
  2798     "truncate" for the reverse; cf. theorems "xxx.defs", which are *not*
  2799     declared as simp by default;
  2800   - shared operations ("more", "fields", etc.) now need to be always
  2801     qualified) --- potential INCOMPATIBILITY;
  2802   - removed "make_scheme" operations (use "make" with "extend") --
  2803     INCOMPATIBILITY;
  2804   - removed "more" class (simply use "term") -- INCOMPATIBILITY;
  2805   - provides cases/induct rules for use with corresponding Isar
  2806     methods (for concrete records, record schemes, concrete more
  2807     parts, and schematic more parts -- in that order);
  2808   - internal definitions directly based on a light-weight abstract
  2809     theory of product types over typedef rather than datatype;
  2810 
  2811 * HOL: generic code generator for generating executable ML code from
  2812 specifications; specific support for HOL constructs such as inductive
  2813 datatypes and sets, as well as recursive functions; can be invoked
  2814 via 'generate_code' theory section;
  2815 
  2816 * HOL: canonical cases/induct rules for n-tuples (n = 3..7);
  2817 
  2818 * HOL: consolidated and renamed several theories.  In particular:
  2819         Ord.thy has been absorbed into HOL.thy
  2820         String.thy has been absorbed into List.thy
  2821 
  2822 * HOL: concrete setsum syntax "\<Sum>i:A. b" == "setsum (%i. b) A"
  2823 (beware of argument permutation!);
  2824 
  2825 * HOL: linorder_less_split superseded by linorder_cases;
  2826 
  2827 * HOL/List: "nodups" renamed to "distinct";
  2828 
  2829 * HOL: added "The" definite description operator; move Hilbert's "Eps"
  2830 to peripheral theory "Hilbert_Choice"; some INCOMPATIBILITIES:
  2831   - Ex_def has changed, now need to use some_eq_ex
  2832 
  2833 * HOL: made split_all_tac safe; EXISTING PROOFS MAY FAIL OR LOOP, so
  2834 in this (rare) case use:
  2835 
  2836   delSWrapper "split_all_tac"
  2837   addSbefore ("unsafe_split_all_tac", unsafe_split_all_tac)
  2838 
  2839 * HOL: added safe wrapper "split_conv_tac" to claset; EXISTING PROOFS
  2840 MAY FAIL;
  2841 
  2842 * HOL: introduced f^n = f o ... o f; warning: due to the limits of
  2843 Isabelle's type classes, ^ on functions and relations has too general
  2844 a domain, namely ('a * 'b) set and 'a => 'b; this means that it may be
  2845 necessary to attach explicit type constraints;
  2846 
  2847 * HOL/Relation: the prefix name of the infix "O" has been changed from
  2848 "comp" to "rel_comp"; INCOMPATIBILITY: a few theorems have been
  2849 renamed accordingly (eg "compI" -> "rel_compI").
  2850 
  2851 * HOL: syntax translations now work properly with numerals and records
  2852 expressions;
  2853 
  2854 * HOL: bounded abstraction now uses syntax "%" / "\<lambda>" instead
  2855 of "lam" -- INCOMPATIBILITY;
  2856 
  2857 * HOL: got rid of some global declarations (potential INCOMPATIBILITY
  2858 for ML tools): const "()" renamed "Product_Type.Unity", type "unit"
  2859 renamed "Product_Type.unit";
  2860 
  2861 * HOL: renamed rtrancl_into_rtrancl2 to converse_rtrancl_into_rtrancl
  2862 
  2863 * HOL: removed obsolete theorem "optionE" (use "option.exhaust", or
  2864 the "cases" method);
  2865 
  2866 * HOL/GroupTheory: group theory examples including Sylow's theorem (by
  2867 Florian Kammller);
  2868 
  2869 * HOL/IMP: updated and converted to new-style theory format; several
  2870 parts turned into readable document, with proper Isar proof texts and
  2871 some explanations (by Gerwin Klein);
  2872 
  2873 * HOL-Real: added Complex_Numbers (by Gertrud Bauer);
  2874 
  2875 * HOL-Hyperreal is now a logic image;
  2876 
  2877 
  2878 *** HOLCF ***
  2879 
  2880 * Isar: consts/constdefs supports mixfix syntax for continuous
  2881 operations;
  2882 
  2883 * Isar: domain package adapted to new-style theory format, e.g. see
  2884 HOLCF/ex/Dnat.thy;
  2885 
  2886 * theory Lift: proper use of rep_datatype lift instead of ML hacks --
  2887 potential INCOMPATIBILITY; now use plain induct_tac instead of former
  2888 lift.induct_tac, always use UU instead of Undef;
  2889 
  2890 * HOLCF/IMP: updated and converted to new-style theory;
  2891 
  2892 
  2893 *** ZF ***
  2894 
  2895 * Isar: proper integration of logic-specific tools and packages,
  2896 including theory commands '(co)inductive', '(co)datatype',
  2897 'rep_datatype', 'inductive_cases', as well as methods 'ind_cases',
  2898 'induct_tac', 'case_tac', and 'typecheck' (with attribute 'TC');
  2899 
  2900 * theory Main no longer includes AC; for the Axiom of Choice, base
  2901 your theory on Main_ZFC;
  2902 
  2903 * the integer library now covers quotients and remainders, with many
  2904 laws relating division to addition, multiplication, etc.;
  2905 
  2906 * ZF/UNITY: Chandy and Misra's UNITY is now available in ZF, giving a
  2907 typeless version of the formalism;
  2908 
  2909 * ZF/AC, Coind, IMP, Resid: updated and converted to new-style theory
  2910 format;
  2911 
  2912 * ZF/Induct: new directory for examples of inductive definitions,
  2913 including theory Multiset for multiset orderings; converted to
  2914 new-style theory format;
  2915 
  2916 * ZF: many new theorems about lists, ordinals, etc.;
  2917 
  2918 
  2919 *** General ***
  2920 
  2921 * Pure/kernel: meta-level proof terms (by Stefan Berghofer); reference
  2922 variable proof controls level of detail: 0 = no proofs (only oracle
  2923 dependencies), 1 = lemma dependencies, 2 = compact proof terms; see
  2924 also ref manual for further ML interfaces;
  2925 
  2926 * Pure/axclass: removed obsolete ML interface
  2927 goal_subclass/goal_arity;
  2928 
  2929 * Pure/syntax: new token syntax "num" for plain numerals (without "#"
  2930 of "xnum"); potential INCOMPATIBILITY, since -0, -1 etc. are now
  2931 separate tokens, so expressions involving minus need to be spaced
  2932 properly;
  2933 
  2934 * Pure/syntax: support non-oriented infixes, using keyword "infix"
  2935 rather than "infixl" or "infixr";
  2936 
  2937 * Pure/syntax: concrete syntax for dummy type variables admits genuine
  2938 sort constraint specifications in type inference; e.g. "x::_::foo"
  2939 ensures that the type of "x" is of sort "foo" (but not necessarily a
  2940 type variable);
  2941 
  2942 * Pure/syntax: print modes "type_brackets" and "no_type_brackets"
  2943 control output of nested => (types); the default behavior is
  2944 "type_brackets";
  2945 
  2946 * Pure/syntax: builtin parse translation for "_constify" turns valued
  2947 tokens into AST constants;
  2948 
  2949 * Pure/syntax: prefer later declarations of translations and print
  2950 translation functions; potential INCOMPATIBILITY: need to reverse
  2951 multiple declarations for same syntax element constant;
  2952 
  2953 * Pure/show_hyps reset by default (in accordance to existing Isar
  2954 practice);
  2955 
  2956 * Provers/classical: renamed addaltern to addafter, addSaltern to
  2957 addSafter;
  2958 
  2959 * Provers/clasimp: ``iff'' declarations now handle conditional rules
  2960 as well;
  2961 
  2962 * system: tested support for MacOS X; should be able to get Isabelle +
  2963 Proof General to work in a plain Terminal after installing Poly/ML
  2964 (e.g. from the Isabelle distribution area) and GNU bash alone
  2965 (e.g. from http://www.apple.com); full X11, XEmacs and X-Symbol
  2966 support requires further installations, e.g. from
  2967 http://fink.sourceforge.net/);
  2968 
  2969 * system: support Poly/ML 4.1.1 (able to manage larger heaps);
  2970 
  2971 * system: reduced base memory usage by Poly/ML (approx. 20 MB instead
  2972 of 40 MB), cf. ML_OPTIONS;
  2973 
  2974 * system: Proof General keywords specification is now part of the
  2975 Isabelle distribution (see etc/isar-keywords.el);
  2976 
  2977 * system: support for persistent Proof General sessions (refrain from
  2978 outdating all loaded theories on startup); user may create writable
  2979 logic images like this: ``isabelle -q HOL Test'';
  2980 
  2981 * system: smart selection of Isabelle process versus Isabelle
  2982 interface, accommodates case-insensitive file systems (e.g. HFS+); may
  2983 run both "isabelle" and "Isabelle" even if file names are badly
  2984 damaged (executable inspects the case of the first letter of its own
  2985 name); added separate "isabelle-process" and "isabelle-interface";
  2986 
  2987 * system: refrain from any attempt at filtering input streams; no
  2988 longer support ``8bit'' encoding of old isabelle font, instead proper
  2989 iso-latin characters may now be used; the related isatools
  2990 "symbolinput" and "nonascii" have disappeared as well;
  2991 
  2992 * system: removed old "xterm" interface (the print modes "xterm" and
  2993 "xterm_color" are still available for direct use in a suitable
  2994 terminal);
  2995 
  2996 
  2997 
  2998 New in Isabelle99-2 (February 2001)
  2999 -----------------------------------
  3000 
  3001 *** Overview of INCOMPATIBILITIES ***
  3002 
  3003 * HOL: please note that theories in the Library and elsewhere often use the
  3004 new-style (Isar) format; to refer to their theorems in an ML script you must
  3005 bind them to ML identifers by e.g.      val thm_name = thm "thm_name";
  3006 
  3007 * HOL: inductive package no longer splits induction rule aggressively,
  3008 but only as far as specified by the introductions given; the old
  3009 format may be recovered via ML function complete_split_rule or attribute
  3010 'split_rule (complete)';
  3011 
  3012 * HOL: induct renamed to lfp_induct, lfp_Tarski to lfp_unfold,
  3013 gfp_Tarski to gfp_unfold;
  3014 
  3015 * HOL: contrapos, contrapos2 renamed to contrapos_nn, contrapos_pp;
  3016 
  3017 * HOL: infix "dvd" now has priority 50 rather than 70 (because it is a
  3018 relation); infix "^^" has been renamed "``"; infix "``" has been
  3019 renamed "`"; "univalent" has been renamed "single_valued";
  3020 
  3021 * HOL/Real: "rinv" and "hrinv" replaced by overloaded "inverse"
  3022 operation;
  3023 
  3024 * HOLCF: infix "`" has been renamed "$"; the symbol syntax is \<cdot>;
  3025 
  3026 * Isar: 'obtain' no longer declares "that" fact as simp/intro;
  3027 
  3028 * Isar/HOL: method 'induct' now handles non-atomic goals; as a
  3029 consequence, it is no longer monotonic wrt. the local goal context
  3030 (which is now passed through the inductive cases);
  3031 
  3032 * Document preparation: renamed standard symbols \<ll> to \<lless> and
  3033 \<gg> to \<ggreater>;
  3034 
  3035 
  3036 *** Document preparation ***
  3037 
  3038 * \isabellestyle{NAME} selects version of Isabelle output (currently
  3039 available: are "it" for near math-mode best-style output, "sl" for
  3040 slanted text style, and "tt" for plain type-writer; if no
  3041 \isabellestyle command is given, output is according to slanted
  3042 type-writer);
  3043 
  3044 * support sub/super scripts (for single symbols only), input syntax is
  3045 like this: "A\<^sup>*" or "A\<^sup>\<star>";
  3046 
  3047 * some more standard symbols; see Appendix A of the system manual for
  3048 the complete list of symbols defined in isabellesym.sty;
  3049 
  3050 * improved isabelle style files; more abstract symbol implementation
  3051 (should now use \isamath{...} and \isatext{...} in custom symbol
  3052 definitions);
  3053 
  3054 * antiquotation @{goals} and @{subgoals} for output of *dynamic* goals
  3055 state; Note that presentation of goal states does not conform to
  3056 actual human-readable proof documents.  Please do not include goal
  3057 states into document output unless you really know what you are doing!
  3058 
  3059 * proper indentation of antiquoted output with proportional LaTeX
  3060 fonts;
  3061 
  3062 * no_document ML operator temporarily disables LaTeX document
  3063 generation;
  3064 
  3065 * isatool unsymbolize tunes sources for plain ASCII communication;
  3066 
  3067 
  3068 *** Isar ***
  3069 
  3070 * Pure: Isar now suffers initial goal statements to contain unbound
  3071 schematic variables (this does not conform to actual readable proof
  3072 documents, due to unpredictable outcome and non-compositional proof
  3073 checking); users who know what they are doing may use schematic goals
  3074 for Prolog-style synthesis of proven results;
  3075 
  3076 * Pure: assumption method (an implicit finishing) now handles actual
  3077 rules as well;
  3078 
  3079 * Pure: improved 'obtain' --- moved to Pure, insert "that" into
  3080 initial goal, declare "that" only as Pure intro (only for single
  3081 steps); the "that" rule assumption may now be involved in implicit
  3082 finishing, thus ".." becomes a feasible for trivial obtains;
  3083 
  3084 * Pure: default proof step now includes 'intro_classes'; thus trivial
  3085 instance proofs may be performed by "..";
  3086 
  3087 * Pure: ?thesis / ?this / "..." now work for pure meta-level
  3088 statements as well;
  3089 
  3090 * Pure: more robust selection of calculational rules;
  3091 
  3092 * Pure: the builtin notion of 'finished' goal now includes the ==-refl
  3093 rule (as well as the assumption rule);
  3094 
  3095 * Pure: 'thm_deps' command visualizes dependencies of theorems and
  3096 lemmas, using the graph browser tool;
  3097 
  3098 * Pure: predict failure of "show" in interactive mode;
  3099 
  3100 * Pure: 'thms_containing' now takes actual terms as arguments;
  3101 
  3102 * HOL: improved method 'induct' --- now handles non-atomic goals
  3103 (potential INCOMPATIBILITY); tuned error handling;
  3104 
  3105 * HOL: cases and induct rules now provide explicit hints about the
  3106 number of facts to be consumed (0 for "type" and 1 for "set" rules);
  3107 any remaining facts are inserted into the goal verbatim;
  3108 
  3109 * HOL: local contexts (aka cases) may now contain term bindings as
  3110 well; the 'cases' and 'induct' methods new provide a ?case binding for
  3111 the result to be shown in each case;
  3112 
  3113 * HOL: added 'recdef_tc' command;
  3114 
  3115 * isatool convert assists in eliminating legacy ML scripts;
  3116 
  3117 
  3118 *** HOL ***
  3119 
  3120 * HOL/Library: a collection of generic theories to be used together
  3121 with main HOL; the theory loader path already includes this directory
  3122 by default; the following existing theories have been moved here:
  3123 HOL/Induct/Multiset, HOL/Induct/Acc (as Accessible_Part), HOL/While
  3124 (as While_Combinator), HOL/Lex/Prefix (as List_Prefix);
  3125 
  3126 * HOL/Unix: "Some aspects of Unix file-system security", a typical
  3127 modelling and verification task performed in Isabelle/HOL +
  3128 Isabelle/Isar + Isabelle document preparation (by Markus Wenzel).
  3129 
  3130 * HOL/Algebra: special summation operator SUM no longer exists, it has
  3131 been replaced by setsum; infix 'assoc' now has priority 50 (like
  3132 'dvd'); axiom 'one_not_zero' has been moved from axclass 'ring' to
  3133 'domain', this makes the theory consistent with mathematical
  3134 literature;
  3135 
  3136 * HOL basics: added overloaded operations "inverse" and "divide"
  3137 (infix "/"), syntax for generic "abs" operation, generic summation
  3138 operator \<Sum>;
  3139 
  3140 * HOL/typedef: simplified package, provide more useful rules (see also
  3141 HOL/subset.thy);
  3142 
  3143 * HOL/datatype: induction rule for arbitrarily branching datatypes is
  3144 now expressed as a proper nested rule (old-style tactic scripts may
  3145 require atomize_strip_tac to cope with non-atomic premises);
  3146 
  3147 * HOL: renamed theory "Prod" to "Product_Type", renamed "split" rule
  3148 to "split_conv" (old name still available for compatibility);
  3149 
  3150 * HOL: improved concrete syntax for strings (e.g. allows translation
  3151 rules with string literals);
  3152 
  3153 * HOL-Real-Hyperreal: this extends HOL-Real with the hyperreals
  3154  and Fleuriot's mechanization of analysis, including the transcendental
  3155  functions for the reals;
  3156 
  3157 * HOL/Real, HOL/Hyperreal: improved arithmetic simplification;
  3158 
  3159 
  3160 *** CTT ***
  3161 
  3162 * CTT: x-symbol support for Pi, Sigma, -->, : (membership); note that
  3163 "lam" is displayed as TWO lambda-symbols
  3164 
  3165 * CTT: theory Main now available, containing everything (that is, Bool
  3166 and Arith);
  3167 
  3168 
  3169 *** General ***
  3170 
  3171 * Pure: the Simplifier has been implemented properly as a derived rule
  3172 outside of the actual kernel (at last!); the overall performance
  3173 penalty in practical applications is about 50%, while reliability of
  3174 the Isabelle inference kernel has been greatly improved;
  3175 
  3176 * print modes "brackets" and "no_brackets" control output of nested =>
  3177 (types) and ==> (props); the default behaviour is "brackets";
  3178 
  3179 * Provers: fast_tac (and friends) now handle actual object-logic rules
  3180 as assumptions as well;
  3181 
  3182 * system: support Poly/ML 4.0;
  3183 
  3184 * system: isatool install handles KDE version 1 or 2;
  3185 
  3186 
  3187 
  3188 New in Isabelle99-1 (October 2000)
  3189 ----------------------------------
  3190 
  3191 *** Overview of INCOMPATIBILITIES ***
  3192 
  3193 * HOL: simplification of natural numbers is much changed; to partly
  3194 recover the old behaviour (e.g. to prevent n+n rewriting to #2*n)
  3195 issue the following ML commands:
  3196 
  3197   Delsimprocs Nat_Numeral_Simprocs.cancel_numerals;
  3198   Delsimprocs [Nat_Numeral_Simprocs.combine_numerals];
  3199 
  3200 * HOL: simplification no longer dives into case-expressions; this is
  3201 controlled by "t.weak_case_cong" for each datatype t;
  3202 
  3203 * HOL: nat_less_induct renamed to less_induct;
  3204 
  3205 * HOL: systematic renaming of the SOME (Eps) rules, may use isatool
  3206 fixsome to patch .thy and .ML sources automatically;
  3207 
  3208   select_equality  -> some_equality
  3209   select_eq_Ex     -> some_eq_ex
  3210   selectI2EX       -> someI2_ex
  3211   selectI2         -> someI2
  3212   selectI          -> someI
  3213   select1_equality -> some1_equality
  3214   Eps_sym_eq       -> some_sym_eq_trivial
  3215   Eps_eq           -> some_eq_trivial
  3216 
  3217 * HOL: exhaust_tac on datatypes superceded by new generic case_tac;
  3218 
  3219 * HOL: removed obsolete theorem binding expand_if (refer to split_if
  3220 instead);
  3221 
  3222 * HOL: the recursion equations generated by 'recdef' are now called
  3223 f.simps instead of f.rules;
  3224 
  3225 * HOL: qed_spec_mp now also handles bounded ALL as well;
  3226 
  3227 * HOL: 0 is now overloaded, so the type constraint ":: nat" may
  3228 sometimes be needed;
  3229 
  3230 * HOL: the constant for "f``x" is now "image" rather than "op ``";
  3231 
  3232 * HOL: the constant for "f-``x" is now "vimage" rather than "op -``";
  3233 
  3234 * HOL: the disjoint sum is now "<+>" instead of "Plus"; the cartesian
  3235 product is now "<*>" instead of "Times"; the lexicographic product is
  3236 now "<*lex*>" instead of "**";
  3237 
  3238 * HOL: theory Sexp is now in HOL/Induct examples (it used to be part
  3239 of main HOL, but was unused); better use HOL's datatype package;
  3240 
  3241 * HOL: removed "symbols" syntax for constant "override" of theory Map;
  3242 the old syntax may be recovered as follows:
  3243 
  3244   syntax (symbols)
  3245     override  :: "('a ~=> 'b) => ('a ~=> 'b) => ('a ~=> 'b)"
  3246       (infixl "\\<oplus>" 100)
  3247 
  3248 * HOL/Real: "rabs" replaced by overloaded "abs" function;
  3249 
  3250 * HOL/ML: even fewer consts are declared as global (see theories Ord,
  3251 Lfp, Gfp, WF); this only affects ML packages that refer to const names
  3252 internally;
  3253 
  3254 * HOL and ZF: syntax for quotienting wrt an equivalence relation
  3255 changed from A/r to A//r;
  3256 
  3257 * ZF: new treatment of arithmetic (nat & int) may break some old
  3258 proofs;
  3259 
  3260 * Isar: renamed some attributes (RS -> THEN, simplify -> simplified,
  3261 rulify -> rule_format, elimify -> elim_format, ...);
  3262 
  3263 * Isar/Provers: intro/elim/dest attributes changed; renamed
  3264 intro/intro!/intro!! flags to intro!/intro/intro? (in most cases, one
  3265 should have to change intro!! to intro? only); replaced "delrule" by
  3266 "rule del";
  3267 
  3268 * Isar/HOL: renamed "intrs" to "intros" in inductive definitions;
  3269 
  3270 * Provers: strengthened force_tac by using new first_best_tac;
  3271 
  3272 * LaTeX document preparation: several changes of isabelle.sty (see
  3273 lib/texinputs);
  3274 
  3275 
  3276 *** Document preparation ***
  3277 
  3278 * formal comments (text blocks etc.) in new-style theories may now
  3279 contain antiquotations of thm/prop/term/typ/text to be presented
  3280 according to latex print mode; concrete syntax is like this:
  3281 @{term[show_types] "f(x) = a + x"};
  3282 
  3283 * isatool mkdir provides easy setup of Isabelle session directories,
  3284 including proper document sources;
  3285 
  3286 * generated LaTeX sources are now deleted after successful run
  3287 (isatool document -c); may retain a copy somewhere else via -D option
  3288 of isatool usedir;
  3289 
  3290 * isatool usedir -D now lets isatool latex -o sty update the Isabelle
  3291 style files, achieving self-contained LaTeX sources and simplifying
  3292 LaTeX debugging;
  3293 
  3294 * old-style theories now produce (crude) LaTeX output as well;
  3295 
  3296 * browser info session directories are now self-contained (may be put
  3297 on WWW server seperately); improved graphs of nested sessions; removed
  3298 graph for 'all sessions';
  3299 
  3300 * several improvements in isabelle style files; \isabellestyle{it}
  3301 produces fake math mode output; \isamarkupheader is now \section by
  3302 default; see lib/texinputs/isabelle.sty etc.;
  3303 
  3304 
  3305 *** Isar ***
  3306 
  3307 * Isar/Pure: local results and corresponding term bindings are now
  3308 subject to Hindley-Milner polymorphism (similar to ML); this
  3309 accommodates incremental type-inference very nicely;
  3310 
  3311 * Isar/Pure: new derived language element 'obtain' supports
  3312 generalized existence reasoning;
  3313 
  3314 * Isar/Pure: new calculational elements 'moreover' and 'ultimately'
  3315 support accumulation of results, without applying any rules yet;
  3316 useful to collect intermediate results without explicit name
  3317 references, and for use with transitivity rules with more than 2
  3318 premises;
  3319 
  3320 * Isar/Pure: scalable support for case-analysis type proofs: new
  3321 'case' language element refers to local contexts symbolically, as
  3322 produced by certain proof methods; internally, case names are attached
  3323 to theorems as "tags";
  3324 
  3325 * Isar/Pure: theory command 'hide' removes declarations from
  3326 class/type/const name spaces;
  3327 
  3328 * Isar/Pure: theory command 'defs' supports option "(overloaded)" to
  3329 indicate potential overloading;
  3330 
  3331 * Isar/Pure: changed syntax of local blocks from {{ }} to { };
  3332 
  3333 * Isar/Pure: syntax of sorts made 'inner', i.e. have to write
  3334 "{a,b,c}" instead of {a,b,c};
  3335 
  3336 * Isar/Pure now provides its own version of intro/elim/dest
  3337 attributes; useful for building new logics, but beware of confusion
  3338 with the version in Provers/classical;
  3339 
  3340 * Isar/Pure: the local context of (non-atomic) goals is provided via
  3341 case name 'antecedent';
  3342 
  3343 * Isar/Pure: removed obsolete 'transfer' attribute (transfer of thms
  3344 to the current context is now done automatically);
  3345 
  3346 * Isar/Pure: theory command 'method_setup' provides a simple interface
  3347 for definining proof methods in ML;
  3348 
  3349 * Isar/Provers: intro/elim/dest attributes changed; renamed
  3350 intro/intro!/intro!! flags to intro!/intro/intro? (INCOMPATIBILITY, in
  3351 most cases, one should have to change intro!! to intro? only);
  3352 replaced "delrule" by "rule del";
  3353 
  3354 * Isar/Provers: new 'hypsubst' method, plain 'subst' method and
  3355 'symmetric' attribute (the latter supercedes [RS sym]);
  3356 
  3357 * Isar/Provers: splitter support (via 'split' attribute and 'simp'
  3358 method modifier); 'simp' method: 'only:' modifier removes loopers as
  3359 well (including splits);
  3360 
  3361 * Isar/Provers: Simplifier and Classical methods now support all kind
  3362 of modifiers used in the past, including 'cong', 'iff', etc.
  3363 
  3364 * Isar/Provers: added 'fastsimp' and 'clarsimp' methods (combination
  3365 of Simplifier and Classical reasoner);
  3366 
  3367 * Isar/HOL: new proof method 'cases' and improved version of 'induct'
  3368 now support named cases; major packages (inductive, datatype, primrec,
  3369 recdef) support case names and properly name parameters;
  3370 
  3371 * Isar/HOL: new transitivity rules for substitution in inequalities --
  3372 monotonicity conditions are extracted to be proven at end of
  3373 calculations;
  3374 
  3375 * Isar/HOL: removed 'case_split' thm binding, should use 'cases' proof
  3376 method anyway;
  3377 
  3378 * Isar/HOL: removed old expand_if = split_if; theorems if_splits =
  3379 split_if split_if_asm; datatype package provides theorems foo.splits =
  3380 foo.split foo.split_asm for each datatype;
  3381 
  3382 * Isar/HOL: tuned inductive package, rename "intrs" to "intros"
  3383 (potential INCOMPATIBILITY), emulation of mk_cases feature for proof
  3384 scripts: new 'inductive_cases' command and 'ind_cases' method; (Note:
  3385 use "(cases (simplified))" method in proper proof texts);
  3386 
  3387 * Isar/HOL: added global 'arith_split' attribute for 'arith' method;
  3388 
  3389 * Isar: names of theorems etc. may be natural numbers as well;
  3390 
  3391 * Isar: 'pr' command: optional arguments for goals_limit and
  3392 ProofContext.prems_limit; no longer prints theory contexts, but only
  3393 proof states;
  3394 
  3395 * Isar: diagnostic commands 'pr', 'thm', 'prop', 'term', 'typ' admit
  3396 additional print modes to be specified; e.g. "pr(latex)" will print
  3397 proof state according to the Isabelle LaTeX style;
  3398 
  3399 * Isar: improved support for emulating tactic scripts, including proof
  3400 methods 'rule_tac' etc., 'cut_tac', 'thin_tac', 'subgoal_tac',
  3401 'rename_tac', 'rotate_tac', 'tactic', and 'case_tac' / 'induct_tac'
  3402 (for HOL datatypes);
  3403 
  3404 * Isar: simplified (more robust) goal selection of proof methods: 1st
  3405 goal, all goals, or explicit goal specifier (tactic emulation); thus
  3406 'proof method scripts' have to be in depth-first order;
  3407 
  3408 * Isar: tuned 'let' syntax: replaced 'as' keyword by 'and';
  3409 
  3410 * Isar: removed 'help' command, which hasn't been too helpful anyway;
  3411 should instead use individual commands for printing items
  3412 (print_commands, print_methods etc.);
  3413 
  3414 * Isar: added 'nothing' --- the empty list of theorems;
  3415 
  3416 
  3417 *** HOL ***
  3418 
  3419 * HOL/MicroJava: formalization of a fragment of Java, together with a
  3420 corresponding virtual machine and a specification of its bytecode
  3421 verifier and a lightweight bytecode verifier, including proofs of
  3422 type-safety; by Gerwin Klein, Tobias Nipkow, David von Oheimb, and
  3423 Cornelia Pusch (see also the homepage of project Bali at
  3424 http://isabelle.in.tum.de/Bali/);
  3425 
  3426 * HOL/Algebra: new theory of rings and univariate polynomials, by
  3427 Clemens Ballarin;
  3428 
  3429 * HOL/NumberTheory: fundamental Theorem of Arithmetic, Chinese
  3430 Remainder Theorem, Fermat/Euler Theorem, Wilson's Theorem, by Thomas M
  3431 Rasmussen;
  3432 
  3433 * HOL/Lattice: fundamental concepts of lattice theory and order
  3434 structures, including duals, properties of bounds versus algebraic
  3435 laws, lattice operations versus set-theoretic ones, the Knaster-Tarski
  3436 Theorem for complete lattices etc.; may also serve as a demonstration
  3437 for abstract algebraic reasoning using axiomatic type classes, and
  3438 mathematics-style proof in Isabelle/Isar; by Markus Wenzel;
  3439 
  3440 * HOL/Prolog: a (bare-bones) implementation of Lambda-Prolog, by David
  3441 von Oheimb;
  3442 
  3443 * HOL/IMPP: extension of IMP with local variables and mutually
  3444 recursive procedures, by David von Oheimb;
  3445 
  3446 * HOL/Lambda: converted into new-style theory and document;
  3447 
  3448 * HOL/ex/Multiquote: example of multiple nested quotations and
  3449 anti-quotations -- basically a generalized version of de-Bruijn
  3450 representation; very useful in avoiding lifting of operations;
  3451 
  3452 * HOL/record: added general record equality rule to simpset; fixed
  3453 select-update simplification procedure to handle extended records as
  3454 well; admit "r" as field name;
  3455 
  3456 * HOL: 0 is now overloaded over the new sort "zero", allowing its use with
  3457 other numeric types and also as the identity of groups, rings, etc.;
  3458 
  3459 * HOL: new axclass plus_ac0 for addition with the AC-laws and 0 as identity.
  3460 Types nat and int belong to this axclass;
  3461 
  3462 * HOL: greatly improved simplification involving numerals of type nat, int, real:
  3463    (i + #8 + j) = Suc k simplifies to  #7 + (i + j) = k
  3464    i*j + k + j*#3*i     simplifies to  #4*(i*j) + k
  3465   two terms #m*u and #n*u are replaced by #(m+n)*u
  3466     (where #m, #n and u can implicitly be 1; this is simproc combine_numerals)
  3467   and the term/formula #m*u+x ~~ #n*u+y simplifies simplifies to #(m-n)+x ~~ y
  3468     or x ~~ #(n-m)+y, where ~~ is one of = < <= or - (simproc cancel_numerals);
  3469 
  3470 * HOL: meson_tac is available (previously in ex/meson.ML); it is a
  3471 powerful prover for predicate logic but knows nothing of clasets; see
  3472 ex/mesontest.ML and ex/mesontest2.ML for example applications;
  3473 
  3474 * HOL: new version of "case_tac" subsumes both boolean case split and
  3475 "exhaust_tac" on datatypes; INCOMPATIBILITY: exhaust_tac no longer
  3476 exists, may define val exhaust_tac = case_tac for ad-hoc portability;
  3477 
  3478 * HOL: simplification no longer dives into case-expressions: only the
  3479 selector expression is simplified, but not the remaining arms; to
  3480 enable full simplification of case-expressions for datatype t, you may
  3481 remove t.weak_case_cong from the simpset, either globally (Delcongs
  3482 [thm"t.weak_case_cong"];) or locally (delcongs [...]).
  3483 
  3484 * HOL/recdef: the recursion equations generated by 'recdef' for
  3485 function 'f' are now called f.simps instead of f.rules; if all
  3486 termination conditions are proved automatically, these simplification
  3487 rules are added to the simpset, as in primrec; rules may be named
  3488 individually as well, resulting in a separate list of theorems for
  3489 each equation;
  3490 
  3491 * HOL/While is a new theory that provides a while-combinator. It
  3492 permits the definition of tail-recursive functions without the
  3493 provision of a termination measure. The latter is necessary once the
  3494 invariant proof rule for while is applied.
  3495 
  3496 * HOL: new (overloaded) notation for the set of elements below/above
  3497 some element: {..u}, {..u(}, {l..}, {)l..}. See theory SetInterval.
  3498 
  3499 * HOL: theorems impI, allI, ballI bound as "strip";
  3500 
  3501 * HOL: new tactic induct_thm_tac: thm -> string -> int -> tactic
  3502 induct_tac th "x1 ... xn" expects th to have a conclusion of the form
  3503 P v1 ... vn and abbreviates res_inst_tac [("v1","x1"),...,("vn","xn")] th;
  3504 
  3505 * HOL/Real: "rabs" replaced by overloaded "abs" function;
  3506 
  3507 * HOL: theory Sexp now in HOL/Induct examples (it used to be part of
  3508 main HOL, but was unused);
  3509 
  3510 * HOL: fewer consts declared as global (e.g. have to refer to
  3511 "Lfp.lfp" instead of "lfp" internally; affects ML packages only);
  3512 
  3513 * HOL: tuned AST representation of nested pairs, avoiding bogus output
  3514 in case of overlap with user translations (e.g. judgements over
  3515 tuples); (note that the underlying logical represenation is still
  3516 bogus);
  3517 
  3518 
  3519 *** ZF ***
  3520 
  3521 * ZF: simplification automatically cancels common terms in arithmetic
  3522 expressions over nat and int;
  3523 
  3524 * ZF: new treatment of nat to minimize type-checking: all operators
  3525 coerce their operands to a natural number using the function natify,
  3526 making the algebraic laws unconditional;
  3527 
  3528 * ZF: as above, for int: operators coerce their operands to an integer
  3529 using the function intify;
  3530 
  3531 * ZF: the integer library now contains many of the usual laws for the
  3532 orderings, including $<=, and monotonicity laws for $+ and $*;
  3533 
  3534 * ZF: new example ZF/ex/NatSum to demonstrate integer arithmetic
  3535 simplification;
  3536 
  3537 * FOL and ZF: AddIffs now available, giving theorems of the form P<->Q
  3538 to the simplifier and classical reasoner simultaneously;
  3539 
  3540 
  3541 *** General ***
  3542 
  3543 * Provers: blast_tac now handles actual object-logic rules as
  3544 assumptions; note that auto_tac uses blast_tac internally as well;
  3545 
  3546 * Provers: new functions rulify/rulify_no_asm: thm -> thm for turning
  3547 outer -->/All/Ball into ==>/!!; qed_spec_mp now uses rulify_no_asm;
  3548 
  3549 * Provers: delrules now handles destruct rules as well (no longer need
  3550 explicit make_elim);
  3551 
  3552 * Provers: Blast_tac now warns of and ignores "weak elimination rules" e.g.
  3553   [| inj ?f;          ?f ?x = ?f ?y; ?x = ?y ==> ?W |] ==> ?W
  3554 use instead the strong form,
  3555   [| inj ?f; ~ ?W ==> ?f ?x = ?f ?y; ?x = ?y ==> ?W |] ==> ?W
  3556 in HOL, FOL and ZF the function cla_make_elim will create such rules
  3557 from destruct-rules;
  3558 
  3559 * Provers: Simplifier.easy_setup provides a fast path to basic
  3560 Simplifier setup for new object-logics;
  3561 
  3562 * Pure: AST translation rules no longer require constant head on LHS;
  3563 
  3564 * Pure: improved name spaces: ambiguous output is qualified; support
  3565 for hiding of names;
  3566 
  3567 * system: smart setup of canonical ML_HOME, ISABELLE_INTERFACE, and
  3568 XSYMBOL_HOME; no longer need to do manual configuration in most
  3569 situations;
  3570 
  3571 * system: compression of ML heaps images may now be controlled via -c
  3572 option of isabelle and isatool usedir (currently only observed by
  3573 Poly/ML);
  3574 
  3575 * system: isatool installfonts may handle X-Symbol fonts as well (very
  3576 useful for remote X11);
  3577 
  3578 * system: provide TAGS file for Isabelle sources;
  3579 
  3580 * ML: infix 'OF' is a version of 'MRS' with more appropriate argument
  3581 order;
  3582 
  3583 * ML: renamed flags Syntax.trace_norm_ast to Syntax.trace_ast; global
  3584 timing flag supersedes proof_timing and Toplevel.trace;
  3585 
  3586 * ML: new combinators |>> and |>>> for incremental transformations
  3587 with secondary results (e.g. certain theory extensions):
  3588 
  3589 * ML: PureThy.add_defs gets additional argument to indicate potential
  3590 overloading (usually false);
  3591 
  3592 * ML: PureThy.add_thms/add_axioms/add_defs now return theorems as
  3593 results;
  3594 
  3595 
  3596 
  3597 New in Isabelle99 (October 1999)
  3598 --------------------------------
  3599 
  3600 *** Overview of INCOMPATIBILITIES (see below for more details) ***
  3601 
  3602 * HOL: The THEN and ELSE parts of conditional expressions (if P then x else y)
  3603 are no longer simplified.  (This allows the simplifier to unfold recursive
  3604 functional programs.)  To restore the old behaviour, declare
  3605 
  3606     Delcongs [if_weak_cong];
  3607 
  3608 * HOL: Removed the obsolete syntax "Compl A"; use -A for set
  3609 complement;
  3610 
  3611 * HOL: the predicate "inj" is now defined by translation to "inj_on";
  3612 
  3613 * HOL/datatype: mutual_induct_tac no longer exists --
  3614   use induct_tac "x_1 ... x_n" instead of mutual_induct_tac ["x_1", ..., "x_n"]
  3615 
  3616 * HOL/typedef: fixed type inference for representing set; type
  3617 arguments now have to occur explicitly on the rhs as type constraints;
  3618 
  3619 * ZF: The con_defs part of an inductive definition may no longer refer
  3620 to constants declared in the same theory;
  3621 
  3622 * HOL, ZF: the function mk_cases, generated by the inductive
  3623 definition package, has lost an argument.  To simplify its result, it
  3624 uses the default simpset instead of a supplied list of theorems.
  3625 
  3626 * HOL/List: the constructors of type list are now Nil and Cons;
  3627 
  3628 * Simplifier: the type of the infix ML functions
  3629         setSSolver addSSolver setSolver addSolver
  3630 is now  simpset * solver -> simpset  where `solver' is a new abstract type
  3631 for packaging solvers. A solver is created via
  3632         mk_solver: string -> (thm list -> int -> tactic) -> solver
  3633 where the string argument is only a comment.
  3634 
  3635 
  3636 *** Proof tools ***
  3637 
  3638 * Provers/Arith/fast_lin_arith.ML contains a functor for creating a
  3639 decision procedure for linear arithmetic. Currently it is used for
  3640 types `nat', `int', and `real' in HOL (see below); it can, should and
  3641 will be instantiated for other types and logics as well.
  3642 
  3643 * The simplifier now accepts rewrite rules with flexible heads, eg
  3644      hom ?f ==> ?f(?x+?y) = ?f ?x + ?f ?y
  3645   They are applied like any rule with a non-pattern lhs, i.e. by first-order
  3646   matching.
  3647 
  3648 
  3649 *** General ***
  3650 
  3651 * New Isabelle/Isar subsystem provides an alternative to traditional
  3652 tactical theorem proving; together with the ProofGeneral/isar user
  3653 interface it offers an interactive environment for developing human
  3654 readable proof documents (Isar == Intelligible semi-automated
  3655 reasoning); for further information see isatool doc isar-ref,
  3656 src/HOL/Isar_examples and http://isabelle.in.tum.de/Isar/
  3657 
  3658 * improved and simplified presentation of theories: better HTML markup
  3659 (including colors), graph views in several sizes; isatool usedir now
  3660 provides a proper interface for user theories (via -P option); actual
  3661 document preparation based on (PDF)LaTeX is available as well (for
  3662 new-style theories only); see isatool doc system for more information;
  3663 
  3664 * native support for Proof General, both for classic Isabelle and
  3665 Isabelle/Isar;
  3666 
  3667 * ML function thm_deps visualizes dependencies of theorems and lemmas,
  3668 using the graph browser tool;
  3669 
  3670 * Isabelle manuals now also available as PDF;
  3671 
  3672 * theory loader rewritten from scratch (may not be fully
  3673 bug-compatible); old loadpath variable has been replaced by show_path,
  3674 add_path, del_path, reset_path functions; new operations such as
  3675 update_thy, touch_thy, remove_thy, use/update_thy_only (see also
  3676 isatool doc ref);
  3677 
  3678 * improved isatool install: option -k creates KDE application icon,
  3679 option -p DIR installs standalone binaries;
  3680 
  3681 * added ML_PLATFORM setting (useful for cross-platform installations);
  3682 more robust handling of platform specific ML images for SML/NJ;
  3683 
  3684 * the settings environment is now statically scoped, i.e. it is never
  3685 created again in sub-processes invoked from isabelle, isatool, or
  3686 Isabelle;
  3687 
  3688 * path element specification '~~' refers to '$ISABELLE_HOME';
  3689 
  3690 * in locales, the "assumes" and "defines" parts may be omitted if
  3691 empty;
  3692 
  3693 * new print_mode "xsymbols" for extended symbol support (e.g. genuine
  3694 long arrows);
  3695 
  3696 * new print_mode "HTML";
  3697 
  3698 * new flag show_tags controls display of tags of theorems (which are
  3699 basically just comments that may be attached by some tools);
  3700 
  3701 * Isamode 2.6 requires patch to accomodate change of Isabelle font
  3702 mode and goal output format:
  3703 
  3704 diff -r Isamode-2.6/elisp/isa-load.el Isamode/elisp/isa-load.el
  3705 244c244
  3706 <       (list (isa-getenv "ISABELLE") "-msymbols" logic-name)
  3707 ---
  3708 >       (list (isa-getenv "ISABELLE") "-misabelle_font" "-msymbols" logic-name)
  3709 diff -r Isabelle-2.6/elisp/isa-proofstate.el Isamode/elisp/isa-proofstate.el
  3710 181c181
  3711 < (defconst proofstate-proofstart-regexp "^Level [0-9]+$"
  3712 ---
  3713 > (defconst proofstate-proofstart-regexp "^Level [0-9]+"
  3714 
  3715 * function bind_thms stores lists of theorems (cf. bind_thm);
  3716 
  3717 * new shorthand tactics ftac, eatac, datac, fatac;
  3718 
  3719 * qed (and friends) now accept "" as result name; in that case the
  3720 theorem is not stored, but proper checks and presentation of the
  3721 result still apply;
  3722 
  3723 * theorem database now also indexes constants "Trueprop", "all",
  3724 "==>", "=="; thus thms_containing, findI etc. may retrieve more rules;
  3725 
  3726 
  3727 *** HOL ***
  3728 
  3729 ** HOL arithmetic **
  3730 
  3731 * There are now decision procedures for linear arithmetic over nat and
  3732 int:
  3733 
  3734 1. arith_tac copes with arbitrary formulae involving `=', `<', `<=',
  3735 `+', `-', `Suc', `min', `max' and numerical constants; other subterms
  3736 are treated as atomic; subformulae not involving type `nat' or `int'
  3737 are ignored; quantified subformulae are ignored unless they are
  3738 positive universal or negative existential. The tactic has to be
  3739 invoked by hand and can be a little bit slow. In particular, the
  3740 running time is exponential in the number of occurrences of `min' and
  3741 `max', and `-' on `nat'.
  3742 
  3743 2. fast_arith_tac is a cut-down version of arith_tac: it only takes
  3744 (negated) (in)equalities among the premises and the conclusion into
  3745 account (i.e. no compound formulae) and does not know about `min' and
  3746 `max', and `-' on `nat'. It is fast and is used automatically by the
  3747 simplifier.
  3748 
  3749 NB: At the moment, these decision procedures do not cope with mixed
  3750 nat/int formulae where the two parts interact, such as `m < n ==>
  3751 int(m) < int(n)'.
  3752 
  3753 * HOL/Numeral provides a generic theory of numerals (encoded
  3754 efficiently as bit strings); setup for types nat/int/real is in place;
  3755 INCOMPATIBILITY: since numeral syntax is now polymorphic, rather than
  3756 int, existing theories and proof scripts may require a few additional
  3757 type constraints;
  3758 
  3759 * integer division and remainder can now be performed on constant
  3760 arguments;
  3761 
  3762 * many properties of integer multiplication, division and remainder
  3763 are now available;
  3764 
  3765 * An interface to the Stanford Validity Checker (SVC) is available through the
  3766 tactic svc_tac.  Propositional tautologies and theorems of linear arithmetic
  3767 are proved automatically.  SVC must be installed separately, and its results
  3768 must be TAKEN ON TRUST (Isabelle does not check the proofs, but tags any
  3769 invocation of the underlying oracle).  For SVC see
  3770   http://verify.stanford.edu/SVC
  3771 
  3772 * IsaMakefile: the HOL-Real target now builds an actual image;
  3773 
  3774 
  3775 ** HOL misc **
  3776 
  3777 * HOL/Real/HahnBanach: the Hahn-Banach theorem for real vector spaces
  3778 (in Isabelle/Isar) -- by Gertrud Bauer;
  3779 
  3780 * HOL/BCV: generic model of bytecode verification, i.e. data-flow
  3781 analysis for assembly languages with subtypes;
  3782 
  3783 * HOL/TLA (Lamport's Temporal Logic of Actions): major reorganization
  3784 -- avoids syntactic ambiguities and treats state, transition, and
  3785 temporal levels more uniformly; introduces INCOMPATIBILITIES due to
  3786 changed syntax and (many) tactics;
  3787 
  3788 * HOL/inductive: Now also handles more general introduction rules such
  3789   as "ALL y. (y, x) : r --> y : acc r ==> x : acc r"; monotonicity
  3790   theorems are now maintained within the theory (maintained via the
  3791   "mono" attribute);
  3792 
  3793 * HOL/datatype: Now also handles arbitrarily branching datatypes
  3794   (using function types) such as
  3795 
  3796   datatype 'a tree = Atom 'a | Branch "nat => 'a tree"
  3797 
  3798 * HOL/record: record_simproc (part of the default simpset) takes care
  3799 of selectors applied to updated records; record_split_tac is no longer
  3800 part of the default claset; update_defs may now be removed from the
  3801 simpset in many cases; COMPATIBILITY: old behavior achieved by
  3802 
  3803   claset_ref () := claset() addSWrapper record_split_wrapper;
  3804   Delsimprocs [record_simproc]
  3805 
  3806 * HOL/typedef: fixed type inference for representing set; type
  3807 arguments now have to occur explicitly on the rhs as type constraints;
  3808 
  3809 * HOL/recdef (TFL): 'congs' syntax now expects comma separated list of theorem
  3810 names rather than an ML expression;
  3811 
  3812 * HOL/defer_recdef (TFL): like recdef but the well-founded relation can be
  3813 supplied later.  Program schemes can be defined, such as
  3814     "While B C s = (if B s then While B C (C s) else s)"
  3815 where the well-founded relation can be chosen after B and C have been given.
  3816 
  3817 * HOL/List: the constructors of type list are now Nil and Cons;
  3818 INCOMPATIBILITY: while [] and infix # syntax is still there, of
  3819 course, ML tools referring to List.list.op # etc. have to be adapted;
  3820 
  3821 * HOL_quantifiers flag superseded by "HOL" print mode, which is
  3822 disabled by default; run isabelle with option -m HOL to get back to
  3823 the original Gordon/HOL-style output;
  3824 
  3825 * HOL/Ord.thy: new bounded quantifier syntax (input only): ALL x<y. P,
  3826 ALL x<=y. P, EX x<y. P, EX x<=y. P;
  3827 
  3828 * HOL basic syntax simplified (more orthogonal): all variants of
  3829 All/Ex now support plain / symbolic / HOL notation; plain syntax for
  3830 Eps operator is provided as well: "SOME x. P[x]";
  3831 
  3832 * HOL/Sum.thy: sum_case has been moved to HOL/Datatype;
  3833 
  3834 * HOL/Univ.thy: infix syntax <*>, <+>, <**>, <+> eliminated and made
  3835 thus available for user theories;
  3836 
  3837 * HOLCF/IOA/Sequents: renamed 'Cons' to 'Consq' to avoid clash with
  3838 HOL/List; hardly an INCOMPATIBILITY since '>>' syntax is used all the
  3839 time;
  3840 
  3841 * HOL: new tactic smp_tac: int -> int -> tactic, which applies spec
  3842 several times and then mp;
  3843 
  3844 
  3845 *** LK ***
  3846 
  3847 * the notation <<...>> is now available as a notation for sequences of
  3848 formulas;
  3849 
  3850 * the simplifier is now installed
  3851 
  3852 * the axiom system has been generalized (thanks to Soren Heilmann)
  3853 
  3854 * the classical reasoner now has a default rule database
  3855 
  3856 
  3857 *** ZF ***
  3858 
  3859 * new primrec section allows primitive recursive functions to be given
  3860 directly (as in HOL) over datatypes and the natural numbers;
  3861 
  3862 * new tactics induct_tac and exhaust_tac for induction (or case
  3863 analysis) over datatypes and the natural numbers;
  3864 
  3865 * the datatype declaration of type T now defines the recursor T_rec;
  3866 
  3867 * simplification automatically does freeness reasoning for datatype
  3868 constructors;
  3869 
  3870 * automatic type-inference, with AddTCs command to insert new
  3871 type-checking rules;
  3872 
  3873 * datatype introduction rules are now added as Safe Introduction rules
  3874 to the claset;
  3875 
  3876 * the syntax "if P then x else y" is now available in addition to
  3877 if(P,x,y);
  3878 
  3879 
  3880 *** Internal programming interfaces ***
  3881 
  3882 * tuned simplifier trace output; new flag debug_simp;
  3883 
  3884 * structures Vartab / Termtab (instances of TableFun) offer efficient
  3885 tables indexed by indexname_ord / term_ord (compatible with aconv);
  3886 
  3887 * AxClass.axclass_tac lost the theory argument;
  3888 
  3889 * tuned current_goals_markers semantics: begin / end goal avoids
  3890 printing empty lines;
  3891 
  3892 * removed prs and prs_fn hook, which was broken because it did not
  3893 include \n in its semantics, forcing writeln to add one
  3894 uncoditionally; replaced prs_fn by writeln_fn; consider std_output:
  3895 string -> unit if you really want to output text without newline;
  3896 
  3897 * Symbol.output subject to print mode; INCOMPATIBILITY: defaults to
  3898 plain output, interface builders may have to enable 'isabelle_font'
  3899 mode to get Isabelle font glyphs as before;
  3900 
  3901 * refined token_translation interface; INCOMPATIBILITY: output length
  3902 now of type real instead of int;
  3903 
  3904 * theory loader actions may be traced via new ThyInfo.add_hook
  3905 interface (see src/Pure/Thy/thy_info.ML); example application: keep
  3906 your own database of information attached to *whole* theories -- as
  3907 opposed to intra-theory data slots offered via TheoryDataFun;
  3908 
  3909 * proper handling of dangling sort hypotheses (at last!);
  3910 Thm.strip_shyps and Drule.strip_shyps_warning take care of removing
  3911 extra sort hypotheses that can be witnessed from the type signature;
  3912 the force_strip_shyps flag is gone, any remaining shyps are simply
  3913 left in the theorem (with a warning issued by strip_shyps_warning);
  3914 
  3915 
  3916 
  3917 New in Isabelle98-1 (October 1998)
  3918 ----------------------------------
  3919 
  3920 *** Overview of INCOMPATIBILITIES (see below for more details) ***
  3921 
  3922 * several changes of automated proof tools;
  3923 
  3924 * HOL: major changes to the inductive and datatype packages, including
  3925 some minor incompatibilities of theory syntax;
  3926 
  3927 * HOL: renamed r^-1 to 'converse' from 'inverse'; 'inj_onto' is now
  3928 called `inj_on';
  3929 
  3930 * HOL: removed duplicate thms in Arith:
  3931   less_imp_add_less  should be replaced by  trans_less_add1
  3932   le_imp_add_le      should be replaced by  trans_le_add1
  3933 
  3934 * HOL: unary minus is now overloaded (new type constraints may be
  3935 required);
  3936 
  3937 * HOL and ZF: unary minus for integers is now #- instead of #~.  In
  3938 ZF, expressions such as n#-1 must be changed to n#- 1, since #-1 is
  3939 now taken as an integer constant.
  3940 
  3941 * Pure: ML function 'theory_of' renamed to 'theory';
  3942 
  3943 
  3944 *** Proof tools ***
  3945 
  3946 * Simplifier:
  3947   1. Asm_full_simp_tac is now more aggressive.
  3948      1. It will sometimes reorient premises if that increases their power to
  3949         simplify.
  3950      2. It does no longer proceed strictly from left to right but may also
  3951         rotate premises to achieve further simplification.
  3952      For compatibility reasons there is now Asm_lr_simp_tac which is like the
  3953      old Asm_full_simp_tac in that it does not rotate premises.
  3954   2. The simplifier now knows a little bit about nat-arithmetic.
  3955 
  3956 * Classical reasoner: wrapper mechanism for the classical reasoner now
  3957 allows for selected deletion of wrappers, by introduction of names for
  3958 wrapper functionals.  This implies that addbefore, addSbefore,
  3959 addaltern, and addSaltern now take a pair (name, tactic) as argument,
  3960 and that adding two tactics with the same name overwrites the first
  3961 one (emitting a warning).
  3962   type wrapper = (int -> tactic) -> (int -> tactic)
  3963   setWrapper, setSWrapper, compWrapper and compSWrapper are replaced by
  3964   addWrapper, addSWrapper: claset * (string * wrapper) -> claset
  3965   delWrapper, delSWrapper: claset *  string            -> claset
  3966   getWrapper is renamed to appWrappers, getSWrapper to appSWrappers;
  3967 
  3968 * Classical reasoner: addbefore/addSbefore now have APPEND/ORELSE
  3969 semantics; addbefore now affects only the unsafe part of step_tac
  3970 etc.; this affects addss/auto_tac/force_tac, so EXISTING PROOFS MAY
  3971 FAIL, but proofs should be fixable easily, e.g. by replacing Auto_tac
  3972 by Force_tac;
  3973 
  3974 * Classical reasoner: setwrapper to setWrapper and compwrapper to
  3975 compWrapper; added safe wrapper (and access functions for it);
  3976 
  3977 * HOL/split_all_tac is now much faster and fails if there is nothing
  3978 to split.  Some EXISTING PROOFS MAY REQUIRE ADAPTION because the order
  3979 and the names of the automatically generated variables have changed.
  3980 split_all_tac has moved within claset() from unsafe wrappers to safe
  3981 wrappers, which means that !!-bound variables are split much more
  3982 aggressively, and safe_tac and clarify_tac now split such variables.
  3983 If this splitting is not appropriate, use delSWrapper "split_all_tac".
  3984 Note: the same holds for record_split_tac, which does the job of
  3985 split_all_tac for record fields.
  3986 
  3987 * HOL/Simplifier: Rewrite rules for case distinctions can now be added
  3988 permanently to the default simpset using Addsplits just like
  3989 Addsimps. They can be removed via Delsplits just like
  3990 Delsimps. Lower-case versions are also available.
  3991 
  3992 * HOL/Simplifier: The rule split_if is now part of the default
  3993 simpset. This means that the simplifier will eliminate all occurrences
  3994 of if-then-else in the conclusion of a goal. To prevent this, you can
  3995 either remove split_if completely from the default simpset by
  3996 `Delsplits [split_if]' or remove it in a specific call of the
  3997 simplifier using `... delsplits [split_if]'.  You can also add/delete
  3998 other case splitting rules to/from the default simpset: every datatype
  3999 generates suitable rules `split_t_case' and `split_t_case_asm' (where
  4000 t is the name of the datatype).
  4001 
  4002 * Classical reasoner / Simplifier combination: new force_tac (and
  4003 derivatives Force_tac, force) combines rewriting and classical
  4004 reasoning (and whatever other tools) similarly to auto_tac, but is
  4005 aimed to solve the given subgoal completely.
  4006 
  4007 
  4008 *** General ***
  4009 
  4010 * new top-level commands `Goal' and `Goalw' that improve upon `goal'
  4011 and `goalw': the theory is no longer needed as an explicit argument -
  4012 the current theory context is used; assumptions are no longer returned
  4013 at the ML-level unless one of them starts with ==> or !!; it is
  4014 recommended to convert to these new commands using isatool fixgoal
  4015 (backup your sources first!);
  4016 
  4017 * new top-level commands 'thm' and 'thms' for retrieving theorems from
  4018 the current theory context, and 'theory' to lookup stored theories;
  4019 
  4020 * new theory section 'locale' for declaring constants, assumptions and
  4021 definitions that have local scope;
  4022 
  4023 * new theory section 'nonterminals' for purely syntactic types;
  4024 
  4025 * new theory section 'setup' for generic ML setup functions
  4026 (e.g. package initialization);
  4027 
  4028 * the distribution now includes Isabelle icons: see
  4029 lib/logo/isabelle-{small,tiny}.xpm;
  4030 
  4031 * isatool install - install binaries with absolute references to
  4032 ISABELLE_HOME/bin;
  4033 
  4034 * isatool logo -- create instances of the Isabelle logo (as EPS);
  4035 
  4036 * print mode 'emacs' reserved for Isamode;
  4037 
  4038 * support multiple print (ast) translations per constant name;
  4039 
  4040 * theorems involving oracles are now printed with a suffixed [!];
  4041 
  4042 
  4043 *** HOL ***
  4044 
  4045 * there is now a tutorial on Isabelle/HOL (do 'isatool doc tutorial');
  4046 
  4047 * HOL/inductive package reorganized and improved: now supports mutual
  4048 definitions such as
  4049 
  4050   inductive EVEN ODD
  4051     intrs
  4052       null "0 : EVEN"
  4053       oddI "n : EVEN ==> Suc n : ODD"
  4054       evenI "n : ODD ==> Suc n : EVEN"
  4055 
  4056 new theorem list "elims" contains an elimination rule for each of the
  4057 recursive sets; inductive definitions now handle disjunctive premises
  4058 correctly (also ZF);
  4059 
  4060 INCOMPATIBILITIES: requires Inductive as an ancestor; component
  4061 "mutual_induct" no longer exists - the induction rule is always
  4062 contained in "induct";
  4063 
  4064 
  4065 * HOL/datatype package re-implemented and greatly improved: now
  4066 supports mutually recursive datatypes such as
  4067 
  4068   datatype
  4069     'a aexp = IF_THEN_ELSE ('a bexp) ('a aexp) ('a aexp)
  4070             | SUM ('a aexp) ('a aexp)
  4071             | DIFF ('a aexp) ('a aexp)
  4072             | NUM 'a
  4073   and
  4074     'a bexp = LESS ('a aexp) ('a aexp)
  4075             | AND ('a bexp) ('a bexp)
  4076             | OR ('a bexp) ('a bexp)
  4077 
  4078 as well as indirectly recursive datatypes such as
  4079 
  4080   datatype
  4081     ('a, 'b) term = Var 'a
  4082                   | App 'b ((('a, 'b) term) list)
  4083 
  4084 The new tactic  mutual_induct_tac [<var_1>, ..., <var_n>] i  performs
  4085 induction on mutually / indirectly recursive datatypes.
  4086 
  4087 Primrec equations are now stored in theory and can be accessed via
  4088 <function_name>.simps.
  4089 
  4090 INCOMPATIBILITIES:
  4091 
  4092   - Theories using datatypes must now have theory Datatype as an
  4093     ancestor.
  4094   - The specific <typename>.induct_tac no longer exists - use the
  4095     generic induct_tac instead.
  4096   - natE has been renamed to nat.exhaust - use exhaust_tac
  4097     instead of res_inst_tac ... natE. Note that the variable
  4098     names in nat.exhaust differ from the names in natE, this
  4099     may cause some "fragile" proofs to fail.
  4100   - The theorems split_<typename>_case and split_<typename>_case_asm
  4101     have been renamed to <typename>.split and <typename>.split_asm.
  4102   - Since default sorts of type variables are now handled correctly,
  4103     some datatype definitions may have to be annotated with explicit
  4104     sort constraints.
  4105   - Primrec definitions no longer require function name and type
  4106     of recursive argument.
  4107 
  4108 Consider using isatool fixdatatype to adapt your theories and proof
  4109 scripts to the new package (backup your sources first!).
  4110 
  4111 
  4112 * HOL/record package: considerably improved implementation; now
  4113 includes concrete syntax for record types, terms, updates; theorems
  4114 for surjective pairing and splitting !!-bound record variables; proof
  4115 support is as follows:
  4116 
  4117   1) standard conversions (selectors or updates applied to record
  4118 constructor terms) are part of the standard simpset;
  4119 
  4120   2) inject equations of the form ((x, y) = (x', y')) == x=x' & y=y' are
  4121 made part of standard simpset and claset via addIffs;
  4122 
  4123   3) a tactic for record field splitting (record_split_tac) is part of
  4124 the standard claset (addSWrapper);
  4125 
  4126 To get a better idea about these rules you may retrieve them via
  4127 something like 'thms "foo.simps"' or 'thms "foo.iffs"', where "foo" is
  4128 the name of your record type.
  4129 
  4130 The split tactic 3) conceptually simplifies by the following rule:
  4131 
  4132   "(!!x. PROP ?P x) == (!!a b. PROP ?P (a, b))"
  4133 
  4134 Thus any record variable that is bound by meta-all will automatically
  4135 blow up into some record constructor term, consequently the
  4136 simplifications of 1), 2) apply.  Thus force_tac, auto_tac etc. shall
  4137 solve record problems automatically.
  4138 
  4139 
  4140 * reorganized the main HOL image: HOL/Integ and String loaded by
  4141 default; theory Main includes everything;
  4142 
  4143 * automatic simplification of integer sums and comparisons, using cancellation;
  4144 
  4145 * added option_map_eq_Some and not_Some_eq to the default simpset and claset;
  4146 
  4147 * added disj_not1 = "(~P | Q) = (P --> Q)" to the default simpset;
  4148 
  4149 * many new identities for unions, intersections, set difference, etc.;
  4150 
  4151 * expand_if, expand_split, expand_sum_case and expand_nat_case are now
  4152 called split_if, split_split, split_sum_case and split_nat_case (to go
  4153 with add/delsplits);
  4154 
  4155 * HOL/Prod introduces simplification procedure unit_eq_proc rewriting
  4156 (?x::unit) = (); this is made part of the default simpset, which COULD
  4157 MAKE EXISTING PROOFS FAIL under rare circumstances (consider
  4158 'Delsimprocs [unit_eq_proc];' as last resort); also note that
  4159 unit_abs_eta_conv is added in order to counter the effect of
  4160 unit_eq_proc on (%u::unit. f u), replacing it by f rather than by
  4161 %u.f();
  4162 
  4163 * HOL/Fun INCOMPATIBILITY: `inj_onto' is now called `inj_on' (which
  4164 makes more sense);
  4165 
  4166 * HOL/Set INCOMPATIBILITY: rule `equals0D' is now a well-formed destruct rule;
  4167   It and 'sym RS equals0D' are now in the default  claset, giving automatic
  4168   disjointness reasoning but breaking a few old proofs.
  4169 
  4170 * HOL/Relation INCOMPATIBILITY: renamed the relational operator r^-1
  4171 to 'converse' from 'inverse' (for compatibility with ZF and some
  4172 literature);
  4173 
  4174 * HOL/recdef can now declare non-recursive functions, with {} supplied as
  4175 the well-founded relation;
  4176 
  4177 * HOL/Set INCOMPATIBILITY: the complement of set A is now written -A instead of
  4178     Compl A.  The "Compl" syntax remains available as input syntax for this
  4179     release ONLY.
  4180 
  4181 * HOL/Update: new theory of function updates:
  4182     f(a:=b) == %x. if x=a then b else f x
  4183 may also be iterated as in f(a:=b,c:=d,...);
  4184 
  4185 * HOL/Vimage: new theory for inverse image of a function, syntax f-``B;
  4186 
  4187 * HOL/List:
  4188   - new function list_update written xs[i:=v] that updates the i-th
  4189     list position. May also be iterated as in xs[i:=a,j:=b,...].
  4190   - new function `upt' written [i..j(] which generates the list
  4191     [i,i+1,...,j-1], i.e. the upper bound is excluded. To include the upper
  4192     bound write [i..j], which is a shorthand for [i..j+1(].
  4193   - new lexicographic orderings and corresponding wellfoundedness theorems.
  4194 
  4195 * HOL/Arith:
  4196   - removed 'pred' (predecessor) function;
  4197   - generalized some theorems about n-1;
  4198   - many new laws about "div" and "mod";
  4199   - new laws about greatest common divisors (see theory ex/Primes);
  4200 
  4201 * HOL/Relation: renamed the relational operator r^-1 "converse"
  4202 instead of "inverse";
  4203 
  4204 * HOL/Induct/Multiset: a theory of multisets, including the wellfoundedness
  4205   of the multiset ordering;
  4206 
  4207 * directory HOL/Real: a construction of the reals using Dedekind cuts
  4208   (not included by default);
  4209 
  4210 * directory HOL/UNITY: Chandy and Misra's UNITY formalism;
  4211 
  4212 * directory HOL/Hoare: a new version of Hoare logic which permits many-sorted
  4213   programs, i.e. different program variables may have different types.
  4214 
  4215 * calling (stac rew i) now fails if "rew" has no effect on the goal
  4216   [previously, this check worked only if the rewrite rule was unconditional]
  4217   Now rew can involve either definitions or equalities (either == or =).
  4218 
  4219 
  4220 *** ZF ***
  4221 
  4222 * theory Main includes everything; INCOMPATIBILITY: theory ZF.thy contains
  4223   only the theorems proved on ZF.ML;
  4224 
  4225 * ZF INCOMPATIBILITY: rule `equals0D' is now a well-formed destruct rule;
  4226   It and 'sym RS equals0D' are now in the default  claset, giving automatic
  4227   disjointness reasoning but breaking a few old proofs.
  4228 
  4229 * ZF/Update: new theory of function updates
  4230     with default rewrite rule  f(x:=y) ` z = if(z=x, y, f`z)
  4231   may also be iterated as in f(a:=b,c:=d,...);
  4232 
  4233 * in  let x=t in u(x), neither t nor u(x) has to be an FOL term.
  4234 
  4235 * calling (stac rew i) now fails if "rew" has no effect on the goal
  4236   [previously, this check worked only if the rewrite rule was unconditional]
  4237   Now rew can involve either definitions or equalities (either == or =).
  4238 
  4239 * case_tac provided for compatibility with HOL
  4240     (like the old excluded_middle_tac, but with subgoals swapped)
  4241 
  4242 
  4243 *** Internal programming interfaces ***
  4244 
  4245 * Pure: several new basic modules made available for general use, see
  4246 also src/Pure/README;
  4247 
  4248 * improved the theory data mechanism to support encapsulation (data
  4249 kind name replaced by private Object.kind, acting as authorization
  4250 key); new type-safe user interface via functor TheoryDataFun; generic
  4251 print_data function becomes basically useless;
  4252 
  4253 * removed global_names compatibility flag -- all theory declarations
  4254 are qualified by default;
  4255 
  4256 * module Pure/Syntax now offers quote / antiquote translation
  4257 functions (useful for Hoare logic etc. with implicit dependencies);
  4258 see HOL/ex/Antiquote for an example use;
  4259 
  4260 * Simplifier now offers conversions (asm_)(full_)rewrite: simpset ->
  4261 cterm -> thm;
  4262 
  4263 * new tactical CHANGED_GOAL for checking that a tactic modifies a
  4264 subgoal;
  4265 
  4266 * Display.print_goals function moved to Locale.print_goals;
  4267 
  4268 * standard print function for goals supports current_goals_markers
  4269 variable for marking begin of proof, end of proof, start of goal; the
  4270 default is ("", "", ""); setting current_goals_markers := ("<proof>",
  4271 "</proof>", "<goal>") causes SGML like tagged proof state printing,
  4272 for example;
  4273 
  4274 
  4275 
  4276 New in Isabelle98 (January 1998)
  4277 --------------------------------
  4278 
  4279 *** Overview of INCOMPATIBILITIES (see below for more details) ***
  4280 
  4281 * changed lexical syntax of terms / types: dots made part of long
  4282 identifiers, e.g. "%x.x" no longer possible, should be "%x. x";
  4283 
  4284 * simpset (and claset) reference variable replaced by functions
  4285 simpset / simpset_ref;
  4286 
  4287 * no longer supports theory aliases (via merge) and non-trivial
  4288 implicit merge of thms' signatures;
  4289 
  4290 * most internal names of constants changed due to qualified names;
  4291 
  4292 * changed Pure/Sequence interface (see Pure/seq.ML);
  4293 
  4294 
  4295 *** General Changes ***
  4296 
  4297 * hierachically structured name spaces (for consts, types, axms, thms
  4298 etc.); new lexical class 'longid' (e.g. Foo.bar.x) may render much of
  4299 old input syntactically incorrect (e.g. "%x.x"); COMPATIBILITY:
  4300 isatool fixdots ensures space after dots (e.g. "%x. x"); set
  4301 long_names for fully qualified output names; NOTE: ML programs
  4302 (special tactics, packages etc.) referring to internal names may have
  4303 to be adapted to cope with fully qualified names; in case of severe
  4304 backward campatibility problems try setting 'global_names' at compile
  4305 time to have enrything declared within a flat name space; one may also
  4306 fine tune name declarations in theories via the 'global' and 'local'
  4307 section;
  4308 
  4309 * reimplemented the implicit simpset and claset using the new anytype
  4310 data filed in signatures; references simpset:simpset ref etc. are
  4311 replaced by functions simpset:unit->simpset and
  4312 simpset_ref:unit->simpset ref; COMPATIBILITY: use isatool fixclasimp
  4313 to patch your ML files accordingly;
  4314 
  4315 * HTML output now includes theory graph data for display with Java
  4316 applet or isatool browser; data generated automatically via isatool
  4317 usedir (see -i option, ISABELLE_USEDIR_OPTIONS);
  4318 
  4319 * defs may now be conditional; improved rewrite_goals_tac to handle
  4320 conditional equations;
  4321 
  4322 * defs now admits additional type arguments, using TYPE('a) syntax;
  4323 
  4324 * theory aliases via merge (e.g. M=A+B+C) no longer supported, always
  4325 creates a new theory node; implicit merge of thms' signatures is
  4326 restricted to 'trivial' ones; COMPATIBILITY: one may have to use
  4327 transfer:theory->thm->thm in (rare) cases;
  4328 
  4329 * improved handling of draft signatures / theories; draft thms (and
  4330 ctyps, cterms) are automatically promoted to real ones;
  4331 
  4332 * slightly changed interfaces for oracles: admit many per theory, named
  4333 (e.g. oracle foo = mlfun), additional name argument for invoke_oracle;
  4334 
  4335 * print_goals: optional output of const types (set show_consts and
  4336 show_types);
  4337 
  4338 * improved output of warnings (###) and errors (***);
  4339 
  4340 * subgoal_tac displays a warning if the new subgoal has type variables;
  4341 
  4342 * removed old README and Makefiles;
  4343 
  4344 * replaced print_goals_ref hook by print_current_goals_fn and result_error_fn;
  4345 
  4346 * removed obsolete init_pps and init_database;
  4347 
  4348 * deleted the obsolete tactical STATE, which was declared by
  4349     fun STATE tacfun st = tacfun st st;
  4350 
  4351 * cd and use now support path variables, e.g. $ISABELLE_HOME, or ~
  4352 (which abbreviates $HOME);
  4353 
  4354 * changed Pure/Sequence interface (see Pure/seq.ML); COMPATIBILITY:
  4355 use isatool fixseq to adapt your ML programs (this works for fully
  4356 qualified references to the Sequence structure only!);
  4357 
  4358 * use_thy no longer requires writable current directory; it always
  4359 reloads .ML *and* .thy file, if either one is out of date;
  4360 
  4361 
  4362 *** Classical Reasoner ***
  4363 
  4364 * Clarify_tac, clarify_tac, clarify_step_tac, Clarify_step_tac: new
  4365 tactics that use classical reasoning to simplify a subgoal without
  4366 splitting it into several subgoals;
  4367 
  4368 * Safe_tac: like safe_tac but uses the default claset;
  4369 
  4370 
  4371 *** Simplifier ***
  4372 
  4373 * added simplification meta rules:
  4374     (asm_)(full_)simplify: simpset -> thm -> thm;
  4375 
  4376 * simplifier.ML no longer part of Pure -- has to be loaded by object
  4377 logics (again);
  4378 
  4379 * added prems argument to simplification procedures;
  4380 
  4381 * HOL, FOL, ZF: added infix function `addsplits':
  4382   instead of `<simpset> setloop (split_tac <thms>)'
  4383   you can simply write `<simpset> addsplits <thms>'
  4384 
  4385 
  4386 *** Syntax ***
  4387 
  4388 * TYPE('a) syntax for type reflection terms;
  4389 
  4390 * no longer handles consts with name "" -- declare as 'syntax' instead;
  4391 
  4392 * pretty printer: changed order of mixfix annotation preference (again!);
  4393 
  4394 * Pure: fixed idt/idts vs. pttrn/pttrns syntactic categories;
  4395 
  4396 
  4397 *** HOL ***
  4398 
  4399 * HOL: there is a new splitter `split_asm_tac' that can be used e.g.
  4400   with `addloop' of the simplifier to faciliate case splitting in premises.
  4401 
  4402 * HOL/TLA: Stephan Merz's formalization of Lamport's Temporal Logic of Actions;
  4403 
  4404 * HOL/Auth: new protocol proofs including some for the Internet
  4405   protocol TLS;
  4406 
  4407 * HOL/Map: new theory of `maps' a la VDM;
  4408 
  4409 * HOL/simplifier: simplification procedures nat_cancel_sums for
  4410 cancelling out common nat summands from =, <, <= (in)equalities, or
  4411 differences; simplification procedures nat_cancel_factor for
  4412 cancelling common factor from =, <, <= (in)equalities over natural
  4413 sums; nat_cancel contains both kinds of procedures, it is installed by
  4414 default in Arith.thy -- this COULD MAKE EXISTING PROOFS FAIL;
  4415 
  4416 * HOL/simplifier: terms of the form
  4417   `? x. P1(x) & ... & Pn(x) & x=t & Q1(x) & ... Qn(x)'  (or t=x)
  4418   are rewritten to
  4419   `P1(t) & ... & Pn(t) & Q1(t) & ... Qn(t)',
  4420   and those of the form
  4421   `! x. P1(x) & ... & Pn(x) & x=t & Q1(x) & ... Qn(x) --> R(x)'  (or t=x)
  4422   are rewritten to
  4423   `P1(t) & ... & Pn(t) & Q1(t) & ... Qn(t) --> R(t)',
  4424 
  4425 * HOL/datatype
  4426   Each datatype `t' now comes with a theorem `split_t_case' of the form
  4427 
  4428   P(t_case f1 ... fn x) =
  4429      ( (!y1 ... ym1. x = C1 y1 ... ym1 --> P(f1 y1 ... ym1)) &
  4430         ...
  4431        (!y1 ... ymn. x = Cn y1 ... ymn --> P(f1 y1 ... ymn))
  4432      )
  4433 
  4434   and a theorem `split_t_case_asm' of the form
  4435 
  4436   P(t_case f1 ... fn x) =
  4437     ~( (? y1 ... ym1. x = C1 y1 ... ym1 & ~P(f1 y1 ... ym1)) |
  4438         ...
  4439        (? y1 ... ymn. x = Cn y1 ... ymn & ~P(f1 y1 ... ymn))
  4440      )
  4441   which can be added to a simpset via `addsplits'. The existing theorems
  4442   expand_list_case and expand_option_case have been renamed to
  4443   split_list_case and split_option_case.
  4444 
  4445 * HOL/Arithmetic:
  4446   - `pred n' is automatically converted to `n-1'.
  4447     Users are strongly encouraged not to use `pred' any longer,
  4448     because it will disappear altogether at some point.
  4449   - Users are strongly encouraged to write "0 < n" rather than
  4450     "n ~= 0". Theorems and proof tools have been modified towards this
  4451     `standard'.
  4452 
  4453 * HOL/Lists:
  4454   the function "set_of_list" has been renamed "set" (and its theorems too);
  4455   the function "nth" now takes its arguments in the reverse order and
  4456   has acquired the infix notation "!" as in "xs!n".
  4457 
  4458 * HOL/Set: UNIV is now a constant and is no longer translated to Compl{};
  4459 
  4460 * HOL/Set: The operator (UN x.B x) now abbreviates (UN x:UNIV. B x) and its
  4461   specialist theorems (like UN1_I) are gone.  Similarly for (INT x.B x);
  4462 
  4463 * HOL/record: extensible records with schematic structural subtyping
  4464 (single inheritance); EXPERIMENTAL version demonstrating the encoding,
  4465 still lacks various theorems and concrete record syntax;
  4466 
  4467 
  4468 *** HOLCF ***
  4469 
  4470 * removed "axioms" and "generated by" sections;
  4471 
  4472 * replaced "ops" section by extended "consts" section, which is capable of
  4473   handling the continuous function space "->" directly;
  4474 
  4475 * domain package:
  4476   . proves theorems immediately and stores them in the theory,
  4477   . creates hierachical name space,
  4478   . now uses normal mixfix annotations (instead of cinfix...),
  4479   . minor changes to some names and values (for consistency),
  4480   . e.g. cases -> casedist, dists_eq -> dist_eqs, [take_lemma] -> take_lemmas,
  4481   . separator between mutual domain defs: changed "," to "and",
  4482   . improved handling of sort constraints;  now they have to
  4483     appear on the left-hand side of the equations only;
  4484 
  4485 * fixed LAM <x,y,zs>.b syntax;
  4486 
  4487 * added extended adm_tac to simplifier in HOLCF -- can now discharge
  4488 adm (%x. P (t x)), where P is chainfinite and t continuous;
  4489 
  4490 
  4491 *** FOL and ZF ***
  4492 
  4493 * FOL: there is a new splitter `split_asm_tac' that can be used e.g.
  4494   with `addloop' of the simplifier to faciliate case splitting in premises.
  4495 
  4496 * qed_spec_mp, qed_goal_spec_mp, qed_goalw_spec_mp are available, as
  4497 in HOL, they strip ALL and --> from proved theorems;
  4498 
  4499 
  4500 
  4501 New in Isabelle94-8 (May 1997)
  4502 ------------------------------
  4503 
  4504 *** General Changes ***
  4505 
  4506 * new utilities to build / run / maintain Isabelle etc. (in parts
  4507 still somewhat experimental); old Makefiles etc. still functional;
  4508 
  4509 * new 'Isabelle System Manual';
  4510 
  4511 * INSTALL text, together with ./configure and ./build scripts;
  4512 
  4513 * reimplemented type inference for greater efficiency, better error
  4514 messages and clean internal interface;
  4515 
  4516 * prlim command for dealing with lots of subgoals (an easier way of
  4517 setting goals_limit);
  4518 
  4519 
  4520 *** Syntax ***
  4521 
  4522 * supports alternative (named) syntax tables (parser and pretty
  4523 printer); internal interface is provided by add_modesyntax(_i);
  4524 
  4525 * Pure, FOL, ZF, HOL, HOLCF now support symbolic input and output; to
  4526 be used in conjunction with the Isabelle symbol font; uses the
  4527 "symbols" syntax table;
  4528 
  4529 * added token_translation interface (may translate name tokens in
  4530 arbitrary ways, dependent on their type (free, bound, tfree, ...) and
  4531 the current print_mode); IMPORTANT: user print translation functions
  4532 are responsible for marking newly introduced bounds
  4533 (Syntax.mark_boundT);
  4534 
  4535 * token translations for modes "xterm" and "xterm_color" that display
  4536 names in bold, underline etc. or colors (which requires a color
  4537 version of xterm);
  4538 
  4539 * infixes may now be declared with names independent of their syntax;
  4540 
  4541 * added typed_print_translation (like print_translation, but may
  4542 access type of constant);
  4543 
  4544 
  4545 *** Classical Reasoner ***
  4546 
  4547 Blast_tac: a new tactic!  It is often more powerful than fast_tac, but has
  4548 some limitations.  Blast_tac...
  4549   + ignores addss, addbefore, addafter; this restriction is intrinsic
  4550   + ignores elimination rules that don't have the correct format
  4551         (the conclusion MUST be a formula variable)
  4552   + ignores types, which can make HOL proofs fail
  4553   + rules must not require higher-order unification, e.g. apply_type in ZF
  4554     [message "Function Var's argument not a bound variable" relates to this]
  4555   + its proof strategy is more general but can actually be slower
  4556 
  4557 * substitution with equality assumptions no longer permutes other
  4558 assumptions;
  4559 
  4560 * minor changes in semantics of addafter (now called addaltern); renamed
  4561 setwrapper to setWrapper and compwrapper to compWrapper; added safe wrapper
  4562 (and access functions for it);
  4563 
  4564 * improved combination of classical reasoner and simplifier:
  4565   + functions for handling clasimpsets
  4566   + improvement of addss: now the simplifier is called _after_ the
  4567     safe steps.
  4568   + safe variant of addss called addSss: uses safe simplifications
  4569     _during_ the safe steps. It is more complete as it allows multiple
  4570     instantiations of unknowns (e.g. with slow_tac).
  4571 
  4572 *** Simplifier ***
  4573 
  4574 * added interface for simplification procedures (functions that
  4575 produce *proven* rewrite rules on the fly, depending on current
  4576 redex);
  4577 
  4578 * ordering on terms as parameter (used for ordered rewriting);
  4579 
  4580 * new functions delcongs, deleqcongs, and Delcongs. richer rep_ss;
  4581 
  4582 * the solver is now split into a safe and an unsafe part.
  4583 This should be invisible for the normal user, except that the
  4584 functions setsolver and addsolver have been renamed to setSolver and
  4585 addSolver; added safe_asm_full_simp_tac;
  4586 
  4587 
  4588 *** HOL ***
  4589 
  4590 * a generic induction tactic `induct_tac' which works for all datatypes and
  4591 also for type `nat';
  4592 
  4593 * a generic case distinction tactic `exhaust_tac' which works for all
  4594 datatypes and also for type `nat';
  4595 
  4596 * each datatype comes with a function `size';
  4597 
  4598 * patterns in case expressions allow tuple patterns as arguments to
  4599 constructors, for example `case x of [] => ... | (x,y,z)#ps => ...';
  4600 
  4601 * primrec now also works with type nat;
  4602 
  4603 * recdef: a new declaration form, allows general recursive functions to be
  4604 defined in theory files.  See HOL/ex/Fib, HOL/ex/Primes, HOL/Subst/Unify.
  4605 
  4606 * the constant for negation has been renamed from "not" to "Not" to
  4607 harmonize with FOL, ZF, LK, etc.;
  4608 
  4609 * HOL/ex/LFilter theory of a corecursive "filter" functional for
  4610 infinite lists;
  4611 
  4612 * HOL/Modelcheck demonstrates invocation of model checker oracle;
  4613 
  4614 * HOL/ex/Ring.thy declares cring_simp, which solves equational
  4615 problems in commutative rings, using axiomatic type classes for + and *;
  4616 
  4617 * more examples in HOL/MiniML and HOL/Auth;
  4618 
  4619 * more default rewrite rules for quantifiers, union/intersection;
  4620 
  4621 * a new constant `arbitrary == @x.False';
  4622 
  4623 * HOLCF/IOA replaces old HOL/IOA;
  4624 
  4625 * HOLCF changes: derived all rules and arities
  4626   + axiomatic type classes instead of classes
  4627   + typedef instead of faking type definitions
  4628   + eliminated the internal constants less_fun, less_cfun, UU_fun, UU_cfun etc.
  4629   + new axclasses cpo, chfin, flat with flat < chfin < pcpo < cpo < po
  4630   + eliminated the types void, one, tr
  4631   + use unit lift and bool lift (with translations) instead of one and tr
  4632   + eliminated blift from Lift3.thy (use Def instead of blift)
  4633   all eliminated rules are derived as theorems --> no visible changes ;
  4634 
  4635 
  4636 *** ZF ***
  4637 
  4638 * ZF now has Fast_tac, Simp_tac and Auto_tac.  Union_iff is a now a default
  4639 rewrite rule; this may affect some proofs.  eq_cs is gone but can be put back
  4640 as ZF_cs addSIs [equalityI];
  4641 
  4642 
  4643 
  4644 New in Isabelle94-7 (November 96)
  4645 ---------------------------------
  4646 
  4647 * allowing negative levels (as offsets) in prlev and choplev;
  4648 
  4649 * super-linear speedup for large simplifications;
  4650 
  4651 * FOL, ZF and HOL now use miniscoping: rewriting pushes
  4652 quantifications in as far as possible (COULD MAKE EXISTING PROOFS
  4653 FAIL); can suppress it using the command Delsimps (ex_simps @
  4654 all_simps); De Morgan laws are also now included, by default;
  4655 
  4656 * improved printing of ==>  :  ~:
  4657 
  4658 * new object-logic "Sequents" adds linear logic, while replacing LK
  4659 and Modal (thanks to Sara Kalvala);
  4660 
  4661 * HOL/Auth: correctness proofs for authentication protocols;
  4662 
  4663 * HOL: new auto_tac combines rewriting and classical reasoning (many
  4664 examples on HOL/Auth);
  4665 
  4666 * HOL: new command AddIffs for declaring theorems of the form P=Q to
  4667 the rewriter and classical reasoner simultaneously;
  4668 
  4669 * function uresult no longer returns theorems in "standard" format;
  4670 regain previous version by: val uresult = standard o uresult;
  4671 
  4672 
  4673 
  4674 New in Isabelle94-6
  4675 -------------------
  4676 
  4677 * oracles -- these establish an interface between Isabelle and trusted
  4678 external reasoners, which may deliver results as theorems;
  4679 
  4680 * proof objects (in particular record all uses of oracles);
  4681 
  4682 * Simp_tac, Fast_tac, etc. that refer to implicit simpset / claset;
  4683 
  4684 * "constdefs" section in theory files;
  4685 
  4686 * "primrec" section (HOL) no longer requires names;
  4687 
  4688 * internal type "tactic" now simply "thm -> thm Sequence.seq";
  4689 
  4690 
  4691 
  4692 New in Isabelle94-5
  4693 -------------------
  4694 
  4695 * reduced space requirements;
  4696 
  4697 * automatic HTML generation from theories;
  4698 
  4699 * theory files no longer require "..." (quotes) around most types;
  4700 
  4701 * new examples, including two proofs of the Church-Rosser theorem;
  4702 
  4703 * non-curried (1994) version of HOL is no longer distributed;
  4704 
  4705 
  4706 
  4707 New in Isabelle94-4
  4708 -------------------
  4709 
  4710 * greatly reduced space requirements;
  4711 
  4712 * theory files (.thy) no longer require \...\ escapes at line breaks;
  4713 
  4714 * searchable theorem database (see the section "Retrieving theorems" on
  4715 page 8 of the Reference Manual);
  4716 
  4717 * new examples, including Grabczewski's monumental case study of the
  4718 Axiom of Choice;
  4719 
  4720 * The previous version of HOL renamed to Old_HOL;
  4721 
  4722 * The new version of HOL (previously called CHOL) uses a curried syntax
  4723 for functions.  Application looks like f a b instead of f(a,b);
  4724 
  4725 * Mutually recursive inductive definitions finally work in HOL;
  4726 
  4727 * In ZF, pattern-matching on tuples is now available in all abstractions and
  4728 translates to the operator "split";
  4729 
  4730 
  4731 
  4732 New in Isabelle94-3
  4733 -------------------
  4734 
  4735 * new infix operator, addss, allowing the classical reasoner to
  4736 perform simplification at each step of its search.  Example:
  4737         fast_tac (cs addss ss)
  4738 
  4739 * a new logic, CHOL, the same as HOL, but with a curried syntax
  4740 for functions.  Application looks like f a b instead of f(a,b).  Also pairs
  4741 look like (a,b) instead of <a,b>;
  4742 
  4743 * PLEASE NOTE: CHOL will eventually replace HOL!
  4744 
  4745 * In CHOL, pattern-matching on tuples is now available in all abstractions.
  4746 It translates to the operator "split".  A new theory of integers is available;
  4747 
  4748 * In ZF, integer numerals now denote two's-complement binary integers.
  4749 Arithmetic operations can be performed by rewriting.  See ZF/ex/Bin.ML;
  4750 
  4751 * Many new examples: I/O automata, Church-Rosser theorem, equivalents
  4752 of the Axiom of Choice;
  4753 
  4754 
  4755 
  4756 New in Isabelle94-2
  4757 -------------------
  4758 
  4759 * Significantly faster resolution;
  4760 
  4761 * the different sections in a .thy file can now be mixed and repeated
  4762 freely;
  4763 
  4764 * Database of theorems for FOL, HOL and ZF.  New
  4765 commands including qed, qed_goal and bind_thm store theorems in the database.
  4766 
  4767 * Simple database queries: return a named theorem (get_thm) or all theorems of
  4768 a given theory (thms_of), or find out what theory a theorem was proved in
  4769 (theory_of_thm);
  4770 
  4771 * Bugs fixed in the inductive definition and datatype packages;
  4772 
  4773 * The classical reasoner provides deepen_tac and depth_tac, making FOL_dup_cs
  4774 and HOL_dup_cs obsolete;
  4775 
  4776 * Syntactic ambiguities caused by the new treatment of syntax in Isabelle94-1
  4777 have been removed;
  4778 
  4779 * Simpler definition of function space in ZF;
  4780 
  4781 * new results about cardinal and ordinal arithmetic in ZF;
  4782 
  4783 * 'subtype' facility in HOL for introducing new types as subsets of existing
  4784 types;
  4785 
  4786 
  4787 $Id$