NEWS
author wenzelm
Wed Nov 08 11:23:09 2006 +0100 (2006-11-08)
changeset 21242 d73735bb33c1
parent 21241 a00a16cbc647
child 21256 47195501ecf7
permissions -rw-r--r--
* November 2006: Lukas Bulwahn, TUM -- HOL/function: method "lexicographic_order".
     1 Isabelle NEWS -- history user-relevant changes
     2 ==============================================
     3 
     4 New in this Isabelle version
     5 ----------------------------
     6 
     7 *** General ***
     8 
     9 * Theory syntax: the header format ``theory A = B + C:'' has been
    10 discontinued in favour of ``theory A imports B C begin''.  Use isatool
    11 fixheaders to convert existing theory files.  INCOMPATIBILITY.
    12 
    13 * Theory syntax: the old non-Isar theory file format has been
    14 discontinued altogether.  Note that ML proof scripts may still be used
    15 with Isar theories; migration is usually quite simple with the ML
    16 function use_legacy_bindings.  INCOMPATIBILITY.
    17 
    18 * Theory syntax: some popular names (e.g. "class", "if", "fun") are
    19 now keywords.  INCOMPATIBILITY, use double quotes.
    20 
    21 * Legacy goal package: reduced interface to the bare minimum required
    22 to keep existing proof scripts running.  Most other user-level
    23 functions are now part of the OldGoals structure, which is *not* open
    24 by default (consider isatool expandshort before open OldGoals).
    25 Removed top_sg, prin, printyp, pprint_term/typ altogether, because
    26 these tend to cause confusion about the actual goal (!) context being
    27 used here, which is not necessarily the same as the_context().
    28 
    29 * Command 'find_theorems': support "*" wildcard in "name:" criterion.
    30 
    31 * The ``prems limit'' option (cf. ProofContext.prems_limit) is now -1
    32 by default, which means that "prems" (and also "fixed variables") are
    33 suppressed from proof state output.  Note that the ProofGeneral
    34 settings mechanism allows to change and save options persistently, but
    35 older versions of Isabelle will fail to start up if a negative prems
    36 limit is imposed.
    37 
    38 
    39 *** Document preparation ***
    40 
    41 * Added antiquotations @{ML_type text} and @{ML_struct text} which
    42 check the given source text as ML type/structure, printing verbatim.
    43 
    44 
    45 *** Pure ***
    46 
    47 * class_package.ML offers a combination of axclasses and locales to
    48 achieve Haskell-like type classes in Isabelle. See
    49 HOL/ex/Classpackage.thy for examples.
    50 
    51 * Yet another code generator framework allows to generate executable
    52 code for ML and Haskell (including "class"es). A short usage sketch:
    53 
    54     internal compilation:
    55         code_gen <list of constants (term syntax)> (SML *)
    56     writing SML code to a file:
    57         code_gen <list of constants (term syntax)> (SML <filename>)
    58     writing Haskell code to a bunch of files:
    59         code_gen <list of constants (term syntax)> (Haskell <filename>)
    60 
    61 Reasonable default setup of framework in HOL/Main.
    62 
    63 See HOL/ex/Code*.thy for examples.
    64 
    65 Theorem attributs for selecting and transforming function equations theorems:
    66 
    67     [code fun]:       select a theorem as function equation for a specific constant
    68     [code nofun]:     deselect a theorem as function equation for a specific constant
    69     [code inline]:    select an equation theorem for unfolding (inlining) in place
    70     [code noinline]:  deselect an equation theorem for unfolding (inlining) in place
    71 
    72 User-defined serializations (target in {SML, Haskell}):
    73 
    74     code_const <and-list of constants (term syntax)>
    75       {(target) <and-list of const target syntax>}+
    76 
    77     code_type <and-list of type constructors>
    78       {(target) <and-list of type target syntax>}+
    79 
    80     code_instance <and-list of instances>
    81       {(target)}+
    82         where instance ::= <type constructor> :: <class>
    83 
    84     code_class <and_list of classes>
    85       {(target) <and-list of class target syntax>}+
    86         where class target syntax ::= <class name> {where {<classop> == <target syntax>}+}?
    87 
    88 For code_instance and code_class, target SML is silently ignored.
    89 
    90 See HOL theories and HOL/ex/Code*.thy for usage examples.  Doc/Isar/Advanced/Codegen/
    91 provides a tutorial.
    92 
    93 * Command 'no_translations' removes translation rules from theory
    94 syntax.
    95 
    96 * Overloaded definitions are now actually checked for acyclic
    97 dependencies.  The overloading scheme is slightly more general than
    98 that of Haskell98, although Isabelle does not demand an exact
    99 correspondence to type class and instance declarations.
   100 INCOMPATIBILITY, use ``defs (unchecked overloaded)'' to admit more
   101 exotic versions of overloading -- at the discretion of the user!
   102 
   103 Polymorphic constants are represented via type arguments, i.e. the
   104 instantiation that matches an instance against the most general
   105 declaration given in the signature.  For example, with the declaration
   106 c :: 'a => 'a => 'a, an instance c :: nat => nat => nat is represented
   107 as c(nat).  Overloading is essentially simultaneous structural
   108 recursion over such type arguments.  Incomplete specification patterns
   109 impose global constraints on all occurrences, e.g. c('a * 'a) on the
   110 LHS means that more general c('a * 'b) will be disallowed on any RHS.
   111 Command 'print_theory' outputs the normalized system of recursive
   112 equations, see section "definitions".
   113 
   114 * Isar: improper proof element 'guess' is like 'obtain', but derives
   115 the obtained context from the course of reasoning!  For example:
   116 
   117   assume "EX x y. A x & B y"   -- "any previous fact"
   118   then guess x and y by clarify
   119 
   120 This technique is potentially adventurous, depending on the facts and
   121 proof tools being involved here.
   122 
   123 * Isar: known facts from the proof context may be specified as literal
   124 propositions, using ASCII back-quote syntax.  This works wherever
   125 named facts used to be allowed so far, in proof commands, proof
   126 methods, attributes etc.  Literal facts are retrieved from the context
   127 according to unification of type and term parameters.  For example,
   128 provided that "A" and "A ==> B" and "!!x. P x ==> Q x" are known
   129 theorems in the current context, then these are valid literal facts:
   130 `A` and `A ==> B` and `!!x. P x ==> Q x" as well as `P a ==> Q a` etc.
   131 
   132 There is also a proof method "fact" which does the same composition
   133 for explicit goal states, e.g. the following proof texts coincide with
   134 certain special cases of literal facts:
   135 
   136   have "A" by fact                 ==  note `A`
   137   have "A ==> B" by fact           ==  note `A ==> B`
   138   have "!!x. P x ==> Q x" by fact  ==  note `!!x. P x ==> Q x`
   139   have "P a ==> Q a" by fact       ==  note `P a ==> Q a`
   140 
   141 * Isar: ":" (colon) is no longer a symbolic identifier character in
   142 outer syntax.  Thus symbolic identifiers may be used without
   143 additional white space in declarations like this: ``assume *: A''.
   144 
   145 * Isar: 'print_facts' prints all local facts of the current context,
   146 both named and unnamed ones.
   147 
   148 * Isar: 'def' now admits simultaneous definitions, e.g.:
   149 
   150   def x == "t" and y == "u"
   151 
   152 * Isar: added command 'unfolding', which is structurally similar to
   153 'using', but affects both the goal state and facts by unfolding given
   154 rewrite rules.  Thus many occurrences of the 'unfold' method or
   155 'unfolded' attribute may be replaced by first-class proof text.
   156 
   157 * Isar: methods 'unfold' / 'fold', attributes 'unfolded' / 'folded',
   158 and command 'unfolding' now all support object-level equalities
   159 (potentially conditional).  The underlying notion of rewrite rule is
   160 analogous to the 'rule_format' attribute, but *not* that of the
   161 Simplifier (which is usually more generous).
   162 
   163 * Isar: the goal restriction operator [N] (default N = 1) evaluates a
   164 method expression within a sandbox consisting of the first N
   165 sub-goals, which need to exist.  For example, ``simp_all [3]''
   166 simplifies the first three sub-goals, while (rule foo, simp_all)[]
   167 simplifies all new goals that emerge from applying rule foo to the
   168 originally first one.
   169 
   170 * Isar: schematic goals are no longer restricted to higher-order
   171 patterns; e.g. ``lemma "?P(?x)" by (rule TrueI)'' now works as
   172 expected.
   173 
   174 * Isar: the conclusion of a long theorem statement is now either
   175 'shows' (a simultaneous conjunction, as before), or 'obtains'
   176 (essentially a disjunction of cases with local parameters and
   177 assumptions).  The latter allows to express general elimination rules
   178 adequately; in this notation common elimination rules look like this:
   179 
   180   lemma exE:    -- "EX x. P x ==> (!!x. P x ==> thesis) ==> thesis"
   181     assumes "EX x. P x"
   182     obtains x where "P x"
   183 
   184   lemma conjE:  -- "A & B ==> (A ==> B ==> thesis) ==> thesis"
   185     assumes "A & B"
   186     obtains A and B
   187 
   188   lemma disjE:  -- "A | B ==> (A ==> thesis) ==> (B ==> thesis) ==> thesis"
   189     assumes "A | B"
   190     obtains
   191       A
   192     | B
   193 
   194 The subsequent classical rules even refer to the formal "thesis"
   195 explicitly:
   196 
   197   lemma classical:     -- "(~ thesis ==> thesis) ==> thesis"
   198     obtains "~ thesis"
   199 
   200   lemma Peirce's_Law:  -- "((thesis ==> something) ==> thesis) ==> thesis"
   201     obtains "thesis ==> something"
   202 
   203 The actual proof of an 'obtains' statement is analogous to that of the
   204 Isar proof element 'obtain', only that there may be several cases.
   205 Optional case names may be specified in parentheses; these will be
   206 available both in the present proof and as annotations in the
   207 resulting rule, for later use with the 'cases' method (cf. attribute
   208 case_names).
   209 
   210 * Isar: 'print_statement' prints theorems from the current theory or
   211 proof context in long statement form, according to the syntax of a
   212 top-level lemma.
   213 
   214 * Isar: 'obtain' takes an optional case name for the local context
   215 introduction rule (default "that").
   216 
   217 * Isar: removed obsolete 'concl is' patterns.  INCOMPATIBILITY, use
   218 explicit (is "_ ==> ?foo") in the rare cases where this still happens
   219 to occur.
   220 
   221 * Pure: syntax "CONST name" produces a fully internalized constant
   222 according to the current context.  This is particularly useful for
   223 syntax translations that should refer to internal constant
   224 representations independently of name spaces.
   225 
   226 * Isar/locales: 'notation' provides a robust interface to the 'syntax'
   227 primitive that also works in a locale context (both for constants and
   228 fixed variables).  Type declaration and internal syntactic
   229 representation of given constants retrieved from the context.
   230 
   231 * Isar/locales: new derived specification elements 'axiomatization',
   232 'definition', 'abbreviation', which support type-inference, admit
   233 object-level specifications (equality, equivalence).  See also the
   234 isar-ref manual.  Examples:
   235 
   236   axiomatization
   237     eq  (infix "===" 50)
   238     where eq_refl: "x === x" and eq_subst: "x === y ==> P x ==> P y"
   239 
   240   definition
   241     "f x y = x + y + 1"
   242     "g x = f x x"
   243 
   244   abbreviation
   245     neq  (infix "=!=" 50)
   246     "x =!= y == ~ (x === y)"
   247 
   248 These specifications may be also used in a locale context.  Then the
   249 constants being introduced depend on certain fixed parameters, and the
   250 constant name is qualified by the locale base name.  An internal
   251 abbreviation takes care for convenient input and output, making the
   252 parameters implicit and using the original short name.  See also
   253 HOL/ex/Abstract_NAT.thy for an example of deriving polymorphic
   254 entities from a monomorphic theory.
   255 
   256 Presently, abbreviations are only available 'in' a target locale, but
   257 not inherited by general import expressions.  Also note that
   258 'abbreviation' may be used as a type-safe replacement for 'syntax' +
   259 'translations' in common applications.
   260 
   261 Concrete syntax is attached to specified constants in internal form,
   262 independently of name spaces.  The parse tree representation is
   263 slightly different -- use 'notation' instead of raw 'syntax', and
   264 'translations' with explicit "CONST" markup to accommodate this.
   265 
   266 * Isar/locales: improved parameter handling:
   267 - use of locales "var" and "struct" no longer necessary;
   268 - parameter renamings are no longer required to be injective.
   269   This enables, for example, to define a locale for endomorphisms thus:
   270   locale endom = homom mult mult h.
   271 
   272 * Isar/locales: changed the way locales with predicates are defined.
   273 Instead of accumulating the specification, the imported expression is
   274 now an interpretation.
   275 INCOMPATIBILITY: different normal form of locale expressions.
   276 In particular, in interpretations of locales with predicates,
   277 goals repesenting already interpreted fragments are not removed
   278 automatically.  Use methods `intro_locales' and `unfold_locales'; see below.
   279 
   280 * Isar/locales: new methods `intro_locales' and `unfold_locales' provide
   281 backward reasoning on locales predicates.  The methods are aware of
   282 interpretations and discharge corresponding goals.  `intro_locales' is
   283 less aggressive then `unfold_locales' and does not unfold predicates to
   284 assumptions.
   285 
   286 * Isar/locales: the order in which locale fragments are accumulated
   287 has changed.  This enables to override declarations from fragments
   288 due to interpretations -- for example, unwanted simp rules.
   289 
   290 * Provers/induct: improved internal context management to support
   291 local fixes and defines on-the-fly.  Thus explicit meta-level
   292 connectives !! and ==> are rarely required anymore in inductive goals
   293 (using object-logic connectives for this purpose has been long
   294 obsolete anyway).  The subsequent proof patterns illustrate advanced
   295 techniques of natural induction; general datatypes and inductive sets
   296 work analogously (see also src/HOL/Lambda for realistic examples).
   297 
   298 (1) This is how to ``strengthen'' an inductive goal wrt. certain
   299 parameters:
   300 
   301   lemma
   302     fixes n :: nat and x :: 'a
   303     assumes a: "A n x"
   304     shows "P n x"
   305     using a                     -- {* make induct insert fact a *}
   306   proof (induct n arbitrary: x) -- {* generalize goal to "!!x. A n x ==> P n x" *}
   307     case 0
   308     show ?case sorry
   309   next
   310     case (Suc n)
   311     note `!!x. A n x ==> P n x` -- {* induction hypothesis, according to induction rule *}
   312     note `A (Suc n) x`          -- {* induction premise, stemming from fact a *}
   313     show ?case sorry
   314   qed
   315 
   316 (2) This is how to perform induction over ``expressions of a certain
   317 form'', using a locally defined inductive parameter n == "a x"
   318 together with strengthening (the latter is usually required to get
   319 sufficiently flexible induction hypotheses):
   320 
   321   lemma
   322     fixes a :: "'a => nat"
   323     assumes a: "A (a x)"
   324     shows "P (a x)"
   325     using a
   326   proof (induct n == "a x" arbitrary: x)
   327     ...
   328 
   329 See also HOL/Isar_examples/Puzzle.thy for an application of the this
   330 particular technique.
   331 
   332 (3) This is how to perform existential reasoning ('obtains' or
   333 'obtain') by induction, while avoiding explicit object-logic
   334 encodings:
   335 
   336   lemma
   337     fixes n :: nat
   338     obtains x :: 'a where "P n x" and "Q n x"
   339   proof (induct n arbitrary: thesis)
   340     case 0
   341     obtain x where "P 0 x" and "Q 0 x" sorry
   342     then show thesis by (rule 0)
   343   next
   344     case (Suc n)
   345     obtain x where "P n x" and "Q n x" by (rule Suc.hyps)
   346     obtain x where "P (Suc n) x" and "Q (Suc n) x" sorry
   347     then show thesis by (rule Suc.prems)
   348   qed
   349 
   350 Here the 'arbitrary: thesis' specification essentially modifies the
   351 scope of the formal thesis parameter, in order to the get the whole
   352 existence statement through the induction as expected.
   353 
   354 * Provers/induct: mutual induction rules are now specified as a list
   355 of rule sharing the same induction cases.  HOL packages usually
   356 provide foo_bar.inducts for mutually defined items foo and bar
   357 (e.g. inductive sets or datatypes).  INCOMPATIBILITY, users need to
   358 specify mutual induction rules differently, i.e. like this:
   359 
   360   (induct rule: foo_bar.inducts)
   361   (induct set: foo bar)
   362   (induct type: foo bar)
   363 
   364 The ML function ProjectRule.projections turns old-style rules into the
   365 new format.
   366 
   367 * Provers/induct: improved handling of simultaneous goals.  Instead of
   368 introducing object-level conjunction, the statement is now split into
   369 several conclusions, while the corresponding symbolic cases are
   370 nested accordingly.  INCOMPATIBILITY, proofs need to be structured
   371 explicitly.  For example:
   372 
   373   lemma
   374     fixes n :: nat
   375     shows "P n" and "Q n"
   376   proof (induct n)
   377     case 0 case 1
   378     show "P 0" sorry
   379   next
   380     case 0 case 2
   381     show "Q 0" sorry
   382   next
   383     case (Suc n) case 1
   384     note `P n` and `Q n`
   385     show "P (Suc n)" sorry
   386   next
   387     case (Suc n) case 2
   388     note `P n` and `Q n`
   389     show "Q (Suc n)" sorry
   390   qed
   391 
   392 The split into subcases may be deferred as follows -- this is
   393 particularly relevant for goal statements with local premises.
   394 
   395   lemma
   396     fixes n :: nat
   397     shows "A n ==> P n" and "B n ==> Q n"
   398   proof (induct n)
   399     case 0
   400     {
   401       case 1
   402       note `A 0`
   403       show "P 0" sorry
   404     next
   405       case 2
   406       note `B 0`
   407       show "Q 0" sorry
   408     }
   409   next
   410     case (Suc n)
   411     note `A n ==> P n` and `B n ==> Q n`
   412     {
   413       case 1
   414       note `A (Suc n)`
   415       show "P (Suc n)" sorry
   416     next
   417       case 2
   418       note `B (Suc n)`
   419       show "Q (Suc n)" sorry
   420     }
   421   qed
   422 
   423 If simultaneous goals are to be used with mutual rules, the statement
   424 needs to be structured carefully as a two-level conjunction, using
   425 lists of propositions separated by 'and':
   426 
   427   lemma
   428     shows "a : A ==> P1 a"
   429           "a : A ==> P2 a"
   430       and "b : B ==> Q1 b"
   431           "b : B ==> Q2 b"
   432           "b : B ==> Q3 b"
   433   proof (induct set: A B)
   434 
   435 * Provers/induct: support coinduction as well.  See
   436 src/HOL/Library/Coinductive_List.thy for various examples.
   437 
   438 * Attribute "symmetric" produces result with standardized schematic
   439 variables (index 0).  Potential INCOMPATIBILITY.
   440 
   441 * Simplifier: by default the simplifier trace only shows top level rewrites
   442 now. That is, trace_simp_depth_limit is set to 1 by default. Thus there is
   443 less danger of being flooded by the trace. The trace indicates where parts
   444 have been suppressed.
   445   
   446 * Provers/classical: removed obsolete classical version of elim_format
   447 attribute; classical elim/dest rules are now treated uniformly when
   448 manipulating the claset.
   449 
   450 * Provers/classical: stricter checks to ensure that supplied intro,
   451 dest and elim rules are well-formed; dest and elim rules must have at
   452 least one premise.
   453 
   454 * Provers/classical: attributes dest/elim/intro take an optional
   455 weight argument for the rule (just as the Pure versions).  Weights are
   456 ignored by automated tools, but determine the search order of single
   457 rule steps.
   458 
   459 * Syntax: input syntax now supports dummy variable binding "%_. b",
   460 where the body does not mention the bound variable.  Note that dummy
   461 patterns implicitly depend on their context of bounds, which makes
   462 "{_. _}" match any set comprehension as expected.  Potential
   463 INCOMPATIBILITY -- parse translations need to cope with syntactic
   464 constant "_idtdummy" in the binding position.
   465 
   466 * Syntax: removed obsolete syntactic constant "_K" and its associated
   467 parse translation.  INCOMPATIBILITY -- use dummy abstraction instead,
   468 for example "A -> B" => "Pi A (%_. B)".
   469 
   470 * Pure: 'class_deps' command visualizes the subclass relation, using
   471 the graph browser tool.
   472 
   473 * Pure: 'print_theory' now suppresses entities with internal name
   474 (trailing "_") by default; use '!' option for full details.
   475 
   476 
   477 *** HOL ***
   478 
   479 * Automated termination proofs "by lexicographic_order" are now
   480 included in the abbreviated function command "fun". No explicit
   481 "termination" command is necessary anymore.
   482 INCOMPATIBILITY: If a "fun"-definition cannot be proved terminating by
   483 a lexicographic size order, then the command fails. Use the expanded
   484 version "function" for these cases.
   485 
   486 * New method "lexicographic_order" automatically synthesizes
   487 termination relations as lexicographic combinations of size measures. 
   488 Usage for (function package) termination proofs:
   489 
   490 termination 
   491 by lexicographic_order
   492 
   493 * Records: generalised field-update to take a function on the field 
   494 rather than the new value: r(|A := x|) is translated to A_update (K x) r
   495 The K-combinator that is internally used is called K_record.
   496 INCOMPATIBILITY: Usage of the plain update functions has to be
   497 adapted.
   498  
   499 * axclass "semiring_0" now contains annihilation axioms 
   500 ("x * 0 = 0","0 * x = 0"), which are required for a semiring. Richer 
   501 structures do not inherit from semiring_0 anymore, because this property 
   502 is a theorem there, not an axiom.
   503 INCOMPATIBILITY: In instances of semiring_0, there is more to prove, but 
   504 this is mostly trivial.
   505 
   506 * axclass "recpower" was generalized to arbitrary monoids, not just 
   507 commutative semirings.
   508 INCOMPATIBILITY: If you use recpower and need commutativity or a semiring
   509 property, add the corresponding classes.
   510 
   511 * Locale Lattic_Locales.partial_order changed (to achieve consistency with
   512   axclass order):
   513   - moved to Orderings.partial_order
   514   - additional parameter ``less''
   515   INCOMPATIBILITY.
   516 
   517 * Constant "List.list_all2" in List.thy now uses authentic syntax.
   518 INCOMPATIBILITY: translations containing list_all2 may go wrong. On Isar
   519 level, use abbreviations instead.
   520 
   521 * Constant "List.op mem" in List.thy now has proper name: "List.memberl"
   522 INCOMPATIBILITY: rarely occuring name references (e.g. ``List.op mem.simps'')
   523 require renaming (e.g. ``List.memberl.simps'').
   524 
   525 * Renamed constants in HOL.thy:
   526     0      ~> HOL.zero
   527     1      ~> HOL.one
   528 INCOMPATIBILITY: ML code directly refering to constant names may need adaption
   529 This in general only affects hand-written proof tactics, simprocs and so on.
   530 
   531 * New theory Code_Generator providing class 'eq' with constant 'eq',
   532 allowing for code generation with polymorphic equality.
   533 
   534 * Numeral syntax: type 'bin' which was a mere type copy of 'int' has been
   535 abandoned in favour of plain 'int'. INCOMPATIBILITY -- Significant changes
   536 for setting up numeral syntax for types:
   537 
   538   - new constants Numeral.pred and Numeral.succ instead
   539       of former Numeral.bin_pred and Numeral.bin_succ.
   540   - Use integer operations instead of bin_add, bin_mult and so on.
   541   - Numeral simplification theorems named Numeral.numeral_simps instead of Bin_simps.
   542   - ML structure Bin_Simprocs now named Int_Numeral_Base_Simprocs.
   543 
   544 See HOL/Integ/IntArith.thy for an example setup.
   545 
   546 * New top level command 'normal_form' computes the normal form of a term
   547 that may contain free variables. For example 'normal_form "rev[a,b,c]"'
   548 prints '[b,c,a]'. This command is suitable for heavy-duty computations
   549 because the functions are compiled to ML first.
   550 INCOMPATIBILITY: new keywords 'normal_form' must quoted when used as
   551 an identifier.
   552 
   553 * Alternative iff syntax "A <-> B" for equality on bool (with priority
   554 25 like -->); output depends on the "iff" print_mode, the default is
   555 "A = B" (with priority 50).
   556 
   557 * Renamed constants in HOL.thy and Orderings.thy:
   558     op +   ~> HOL.plus
   559     op -   ~> HOL.minus
   560     uminus ~> HOL.uminus
   561     op *   ~> HOL.times
   562     op <   ~> Orderings.less
   563     op <=  ~> Orderings.less_eq
   564 
   565 Adaptions may be required in the following cases:
   566 
   567 a) User-defined constants using any of the names "plus", "minus", "times",
   568 "less" or "less_eq". The standard syntax translations for "+", "-" and "*"
   569 may go wrong.
   570 INCOMPATIBILITY: use more specific names.
   571 
   572 b) Variables named "plus", "minus", "times", "less", "less_eq"
   573 INCOMPATIBILITY: use more specific names.
   574 
   575 c) Permutative equations (e.g. "a + b = b + a")
   576 Since the change of names also changes the order of terms, permutative
   577 rewrite rules may get applied in a different order. Experience shows that
   578 this is rarely the case (only two adaptions in the whole Isabelle
   579 distribution).
   580 INCOMPATIBILITY: rewrite proofs
   581 
   582 d) ML code directly refering to constant names
   583 This in general only affects hand-written proof tactics, simprocs and so on.
   584 INCOMPATIBILITY: grep your sourcecode and replace names.
   585 
   586 * "LEAST x:A. P" expands to "LEAST x. x:A & P" (input only).
   587 
   588 * Relation composition operator "op O" now has precedence 75 and binds
   589 stronger than union and intersection. INCOMPATIBILITY.
   590 
   591 * The old set interval syntax "{m..n(}" (and relatives) has been removed.
   592   Use "{m..<n}" (and relatives) instead.
   593 
   594 * In the context of the assumption "~(s = t)" the Simplifier rewrites
   595 "t = s" to False (by simproc "neq_simproc").  For backward
   596 compatibility this can be disabled by ML "reset use_neq_simproc".
   597 
   598 * "m dvd n" where m and n are numbers is evaluated to True/False by simp.
   599 
   600 * Theorem Cons_eq_map_conv no longer has attribute `simp'.
   601 
   602 * Theorem setsum_mult renamed to setsum_right_distrib.
   603 
   604 * Prefer ex1I over ex_ex1I in single-step reasoning, e.g. by the
   605 'rule' method.
   606 
   607 * Tactics 'sat' and 'satx' reimplemented, several improvements: goals
   608 no longer need to be stated as "<prems> ==> False", equivalences (i.e.
   609 "=" on type bool) are handled, variable names of the form "lit_<n>"
   610 are no longer reserved, significant speedup.
   611 
   612 * Tactics 'sat' and 'satx' can now replay MiniSat proof traces.  zChaff is
   613 still supported as well.
   614 
   615 * inductive and datatype: provide projections of mutual rules, bundled
   616 as foo_bar.inducts;
   617 
   618 * Library: theory Infinite_Set has been moved to the library.
   619 
   620 * Library: theory Accessible_Part has been move to main HOL.
   621 
   622 * Library: added theory Coinductive_List of potentially infinite lists
   623 as greatest fixed-point.
   624 
   625 * Library: added theory AssocList which implements (finite) maps as
   626 association lists.
   627 
   628 * New proof method "evaluation" for efficiently solving a goal (i.e. a
   629 boolean expression) by compiling it to ML. The goal is "proved" (via
   630 the oracle "Evaluation") if it evaluates to True.
   631 
   632 * Linear arithmetic now splits certain operators (e.g. min, max, abs)
   633 also when invoked by the simplifier.  This results in the simplifier
   634 being more powerful on arithmetic goals.  INCOMPATIBILITY.  Set
   635 fast_arith_split_limit to 0 to obtain the old behavior.
   636 
   637 * Support for hex (0x20) and binary (0b1001) numerals. 
   638 
   639 * New method: reify eqs (t), where eqs are equations for an
   640 interpretation I :: 'a list => 'b => 'c and t::'c is an optional
   641 parameter, computes a term s::'b and a list xs::'a list and proves the
   642 theorem I xs s = t. This is also known as reification or quoting. The
   643 resulting theorem is applied to the subgoal to substitute t with I xs
   644 s.  If t is omitted, the subgoal itself is reified.
   645 
   646 * New method: reflection corr_thm eqs (t). The parameters eqs and (t)
   647 are as explained above. corr_thm is a theorem for I vs (f t) = I vs t,
   648 where f is supposed to be a computable function (in the sense of code
   649 generattion). The method uses reify to compute s and xs as above then
   650 applies corr_thm and uses normalization by evaluation to "prove" f s =
   651 r and finally gets the theorem t = r, which is again applied to the
   652 subgoal. An Example is available in HOL/ex/ReflectionEx.thy.
   653 
   654 * Reflection: Automatic refification now handels binding, an example
   655 is available in HOL/ex/ReflectionEx.thy
   656 
   657 
   658 *** HOL-Algebra ***
   659 
   660 * Formalisation of ideals and the quotient construction over rings.
   661 
   662 * Order and lattice theory no longer based on records.
   663 INCOMPATIBILITY.
   664 
   665 * Method algebra is now set up via an attribute.  For examples see
   666 CRing.thy.  INCOMPATIBILITY: the method is now weaker on combinations
   667 of algebraic structures.
   668 
   669 * Renamed `CRing.thy' to `Ring.thy'.  INCOMPATIBILITY.
   670 
   671 
   672 *** HOL-Complex ***
   673 
   674 * Theory Real: new method ferrack implements quantifier elimination
   675 for linear arithmetic over the reals. The quantifier elimination
   676 feature is used only for decision, for compatibility with arith. This
   677 means a goal is either solved or left unchanged, no simplification.
   678 
   679 
   680 *** ML ***
   681 
   682 * Pure/table:
   683 
   684 Function `...tab.foldl` removed.
   685 INCOMPATIBILITY: use `...tabfold` instead
   686 
   687 * Pure/library:
   688 
   689 `gen_rem(s)` abandoned in favour of `remove` / `subtract`.
   690 INCOMPATIBILITY:
   691 rewrite "gen_rem eq (xs, x)" to "remove (eq o swap) x xs"
   692 rewrite "gen_rems eq (xs, ys)" to "subtract (eq o swap) ys xs"
   693 drop "swap" if "eq" is symmetric.
   694 
   695 * Pure/library:
   696 
   697 infixes `ins` `ins_string` `ins_int` have been abandoned in favour of `insert`.
   698 INCOMPATIBILITY: rewrite "x ins(_...) xs" to "insert (op =) x xs"
   699 
   700 * Pure/General/susp.ML:
   701 
   702 New module for delayed evaluations.
   703 
   704 * Pure/library:
   705 
   706 Semantically identical functions "equal_list" and "eq_list" have been
   707 unified to "eq_list".
   708 
   709 * Pure/library:
   710 
   711   val burrow: ('a list -> 'b list) -> 'a list list -> 'b list list
   712   val fold_burrow: ('a list -> 'c -> 'b list * 'd) -> 'a list list -> 'c -> 'b list list * 'd
   713 
   714 The semantics of "burrow" is: "take a function with *simulatanously*
   715 transforms a list of value, and apply it *simulatanously* to a list of
   716 list of values of the appropriate type". Confer this with "map" which
   717 would *not* apply its argument function simulatanously but in
   718 sequence. "fold_burrow" has an additional context.
   719 
   720 Both actually avoid the usage of "unflat" since they hide away
   721 "unflat" from the user.
   722 
   723 * Pure/library: functions map2 and fold2 with curried syntax for
   724 simultanous mapping and folding:
   725 
   726     val map2: ('a -> 'b -> 'c) -> 'a list -> 'b list -> 'c list
   727     val fold2: ('a -> 'b -> 'c -> 'c) -> 'a list -> 'b list -> 'c -> 'c
   728 
   729 * Pure/library: indexed lists - some functions in the Isabelle library
   730 treating lists over 'a as finite mappings from [0...n] to 'a have been
   731 given more convenient names and signatures reminiscent of similar
   732 functions for alists, tables, etc:
   733 
   734   val nth: 'a list -> int -> 'a 
   735   val nth_map: int -> ('a -> 'a) -> 'a list -> 'a list
   736   val fold_index: (int * 'a -> 'b -> 'b) -> 'a list -> 'b -> 'b
   737 
   738 Note that fold_index starts counting at index 0, not 1 like foldln
   739 used to.
   740 
   741 * Pure/library: general ``divide_and_conquer'' combinator on lists.
   742 
   743 * Pure/General/name_mangler.ML provides a functor for generic name
   744 mangling (bijective mapping from expression values to strings).
   745 
   746 * Pure/General/rat.ML implements rational numbers.
   747 
   748 * Pure/General/table.ML: the join operations now works via exceptions
   749 DUP/SAME instead of type option.  This is simpler in simple cases, and
   750 admits slightly more efficient complex applications.
   751 
   752 * Pure: datatype Context.generic joins theory/Proof.context and
   753 provides some facilities for code that works in either kind of
   754 context, notably GenericDataFun for uniform theory and proof data.
   755 
   756 * Pure: 'advanced' translation functions (parse_translation etc.) now
   757 use Context.generic instead of just theory.
   758 
   759 * Pure: simplified internal attribute type, which is now always
   760 Context.generic * thm -> Context.generic * thm.  Global (theory)
   761 vs. local (Proof.context) attributes have been discontinued, while
   762 minimizing code duplication.  Thm.rule_attribute and
   763 Thm.declaration_attribute build canonical attributes; see also
   764 structure Context for further operations on Context.generic, notably
   765 GenericDataFun.  INCOMPATIBILITY, need to adapt attribute type
   766 declarations and definitions.
   767 
   768 * Pure/kernel: consts certification ignores sort constraints given in
   769 signature declarations.  (This information is not relevant to the
   770 logic, but only for type inference.)  IMPORTANT INTERNAL CHANGE.
   771 
   772 * Pure: axiomatic type classes are now purely definitional, with
   773 explicit proofs of class axioms and super class relations performed
   774 internally.  See Pure/axclass.ML for the main internal interfaces --
   775 notably AxClass.define_class supercedes AxClass.add_axclass, and
   776 AxClass.axiomatize_class/classrel/arity supercede
   777 Sign.add_classes/classrel/arities.
   778 
   779 * Pure/Isar: Args/Attrib parsers operate on Context.generic --
   780 global/local versions on theory vs. Proof.context have been
   781 discontinued; Attrib.syntax and Method.syntax have been adapted
   782 accordingly.  INCOMPATIBILITY, need to adapt parser expressions for
   783 attributes, methods, etc.
   784 
   785 * Pure: several functions of signature "... -> theory -> theory * ..."
   786 have been reoriented to "... -> theory -> ... * theory" in order to
   787 allow natural usage in combination with the ||>, ||>>, |-> and
   788 fold_map combinators.
   789 
   790 * Pure: primitive rule lift_rule now takes goal cterm instead of an
   791 actual goal state (thm).  Use Thm.lift_rule (Thm.cprem_of st i) to
   792 achieve the old behaviour.
   793 
   794 * Pure: the "Goal" constant is now called "prop", supporting a
   795 slightly more general idea of ``protecting'' meta-level rule
   796 statements.
   797 
   798 * Pure: Logic.(un)varify only works in a global context, which is now
   799 enforced instead of silently assumed.  INCOMPATIBILITY, may use
   800 Logic.legacy_(un)varify as temporary workaround.
   801 
   802 * Pure: structure Name provides scalable operations for generating
   803 internal variable names, notably Name.variants etc.  This replaces
   804 some popular functions from term.ML:
   805 
   806   Term.variant		->  Name.variant
   807   Term.variantlist	->  Name.variant_list  (*canonical argument order*)
   808   Term.invent_names	->  Name.invent_list
   809 
   810 Note that low-level renaming rarely occurs in new code -- operations
   811 from structure Variable are used instead (see below).
   812 
   813 * Pure: structure Variable provides fundamental operations for proper
   814 treatment of fixed/schematic variables in a context.  For example,
   815 Variable.import introduces fixes for schematics of given facts and
   816 Variable.export reverses the effect (up to renaming) -- this replaces
   817 various freeze_thaw operations.
   818 
   819 * Pure: structure Goal provides simple interfaces for
   820 init/conclude/finish and tactical prove operations (replacing former
   821 Tactic.prove).  Goal.prove is the canonical way to prove results
   822 within a given context; Goal.prove_global is a degraded version for
   823 theory level goals, including a global Drule.standard.  Note that
   824 OldGoals.prove_goalw_cterm has long been obsolete, since it is
   825 ill-behaved in a local proof context (e.g. with local fixes/assumes or
   826 in a locale context).
   827 
   828 * Isar: simplified treatment of user-level errors, using exception
   829 ERROR of string uniformly.  Function error now merely raises ERROR,
   830 without any side effect on output channels.  The Isar toplevel takes
   831 care of proper display of ERROR exceptions.  ML code may use plain
   832 handle/can/try; cat_error may be used to concatenate errors like this:
   833 
   834   ... handle ERROR msg => cat_error msg "..."
   835 
   836 Toplevel ML code (run directly or through the Isar toplevel) may be
   837 embedded into the Isar toplevel with exception display/debug like
   838 this:
   839 
   840   Isar.toplevel (fn () => ...)
   841 
   842 INCOMPATIBILITY, removed special transform_error facilities, removed
   843 obsolete variants of user-level exceptions (ERROR_MESSAGE,
   844 Context.PROOF, ProofContext.CONTEXT, Proof.STATE, ProofHistory.FAIL)
   845 -- use plain ERROR instead.
   846 
   847 * Isar: theory setup now has type (theory -> theory), instead of a
   848 list.  INCOMPATIBILITY, may use #> to compose setup functions.
   849 
   850 * Isar: installed ML toplevel pretty printer for type Proof.context,
   851 subject to ProofContext.debug/verbose flags.
   852 
   853 * Isar: Toplevel.theory_to_proof admits transactions that modify the
   854 theory before entering a proof state.  Transactions now always see a
   855 quasi-functional intermediate checkpoint, both in interactive and
   856 batch mode.
   857 
   858 * Simplifier: the simpset of a running simplification process now
   859 contains a proof context (cf. Simplifier.the_context), which is the
   860 very context that the initial simpset has been retrieved from (by
   861 simpset_of/local_simpset_of).  Consequently, all plug-in components
   862 (solver, looper etc.) may depend on arbitrary proof data.
   863 
   864 * Simplifier.inherit_context inherits the proof context (plus the
   865 local bounds) of the current simplification process; any simproc
   866 etc. that calls the Simplifier recursively should do this!  Removed
   867 former Simplifier.inherit_bounds, which is already included here --
   868 INCOMPATIBILITY.  Tools based on low-level rewriting may even have to
   869 specify an explicit context using Simplifier.context/theory_context.
   870 
   871 * Simplifier/Classical Reasoner: more abstract interfaces
   872 change_simpset/claset for modifying the simpset/claset reference of a
   873 theory; raw versions simpset/claset_ref etc. have been discontinued --
   874 INCOMPATIBILITY.
   875 
   876 * Provers: more generic wrt. syntax of object-logics, avoid hardwired
   877 "Trueprop" etc.
   878 
   879 
   880 *** System ***
   881 
   882 * isabelle-process: option -S (secure mode) disables some critical
   883 operations, notably runtime compilation and evaluation of ML source
   884 code.
   885 
   886 
   887 New in Isabelle2005 (October 2005)
   888 ----------------------------------
   889 
   890 *** General ***
   891 
   892 * Theory headers: the new header syntax for Isar theories is
   893 
   894   theory <name>
   895   imports <theory1> ... <theoryN>
   896   uses <file1> ... <fileM>
   897   begin
   898 
   899 where the 'uses' part is optional.  The previous syntax
   900 
   901   theory <name> = <theory1> + ... + <theoryN>:
   902 
   903 will disappear in the next release.  Use isatool fixheaders to convert
   904 existing theory files.  Note that there is no change in ancient
   905 non-Isar theories now, but these will disappear soon.
   906 
   907 * Theory loader: parent theories can now also be referred to via
   908 relative and absolute paths.
   909 
   910 * Command 'find_theorems' searches for a list of criteria instead of a
   911 list of constants. Known criteria are: intro, elim, dest, name:string,
   912 simp:term, and any term. Criteria can be preceded by '-' to select
   913 theorems that do not match. Intro, elim, dest select theorems that
   914 match the current goal, name:s selects theorems whose fully qualified
   915 name contain s, and simp:term selects all simplification rules whose
   916 lhs match term.  Any other term is interpreted as pattern and selects
   917 all theorems matching the pattern. Available in ProofGeneral under
   918 'ProofGeneral -> Find Theorems' or C-c C-f.  Example:
   919 
   920   C-c C-f (100) "(_::nat) + _ + _" intro -name: "HOL."
   921 
   922 prints the last 100 theorems matching the pattern "(_::nat) + _ + _",
   923 matching the current goal as introduction rule and not having "HOL."
   924 in their name (i.e. not being defined in theory HOL).
   925 
   926 * Command 'thms_containing' has been discontinued in favour of
   927 'find_theorems'; INCOMPATIBILITY.
   928 
   929 * Communication with Proof General is now 8bit clean, which means that
   930 Unicode text in UTF-8 encoding may be used within theory texts (both
   931 formal and informal parts).  Cf. option -U of the Isabelle Proof
   932 General interface.  Here are some simple examples (cf. src/HOL/ex):
   933 
   934   http://isabelle.in.tum.de/library/HOL/ex/Hebrew.html
   935   http://isabelle.in.tum.de/library/HOL/ex/Chinese.html
   936 
   937 * Improved efficiency of the Simplifier and, to a lesser degree, the
   938 Classical Reasoner.  Typical big applications run around 2 times
   939 faster.
   940 
   941 
   942 *** Document preparation ***
   943 
   944 * Commands 'display_drafts' and 'print_drafts' perform simple output
   945 of raw sources.  Only those symbols that do not require additional
   946 LaTeX packages (depending on comments in isabellesym.sty) are
   947 displayed properly, everything else is left verbatim.  isatool display
   948 and isatool print are used as front ends (these are subject to the
   949 DVI/PDF_VIEWER and PRINT_COMMAND settings, respectively).
   950 
   951 * Command tags control specific markup of certain regions of text,
   952 notably folding and hiding.  Predefined tags include "theory" (for
   953 theory begin and end), "proof" for proof commands, and "ML" for
   954 commands involving ML code; the additional tags "visible" and
   955 "invisible" are unused by default.  Users may give explicit tag
   956 specifications in the text, e.g. ''by %invisible (auto)''.  The
   957 interpretation of tags is determined by the LaTeX job during document
   958 preparation: see option -V of isatool usedir, or options -n and -t of
   959 isatool document, or even the LaTeX macros \isakeeptag, \isafoldtag,
   960 \isadroptag.
   961 
   962 Several document versions may be produced at the same time via isatool
   963 usedir (the generated index.html will link all of them).  Typical
   964 specifications include ''-V document=theory,proof,ML'' to present
   965 theory/proof/ML parts faithfully, ''-V outline=/proof,/ML'' to fold
   966 proof and ML commands, and ''-V mutilated=-theory,-proof,-ML'' to omit
   967 these parts without any formal replacement text.  The Isabelle site
   968 default settings produce ''document'' and ''outline'' versions as
   969 specified above.
   970 
   971 * Several new antiquotations:
   972 
   973   @{term_type term} prints a term with its type annotated;
   974 
   975   @{typeof term} prints the type of a term;
   976 
   977   @{const const} is the same as @{term const}, but checks that the
   978   argument is a known logical constant;
   979 
   980   @{term_style style term} and @{thm_style style thm} print a term or
   981   theorem applying a "style" to it
   982 
   983   @{ML text}
   984 
   985 Predefined styles are 'lhs' and 'rhs' printing the lhs/rhs of
   986 definitions, equations, inequations etc., 'concl' printing only the
   987 conclusion of a meta-logical statement theorem, and 'prem1' .. 'prem19'
   988 to print the specified premise.  TermStyle.add_style provides an ML
   989 interface for introducing further styles.  See also the "LaTeX Sugar"
   990 document practical applications.  The ML antiquotation prints
   991 type-checked ML expressions verbatim.
   992 
   993 * Markup commands 'chapter', 'section', 'subsection', 'subsubsection',
   994 and 'text' support optional locale specification '(in loc)', which
   995 specifies the default context for interpreting antiquotations.  For
   996 example: 'text (in lattice) {* @{thm inf_assoc}*}'.
   997 
   998 * Option 'locale=NAME' of antiquotations specifies an alternative
   999 context interpreting the subsequent argument.  For example: @{thm
  1000 [locale=lattice] inf_assoc}.
  1001 
  1002 * Proper output of proof terms (@{prf ...} and @{full_prf ...}) within
  1003 a proof context.
  1004 
  1005 * Proper output of antiquotations for theory commands involving a
  1006 proof context (such as 'locale' or 'theorem (in loc) ...').
  1007 
  1008 * Delimiters of outer tokens (string etc.) now produce separate LaTeX
  1009 macros (\isachardoublequoteopen, isachardoublequoteclose etc.).
  1010 
  1011 * isatool usedir: new option -C (default true) controls whether option
  1012 -D should include a copy of the original document directory; -C false
  1013 prevents unwanted effects such as copying of administrative CVS data.
  1014 
  1015 
  1016 *** Pure ***
  1017 
  1018 * Considerably improved version of 'constdefs' command.  Now performs
  1019 automatic type-inference of declared constants; additional support for
  1020 local structure declarations (cf. locales and HOL records), see also
  1021 isar-ref manual.  Potential INCOMPATIBILITY: need to observe strictly
  1022 sequential dependencies of definitions within a single 'constdefs'
  1023 section; moreover, the declared name needs to be an identifier.  If
  1024 all fails, consider to fall back on 'consts' and 'defs' separately.
  1025 
  1026 * Improved indexed syntax and implicit structures.  First of all,
  1027 indexed syntax provides a notational device for subscripted
  1028 application, using the new syntax \<^bsub>term\<^esub> for arbitrary
  1029 expressions.  Secondly, in a local context with structure
  1030 declarations, number indexes \<^sub>n or the empty index (default
  1031 number 1) refer to a certain fixed variable implicitly; option
  1032 show_structs controls printing of implicit structures.  Typical
  1033 applications of these concepts involve record types and locales.
  1034 
  1035 * New command 'no_syntax' removes grammar declarations (and
  1036 translations) resulting from the given syntax specification, which is
  1037 interpreted in the same manner as for the 'syntax' command.
  1038 
  1039 * 'Advanced' translation functions (parse_translation etc.) may depend
  1040 on the signature of the theory context being presently used for
  1041 parsing/printing, see also isar-ref manual.
  1042 
  1043 * Improved 'oracle' command provides a type-safe interface to turn an
  1044 ML expression of type theory -> T -> term into a primitive rule of
  1045 type theory -> T -> thm (i.e. the functionality of Thm.invoke_oracle
  1046 is already included here); see also FOL/ex/IffExample.thy;
  1047 INCOMPATIBILITY.
  1048 
  1049 * axclass: name space prefix for class "c" is now "c_class" (was "c"
  1050 before); "cI" is no longer bound, use "c.intro" instead.
  1051 INCOMPATIBILITY.  This change avoids clashes of fact bindings for
  1052 axclasses vs. locales.
  1053 
  1054 * Improved internal renaming of symbolic identifiers -- attach primes
  1055 instead of base 26 numbers.
  1056 
  1057 * New flag show_question_marks controls printing of leading question
  1058 marks in schematic variable names.
  1059 
  1060 * In schematic variable names, *any* symbol following \<^isub> or
  1061 \<^isup> is now treated as part of the base name.  For example, the
  1062 following works without printing of awkward ".0" indexes:
  1063 
  1064   lemma "x\<^isub>1 = x\<^isub>2 ==> x\<^isub>2 = x\<^isub>1"
  1065     by simp
  1066 
  1067 * Inner syntax includes (*(*nested*) comments*).
  1068 
  1069 * Pretty printer now supports unbreakable blocks, specified in mixfix
  1070 annotations as "(00...)".
  1071 
  1072 * Clear separation of logical types and nonterminals, where the latter
  1073 may only occur in 'syntax' specifications or type abbreviations.
  1074 Before that distinction was only partially implemented via type class
  1075 "logic" vs. "{}".  Potential INCOMPATIBILITY in rare cases of improper
  1076 use of 'types'/'consts' instead of 'nonterminals'/'syntax'.  Some very
  1077 exotic syntax specifications may require further adaption
  1078 (e.g. Cube/Cube.thy).
  1079 
  1080 * Removed obsolete type class "logic", use the top sort {} instead.
  1081 Note that non-logical types should be declared as 'nonterminals'
  1082 rather than 'types'.  INCOMPATIBILITY for new object-logic
  1083 specifications.
  1084 
  1085 * Attributes 'induct' and 'cases': type or set names may now be
  1086 locally fixed variables as well.
  1087 
  1088 * Simplifier: can now control the depth to which conditional rewriting
  1089 is traced via the PG menu Isabelle -> Settings -> Trace Simp Depth
  1090 Limit.
  1091 
  1092 * Simplifier: simplification procedures may now take the current
  1093 simpset into account (cf. Simplifier.simproc(_i) / mk_simproc
  1094 interface), which is very useful for calling the Simplifier
  1095 recursively.  Minor INCOMPATIBILITY: the 'prems' argument of simprocs
  1096 is gone -- use prems_of_ss on the simpset instead.  Moreover, the
  1097 low-level mk_simproc no longer applies Logic.varify internally, to
  1098 allow for use in a context of fixed variables.
  1099 
  1100 * thin_tac now works even if the assumption being deleted contains !!
  1101 or ==>.  More generally, erule now works even if the major premise of
  1102 the elimination rule contains !! or ==>.
  1103 
  1104 * Method 'rules' has been renamed to 'iprover'. INCOMPATIBILITY.
  1105 
  1106 * Reorganized bootstrapping of the Pure theories; CPure is now derived
  1107 from Pure, which contains all common declarations already.  Both
  1108 theories are defined via plain Isabelle/Isar .thy files.
  1109 INCOMPATIBILITY: elements of CPure (such as the CPure.intro /
  1110 CPure.elim / CPure.dest attributes) now appear in the Pure name space;
  1111 use isatool fixcpure to adapt your theory and ML sources.
  1112 
  1113 * New syntax 'name(i-j, i-, i, ...)' for referring to specific
  1114 selections of theorems in named facts via index ranges.
  1115 
  1116 * 'print_theorems': in theory mode, really print the difference
  1117 wrt. the last state (works for interactive theory development only),
  1118 in proof mode print all local facts (cf. 'print_facts');
  1119 
  1120 * 'hide': option '(open)' hides only base names.
  1121 
  1122 * More efficient treatment of intermediate checkpoints in interactive
  1123 theory development.
  1124 
  1125 * Code generator is now invoked via code_module (incremental code
  1126 generation) and code_library (modular code generation, ML structures
  1127 for each theory).  INCOMPATIBILITY: new keywords 'file' and 'contains'
  1128 must be quoted when used as identifiers.
  1129 
  1130 * New 'value' command for reading, evaluating and printing terms using
  1131 the code generator.  INCOMPATIBILITY: command keyword 'value' must be
  1132 quoted when used as identifier.
  1133 
  1134 
  1135 *** Locales ***
  1136 
  1137 * New commands for the interpretation of locale expressions in
  1138 theories (1), locales (2) and proof contexts (3).  These generate
  1139 proof obligations from the expression specification.  After the
  1140 obligations have been discharged, theorems of the expression are added
  1141 to the theory, target locale or proof context.  The synopsis of the
  1142 commands is a follows:
  1143 
  1144   (1) interpretation expr inst
  1145   (2) interpretation target < expr
  1146   (3) interpret expr inst
  1147 
  1148 Interpretation in theories and proof contexts require a parameter
  1149 instantiation of terms from the current context.  This is applied to
  1150 specifications and theorems of the interpreted expression.
  1151 Interpretation in locales only permits parameter renaming through the
  1152 locale expression.  Interpretation is smart in that interpretations
  1153 that are active already do not occur in proof obligations, neither are
  1154 instantiated theorems stored in duplicate.  Use 'print_interps' to
  1155 inspect active interpretations of a particular locale.  For details,
  1156 see the Isar Reference manual.  Examples can be found in
  1157 HOL/Finite_Set.thy and HOL/Algebra/UnivPoly.thy.
  1158 
  1159 INCOMPATIBILITY: former 'instantiate' has been withdrawn, use
  1160 'interpret' instead.
  1161 
  1162 * New context element 'constrains' for adding type constraints to
  1163 parameters.
  1164 
  1165 * Context expressions: renaming of parameters with syntax
  1166 redeclaration.
  1167 
  1168 * Locale declaration: 'includes' disallowed.
  1169 
  1170 * Proper static binding of attribute syntax -- i.e. types / terms /
  1171 facts mentioned as arguments are always those of the locale definition
  1172 context, independently of the context of later invocations.  Moreover,
  1173 locale operations (renaming and type / term instantiation) are applied
  1174 to attribute arguments as expected.
  1175 
  1176 INCOMPATIBILITY of the ML interface: always pass Attrib.src instead of
  1177 actual attributes; rare situations may require Attrib.attribute to
  1178 embed those attributes into Attrib.src that lack concrete syntax.
  1179 Attribute implementations need to cooperate properly with the static
  1180 binding mechanism.  Basic parsers Args.XXX_typ/term/prop and
  1181 Attrib.XXX_thm etc. already do the right thing without further
  1182 intervention.  Only unusual applications -- such as "where" or "of"
  1183 (cf. src/Pure/Isar/attrib.ML), which process arguments depending both
  1184 on the context and the facts involved -- may have to assign parsed
  1185 values to argument tokens explicitly.
  1186 
  1187 * Changed parameter management in theorem generation for long goal
  1188 statements with 'includes'.  INCOMPATIBILITY: produces a different
  1189 theorem statement in rare situations.
  1190 
  1191 * Locale inspection command 'print_locale' omits notes elements.  Use
  1192 'print_locale!' to have them included in the output.
  1193 
  1194 
  1195 *** Provers ***
  1196 
  1197 * Provers/hypsubst.ML: improved version of the subst method, for
  1198 single-step rewriting: it now works in bound variable contexts. New is
  1199 'subst (asm)', for rewriting an assumption.  INCOMPATIBILITY: may
  1200 rewrite a different subterm than the original subst method, which is
  1201 still available as 'simplesubst'.
  1202 
  1203 * Provers/quasi.ML: new transitivity reasoners for transitivity only
  1204 and quasi orders.
  1205 
  1206 * Provers/trancl.ML: new transitivity reasoner for transitive and
  1207 reflexive-transitive closure of relations.
  1208 
  1209 * Provers/blast.ML: new reference depth_limit to make blast's depth
  1210 limit (previously hard-coded with a value of 20) user-definable.
  1211 
  1212 * Provers/simplifier.ML has been moved to Pure, where Simplifier.setup
  1213 is peformed already.  Object-logics merely need to finish their
  1214 initial simpset configuration as before.  INCOMPATIBILITY.
  1215 
  1216 
  1217 *** HOL ***
  1218 
  1219 * Symbolic syntax of Hilbert Choice Operator is now as follows:
  1220 
  1221   syntax (epsilon)
  1222     "_Eps" :: "[pttrn, bool] => 'a"    ("(3\<some>_./ _)" [0, 10] 10)
  1223 
  1224 The symbol \<some> is displayed as the alternative epsilon of LaTeX
  1225 and x-symbol; use option '-m epsilon' to get it actually printed.
  1226 Moreover, the mathematically important symbolic identifier \<epsilon>
  1227 becomes available as variable, constant etc.  INCOMPATIBILITY,
  1228 
  1229 * "x > y" abbreviates "y < x" and "x >= y" abbreviates "y <= x".
  1230 Similarly for all quantifiers: "ALL x > y" etc.  The x-symbol for >=
  1231 is \<ge>. New transitivity rules have been added to HOL/Orderings.thy to
  1232 support corresponding Isar calculations.
  1233 
  1234 * "{x:A. P}" abbreviates "{x. x:A & P}", and similarly for "\<in>"
  1235 instead of ":".
  1236 
  1237 * theory SetInterval: changed the syntax for open intervals:
  1238 
  1239   Old       New
  1240   {..n(}    {..<n}
  1241   {)n..}    {n<..}
  1242   {m..n(}   {m..<n}
  1243   {)m..n}   {m<..n}
  1244   {)m..n(}  {m<..<n}
  1245 
  1246 The old syntax is still supported but will disappear in the next
  1247 release.  For conversion use the following Emacs search and replace
  1248 patterns (these are not perfect but work quite well):
  1249 
  1250   {)\([^\.]*\)\.\.  ->  {\1<\.\.}
  1251   \.\.\([^(}]*\)(}  ->  \.\.<\1}
  1252 
  1253 * Theory Commutative_Ring (in Library): method comm_ring for proving
  1254 equalities in commutative rings; method 'algebra' provides a generic
  1255 interface.
  1256 
  1257 * Theory Finite_Set: changed the syntax for 'setsum', summation over
  1258 finite sets: "setsum (%x. e) A", which used to be "\<Sum>x:A. e", is
  1259 now either "SUM x:A. e" or "\<Sum>x \<in> A. e". The bound variable can
  1260 be a tuple pattern.
  1261 
  1262 Some new syntax forms are available:
  1263 
  1264   "\<Sum>x | P. e"      for     "setsum (%x. e) {x. P}"
  1265   "\<Sum>x = a..b. e"   for     "setsum (%x. e) {a..b}"
  1266   "\<Sum>x = a..<b. e"  for     "setsum (%x. e) {a..<b}"
  1267   "\<Sum>x < k. e"      for     "setsum (%x. e) {..<k}"
  1268 
  1269 The latter form "\<Sum>x < k. e" used to be based on a separate
  1270 function "Summation", which has been discontinued.
  1271 
  1272 * theory Finite_Set: in structured induction proofs, the insert case
  1273 is now 'case (insert x F)' instead of the old counterintuitive 'case
  1274 (insert F x)'.
  1275 
  1276 * The 'refute' command has been extended to support a much larger
  1277 fragment of HOL, including axiomatic type classes, constdefs and
  1278 typedefs, inductive datatypes and recursion.
  1279 
  1280 * New tactics 'sat' and 'satx' to prove propositional tautologies.
  1281 Requires zChaff with proof generation to be installed.  See
  1282 HOL/ex/SAT_Examples.thy for examples.
  1283 
  1284 * Datatype induction via method 'induct' now preserves the name of the
  1285 induction variable. For example, when proving P(xs::'a list) by
  1286 induction on xs, the induction step is now P(xs) ==> P(a#xs) rather
  1287 than P(list) ==> P(a#list) as previously.  Potential INCOMPATIBILITY
  1288 in unstructured proof scripts.
  1289 
  1290 * Reworked implementation of records.  Improved scalability for
  1291 records with many fields, avoiding performance problems for type
  1292 inference. Records are no longer composed of nested field types, but
  1293 of nested extension types. Therefore the record type only grows linear
  1294 in the number of extensions and not in the number of fields.  The
  1295 top-level (users) view on records is preserved.  Potential
  1296 INCOMPATIBILITY only in strange cases, where the theory depends on the
  1297 old record representation. The type generated for a record is called
  1298 <record_name>_ext_type.
  1299 
  1300 Flag record_quick_and_dirty_sensitive can be enabled to skip the
  1301 proofs triggered by a record definition or a simproc (if
  1302 quick_and_dirty is enabled).  Definitions of large records can take
  1303 quite long.
  1304 
  1305 New simproc record_upd_simproc for simplification of multiple record
  1306 updates enabled by default.  Moreover, trivial updates are also
  1307 removed: r(|x := x r|) = r.  INCOMPATIBILITY: old proofs break
  1308 occasionally, since simplification is more powerful by default.
  1309 
  1310 * typedef: proper support for polymorphic sets, which contain extra
  1311 type-variables in the term.
  1312 
  1313 * Simplifier: automatically reasons about transitivity chains
  1314 involving "trancl" (r^+) and "rtrancl" (r^*) by setting up tactics
  1315 provided by Provers/trancl.ML as additional solvers.  INCOMPATIBILITY:
  1316 old proofs break occasionally as simplification may now solve more
  1317 goals than previously.
  1318 
  1319 * Simplifier: converts x <= y into x = y if assumption y <= x is
  1320 present.  Works for all partial orders (class "order"), in particular
  1321 numbers and sets.  For linear orders (e.g. numbers) it treats ~ x < y
  1322 just like y <= x.
  1323 
  1324 * Simplifier: new simproc for "let x = a in f x".  If a is a free or
  1325 bound variable or a constant then the let is unfolded.  Otherwise
  1326 first a is simplified to b, and then f b is simplified to g. If
  1327 possible we abstract b from g arriving at "let x = b in h x",
  1328 otherwise we unfold the let and arrive at g.  The simproc can be
  1329 enabled/disabled by the reference use_let_simproc.  Potential
  1330 INCOMPATIBILITY since simplification is more powerful by default.
  1331 
  1332 * Classical reasoning: the meson method now accepts theorems as arguments.
  1333 
  1334 * Prover support: pre-release of the Isabelle-ATP linkup, which runs background
  1335 jobs to provide advice on the provability of subgoals.
  1336 
  1337 * Theory OrderedGroup and Ring_and_Field: various additions and
  1338 improvements to faciliate calculations involving equalities and
  1339 inequalities.
  1340 
  1341 The following theorems have been eliminated or modified
  1342 (INCOMPATIBILITY):
  1343 
  1344   abs_eq             now named abs_of_nonneg
  1345   abs_of_ge_0        now named abs_of_nonneg
  1346   abs_minus_eq       now named abs_of_nonpos
  1347   imp_abs_id         now named abs_of_nonneg
  1348   imp_abs_neg_id     now named abs_of_nonpos
  1349   mult_pos           now named mult_pos_pos
  1350   mult_pos_le        now named mult_nonneg_nonneg
  1351   mult_pos_neg_le    now named mult_nonneg_nonpos
  1352   mult_pos_neg2_le   now named mult_nonneg_nonpos2
  1353   mult_neg           now named mult_neg_neg
  1354   mult_neg_le        now named mult_nonpos_nonpos
  1355 
  1356 * Theory Parity: added rules for simplifying exponents.
  1357 
  1358 * Theory List:
  1359 
  1360 The following theorems have been eliminated or modified
  1361 (INCOMPATIBILITY):
  1362 
  1363   list_all_Nil       now named list_all.simps(1)
  1364   list_all_Cons      now named list_all.simps(2)
  1365   list_all_conv      now named list_all_iff
  1366   set_mem_eq         now named mem_iff
  1367 
  1368 * Theories SetsAndFunctions and BigO (see HOL/Library) support
  1369 asymptotic "big O" calculations.  See the notes in BigO.thy.
  1370 
  1371 
  1372 *** HOL-Complex ***
  1373 
  1374 * Theory RealDef: better support for embedding natural numbers and
  1375 integers in the reals.
  1376 
  1377 The following theorems have been eliminated or modified
  1378 (INCOMPATIBILITY):
  1379 
  1380   exp_ge_add_one_self  now requires no hypotheses
  1381   real_of_int_add      reversed direction of equality (use [symmetric])
  1382   real_of_int_minus    reversed direction of equality (use [symmetric])
  1383   real_of_int_diff     reversed direction of equality (use [symmetric])
  1384   real_of_int_mult     reversed direction of equality (use [symmetric])
  1385 
  1386 * Theory RComplete: expanded support for floor and ceiling functions.
  1387 
  1388 * Theory Ln is new, with properties of the natural logarithm
  1389 
  1390 * Hyperreal: There is a new type constructor "star" for making
  1391 nonstandard types.  The old type names are now type synonyms:
  1392 
  1393   hypreal = real star
  1394   hypnat = nat star
  1395   hcomplex = complex star
  1396 
  1397 * Hyperreal: Many groups of similarly-defined constants have been
  1398 replaced by polymorphic versions (INCOMPATIBILITY):
  1399 
  1400   star_of <-- hypreal_of_real, hypnat_of_nat, hcomplex_of_complex
  1401 
  1402   starset      <-- starsetNat, starsetC
  1403   *s*          <-- *sNat*, *sc*
  1404   starset_n    <-- starsetNat_n, starsetC_n
  1405   *sn*         <-- *sNatn*, *scn*
  1406   InternalSets <-- InternalNatSets, InternalCSets
  1407 
  1408   starfun      <-- starfun{Nat,Nat2,C,RC,CR}
  1409   *f*          <-- *fNat*, *fNat2*, *fc*, *fRc*, *fcR*
  1410   starfun_n    <-- starfun{Nat,Nat2,C,RC,CR}_n
  1411   *fn*         <-- *fNatn*, *fNat2n*, *fcn*, *fRcn*, *fcRn*
  1412   InternalFuns <-- InternalNatFuns, InternalNatFuns2, Internal{C,RC,CR}Funs
  1413 
  1414 * Hyperreal: Many type-specific theorems have been removed in favor of
  1415 theorems specific to various axiomatic type classes (INCOMPATIBILITY):
  1416 
  1417   add_commute <-- {hypreal,hypnat,hcomplex}_add_commute
  1418   add_assoc   <-- {hypreal,hypnat,hcomplex}_add_assocs
  1419   OrderedGroup.add_0 <-- {hypreal,hypnat,hcomplex}_add_zero_left
  1420   OrderedGroup.add_0_right <-- {hypreal,hcomplex}_add_zero_right
  1421   right_minus <-- hypreal_add_minus
  1422   left_minus <-- {hypreal,hcomplex}_add_minus_left
  1423   mult_commute <-- {hypreal,hypnat,hcomplex}_mult_commute
  1424   mult_assoc <-- {hypreal,hypnat,hcomplex}_mult_assoc
  1425   mult_1_left <-- {hypreal,hypnat}_mult_1, hcomplex_mult_one_left
  1426   mult_1_right <-- hcomplex_mult_one_right
  1427   mult_zero_left <-- hcomplex_mult_zero_left
  1428   left_distrib <-- {hypreal,hypnat,hcomplex}_add_mult_distrib
  1429   right_distrib <-- hypnat_add_mult_distrib2
  1430   zero_neq_one <-- {hypreal,hypnat,hcomplex}_zero_not_eq_one
  1431   right_inverse <-- hypreal_mult_inverse
  1432   left_inverse <-- hypreal_mult_inverse_left, hcomplex_mult_inv_left
  1433   order_refl <-- {hypreal,hypnat}_le_refl
  1434   order_trans <-- {hypreal,hypnat}_le_trans
  1435   order_antisym <-- {hypreal,hypnat}_le_anti_sym
  1436   order_less_le <-- {hypreal,hypnat}_less_le
  1437   linorder_linear <-- {hypreal,hypnat}_le_linear
  1438   add_left_mono <-- {hypreal,hypnat}_add_left_mono
  1439   mult_strict_left_mono <-- {hypreal,hypnat}_mult_less_mono2
  1440   add_nonneg_nonneg <-- hypreal_le_add_order
  1441 
  1442 * Hyperreal: Separate theorems having to do with type-specific
  1443 versions of constants have been merged into theorems that apply to the
  1444 new polymorphic constants (INCOMPATIBILITY):
  1445 
  1446   STAR_UNIV_set <-- {STAR_real,NatStar_real,STARC_complex}_set
  1447   STAR_empty_set <-- {STAR,NatStar,STARC}_empty_set
  1448   STAR_Un <-- {STAR,NatStar,STARC}_Un
  1449   STAR_Int <-- {STAR,NatStar,STARC}_Int
  1450   STAR_Compl <-- {STAR,NatStar,STARC}_Compl
  1451   STAR_subset <-- {STAR,NatStar,STARC}_subset
  1452   STAR_mem <-- {STAR,NatStar,STARC}_mem
  1453   STAR_mem_Compl <-- {STAR,STARC}_mem_Compl
  1454   STAR_diff <-- {STAR,STARC}_diff
  1455   STAR_star_of_image_subset <-- {STAR_hypreal_of_real, NatStar_hypreal_of_real,
  1456     STARC_hcomplex_of_complex}_image_subset
  1457   starset_n_Un <-- starset{Nat,C}_n_Un
  1458   starset_n_Int <-- starset{Nat,C}_n_Int
  1459   starset_n_Compl <-- starset{Nat,C}_n_Compl
  1460   starset_n_diff <-- starset{Nat,C}_n_diff
  1461   InternalSets_Un <-- Internal{Nat,C}Sets_Un
  1462   InternalSets_Int <-- Internal{Nat,C}Sets_Int
  1463   InternalSets_Compl <-- Internal{Nat,C}Sets_Compl
  1464   InternalSets_diff <-- Internal{Nat,C}Sets_diff
  1465   InternalSets_UNIV_diff <-- Internal{Nat,C}Sets_UNIV_diff
  1466   InternalSets_starset_n <-- Internal{Nat,C}Sets_starset{Nat,C}_n
  1467   starset_starset_n_eq <-- starset{Nat,C}_starset{Nat,C}_n_eq
  1468   starset_n_starset <-- starset{Nat,C}_n_starset{Nat,C}
  1469   starfun_n_starfun <-- starfun{Nat,Nat2,C,RC,CR}_n_starfun{Nat,Nat2,C,RC,CR}
  1470   starfun <-- starfun{Nat,Nat2,C,RC,CR}
  1471   starfun_mult <-- starfun{Nat,Nat2,C,RC,CR}_mult
  1472   starfun_add <-- starfun{Nat,Nat2,C,RC,CR}_add
  1473   starfun_minus <-- starfun{Nat,Nat2,C,RC,CR}_minus
  1474   starfun_diff <-- starfun{C,RC,CR}_diff
  1475   starfun_o <-- starfun{NatNat2,Nat2,_stafunNat,C,C_starfunRC,_starfunCR}_o
  1476   starfun_o2 <-- starfun{NatNat2,_stafunNat,C,C_starfunRC,_starfunCR}_o2
  1477   starfun_const_fun <-- starfun{Nat,Nat2,C,RC,CR}_const_fun
  1478   starfun_inverse <-- starfun{Nat,C,RC,CR}_inverse
  1479   starfun_eq <-- starfun{Nat,Nat2,C,RC,CR}_eq
  1480   starfun_eq_iff <-- starfun{C,RC,CR}_eq_iff
  1481   starfun_Id <-- starfunC_Id
  1482   starfun_approx <-- starfun{Nat,CR}_approx
  1483   starfun_capprox <-- starfun{C,RC}_capprox
  1484   starfun_abs <-- starfunNat_rabs
  1485   starfun_lambda_cancel <-- starfun{C,CR,RC}_lambda_cancel
  1486   starfun_lambda_cancel2 <-- starfun{C,CR,RC}_lambda_cancel2
  1487   starfun_mult_HFinite_approx <-- starfunCR_mult_HFinite_capprox
  1488   starfun_mult_CFinite_capprox <-- starfun{C,RC}_mult_CFinite_capprox
  1489   starfun_add_capprox <-- starfun{C,RC}_add_capprox
  1490   starfun_add_approx <-- starfunCR_add_approx
  1491   starfun_inverse_inverse <-- starfunC_inverse_inverse
  1492   starfun_divide <-- starfun{C,CR,RC}_divide
  1493   starfun_n <-- starfun{Nat,C}_n
  1494   starfun_n_mult <-- starfun{Nat,C}_n_mult
  1495   starfun_n_add <-- starfun{Nat,C}_n_add
  1496   starfun_n_add_minus <-- starfunNat_n_add_minus
  1497   starfun_n_const_fun <-- starfun{Nat,C}_n_const_fun
  1498   starfun_n_minus <-- starfun{Nat,C}_n_minus
  1499   starfun_n_eq <-- starfun{Nat,C}_n_eq
  1500 
  1501   star_n_add <-- {hypreal,hypnat,hcomplex}_add
  1502   star_n_minus <-- {hypreal,hcomplex}_minus
  1503   star_n_diff <-- {hypreal,hcomplex}_diff
  1504   star_n_mult <-- {hypreal,hcomplex}_mult
  1505   star_n_inverse <-- {hypreal,hcomplex}_inverse
  1506   star_n_le <-- {hypreal,hypnat}_le
  1507   star_n_less <-- {hypreal,hypnat}_less
  1508   star_n_zero_num <-- {hypreal,hypnat,hcomplex}_zero_num
  1509   star_n_one_num <-- {hypreal,hypnat,hcomplex}_one_num
  1510   star_n_abs <-- hypreal_hrabs
  1511   star_n_divide <-- hcomplex_divide
  1512 
  1513   star_of_add <-- {hypreal_of_real,hypnat_of_nat,hcomplex_of_complex}_add
  1514   star_of_minus <-- {hypreal_of_real,hcomplex_of_complex}_minus
  1515   star_of_diff <-- hypreal_of_real_diff
  1516   star_of_mult <-- {hypreal_of_real,hypnat_of_nat,hcomplex_of_complex}_mult
  1517   star_of_one <-- {hypreal_of_real,hcomplex_of_complex}_one
  1518   star_of_zero <-- {hypreal_of_real,hypnat_of_nat,hcomplex_of_complex}_zero
  1519   star_of_le <-- {hypreal_of_real,hypnat_of_nat}_le_iff
  1520   star_of_less <-- {hypreal_of_real,hypnat_of_nat}_less_iff
  1521   star_of_eq <-- {hypreal_of_real,hypnat_of_nat,hcomplex_of_complex}_eq_iff
  1522   star_of_inverse <-- {hypreal_of_real,hcomplex_of_complex}_inverse
  1523   star_of_divide <-- {hypreal_of_real,hcomplex_of_complex}_divide
  1524   star_of_of_nat <-- {hypreal_of_real,hcomplex_of_complex}_of_nat
  1525   star_of_of_int <-- {hypreal_of_real,hcomplex_of_complex}_of_int
  1526   star_of_number_of <-- {hypreal,hcomplex}_number_of
  1527   star_of_number_less <-- number_of_less_hypreal_of_real_iff
  1528   star_of_number_le <-- number_of_le_hypreal_of_real_iff
  1529   star_of_eq_number <-- hypreal_of_real_eq_number_of_iff
  1530   star_of_less_number <-- hypreal_of_real_less_number_of_iff
  1531   star_of_le_number <-- hypreal_of_real_le_number_of_iff
  1532   star_of_power <-- hypreal_of_real_power
  1533   star_of_eq_0 <-- hcomplex_of_complex_zero_iff
  1534 
  1535 * Hyperreal: new method "transfer" that implements the transfer
  1536 principle of nonstandard analysis. With a subgoal that mentions
  1537 nonstandard types like "'a star", the command "apply transfer"
  1538 replaces it with an equivalent one that mentions only standard types.
  1539 To be successful, all free variables must have standard types; non-
  1540 standard variables must have explicit universal quantifiers.
  1541 
  1542 * Hyperreal: A theory of Taylor series.
  1543 
  1544 
  1545 *** HOLCF ***
  1546 
  1547 * Discontinued special version of 'constdefs' (which used to support
  1548 continuous functions) in favor of the general Pure one with full
  1549 type-inference.
  1550 
  1551 * New simplification procedure for solving continuity conditions; it
  1552 is much faster on terms with many nested lambda abstractions (cubic
  1553 instead of exponential time).
  1554 
  1555 * New syntax for domain package: selector names are now optional.
  1556 Parentheses should be omitted unless argument is lazy, for example:
  1557 
  1558   domain 'a stream = cons "'a" (lazy "'a stream")
  1559 
  1560 * New command 'fixrec' for defining recursive functions with pattern
  1561 matching; defining multiple functions with mutual recursion is also
  1562 supported.  Patterns may include the constants cpair, spair, up, sinl,
  1563 sinr, or any data constructor defined by the domain package. The given
  1564 equations are proven as rewrite rules. See HOLCF/ex/Fixrec_ex.thy for
  1565 syntax and examples.
  1566 
  1567 * New commands 'cpodef' and 'pcpodef' for defining predicate subtypes
  1568 of cpo and pcpo types. Syntax is exactly like the 'typedef' command,
  1569 but the proof obligation additionally includes an admissibility
  1570 requirement. The packages generate instances of class cpo or pcpo,
  1571 with continuity and strictness theorems for Rep and Abs.
  1572 
  1573 * HOLCF: Many theorems have been renamed according to a more standard naming
  1574 scheme (INCOMPATIBILITY):
  1575 
  1576   foo_inject:  "foo$x = foo$y ==> x = y"
  1577   foo_eq:      "(foo$x = foo$y) = (x = y)"
  1578   foo_less:    "(foo$x << foo$y) = (x << y)"
  1579   foo_strict:  "foo$UU = UU"
  1580   foo_defined: "... ==> foo$x ~= UU"
  1581   foo_defined_iff: "(foo$x = UU) = (x = UU)"
  1582 
  1583 
  1584 *** ZF ***
  1585 
  1586 * ZF/ex: theories Group and Ring provide examples in abstract algebra,
  1587 including the First Isomorphism Theorem (on quotienting by the kernel
  1588 of a homomorphism).
  1589 
  1590 * ZF/Simplifier: install second copy of type solver that actually
  1591 makes use of TC rules declared to Isar proof contexts (or locales);
  1592 the old version is still required for ML proof scripts.
  1593 
  1594 
  1595 *** Cube ***
  1596 
  1597 * Converted to Isar theory format; use locales instead of axiomatic
  1598 theories.
  1599 
  1600 
  1601 *** ML ***
  1602 
  1603 * Pure/library.ML no longer defines its own option datatype, but uses
  1604 that of the SML basis, which has constructors NONE and SOME instead of
  1605 None and Some, as well as exception Option.Option instead of OPTION.
  1606 The functions the, if_none, is_some, is_none have been adapted
  1607 accordingly, while Option.map replaces apsome.
  1608 
  1609 * Pure/library.ML: the exception LIST has been given up in favour of
  1610 the standard exceptions Empty and Subscript, as well as
  1611 Library.UnequalLengths.  Function like Library.hd and Library.tl are
  1612 superceded by the standard hd and tl functions etc.
  1613 
  1614 A number of basic list functions are no longer exported to the ML
  1615 toplevel, as they are variants of predefined functions.  The following
  1616 suggests how one can translate existing code:
  1617 
  1618     rev_append xs ys = List.revAppend (xs, ys)
  1619     nth_elem (i, xs) = List.nth (xs, i)
  1620     last_elem xs = List.last xs
  1621     flat xss = List.concat xss
  1622     seq fs = List.app fs
  1623     partition P xs = List.partition P xs
  1624     mapfilter f xs = List.mapPartial f xs
  1625 
  1626 * Pure/library.ML: several combinators for linear functional
  1627 transformations, notably reverse application and composition:
  1628 
  1629   x |> f                f #> g
  1630   (x, y) |-> f          f #-> g
  1631 
  1632 * Pure/library.ML: introduced/changed precedence of infix operators:
  1633 
  1634   infix 1 |> |-> ||> ||>> |>> |>>> #> #->;
  1635   infix 2 ?;
  1636   infix 3 o oo ooo oooo;
  1637   infix 4 ~~ upto downto;
  1638 
  1639 Maybe INCOMPATIBILITY when any of those is used in conjunction with other
  1640 infix operators.
  1641 
  1642 * Pure/library.ML: natural list combinators fold, fold_rev, and
  1643 fold_map support linear functional transformations and nesting.  For
  1644 example:
  1645 
  1646   fold f [x1, ..., xN] y =
  1647     y |> f x1 |> ... |> f xN
  1648 
  1649   (fold o fold) f [xs1, ..., xsN] y =
  1650     y |> fold f xs1 |> ... |> fold f xsN
  1651 
  1652   fold f [x1, ..., xN] =
  1653     f x1 #> ... #> f xN
  1654 
  1655   (fold o fold) f [xs1, ..., xsN] =
  1656     fold f xs1 #> ... #> fold f xsN
  1657 
  1658 * Pure/library.ML: the following selectors on type 'a option are
  1659 available:
  1660 
  1661   the:               'a option -> 'a  (*partial*)
  1662   these:             'a option -> 'a  where 'a = 'b list
  1663   the_default: 'a -> 'a option -> 'a
  1664   the_list:          'a option -> 'a list
  1665 
  1666 * Pure/General: structure AList (cf. Pure/General/alist.ML) provides
  1667 basic operations for association lists, following natural argument
  1668 order; moreover the explicit equality predicate passed here avoids
  1669 potentially expensive polymorphic runtime equality checks.
  1670 The old functions may be expressed as follows:
  1671 
  1672   assoc = uncurry (AList.lookup (op =))
  1673   assocs = these oo AList.lookup (op =)
  1674   overwrite = uncurry (AList.update (op =)) o swap
  1675 
  1676 * Pure/General: structure AList (cf. Pure/General/alist.ML) provides
  1677 
  1678   val make: ('a -> 'b) -> 'a list -> ('a * 'b) list
  1679   val find: ('a * 'b -> bool) -> ('c * 'b) list -> 'a -> 'c list
  1680 
  1681 replacing make_keylist and keyfilter (occassionally used)
  1682 Naive rewrites:
  1683 
  1684   make_keylist = AList.make
  1685   keyfilter = AList.find (op =)
  1686 
  1687 * eq_fst and eq_snd now take explicit equality parameter, thus
  1688   avoiding eqtypes. Naive rewrites:
  1689 
  1690     eq_fst = eq_fst (op =)
  1691     eq_snd = eq_snd (op =)
  1692 
  1693 * Removed deprecated apl and apr (rarely used).
  1694   Naive rewrites:
  1695 
  1696     apl (n, op) =>>= curry op n
  1697     apr (op, m) =>>= fn n => op (n, m)
  1698 
  1699 * Pure/General: structure OrdList (cf. Pure/General/ord_list.ML)
  1700 provides a reasonably efficient light-weight implementation of sets as
  1701 lists.
  1702 
  1703 * Pure/General: generic tables (cf. Pure/General/table.ML) provide a
  1704 few new operations; existing lookup and update are now curried to
  1705 follow natural argument order (for use with fold etc.);
  1706 INCOMPATIBILITY, use (uncurry Symtab.lookup) etc. as last resort.
  1707 
  1708 * Pure/General: output via the Isabelle channels of
  1709 writeln/warning/error etc. is now passed through Output.output, with a
  1710 hook for arbitrary transformations depending on the print_mode
  1711 (cf. Output.add_mode -- the first active mode that provides a output
  1712 function wins).  Already formatted output may be embedded into further
  1713 text via Output.raw; the result of Pretty.string_of/str_of and derived
  1714 functions (string_of_term/cterm/thm etc.) is already marked raw to
  1715 accommodate easy composition of diagnostic messages etc.  Programmers
  1716 rarely need to care about Output.output or Output.raw at all, with
  1717 some notable exceptions: Output.output is required when bypassing the
  1718 standard channels (writeln etc.), or in token translations to produce
  1719 properly formatted results; Output.raw is required when capturing
  1720 already output material that will eventually be presented to the user
  1721 a second time.  For the default print mode, both Output.output and
  1722 Output.raw have no effect.
  1723 
  1724 * Pure/General: Output.time_accumulator NAME creates an operator ('a
  1725 -> 'b) -> 'a -> 'b to measure runtime and count invocations; the
  1726 cumulative results are displayed at the end of a batch session.
  1727 
  1728 * Pure/General: File.sysify_path and File.quote_sysify path have been
  1729 replaced by File.platform_path and File.shell_path (with appropriate
  1730 hooks).  This provides a clean interface for unusual systems where the
  1731 internal and external process view of file names are different.
  1732 
  1733 * Pure: more efficient orders for basic syntactic entities: added
  1734 fast_string_ord, fast_indexname_ord, fast_term_ord; changed sort_ord
  1735 and typ_ord to use fast_string_ord and fast_indexname_ord (term_ord is
  1736 NOT affected); structures Symtab, Vartab, Typtab, Termtab use the fast
  1737 orders now -- potential INCOMPATIBILITY for code that depends on a
  1738 particular order for Symtab.keys, Symtab.dest, etc. (consider using
  1739 Library.sort_strings on result).
  1740 
  1741 * Pure/term.ML: combinators fold_atyps, fold_aterms, fold_term_types,
  1742 fold_types traverse types/terms from left to right, observing natural
  1743 argument order.  Supercedes previous foldl_XXX versions, add_frees,
  1744 add_vars etc. have been adapted as well: INCOMPATIBILITY.
  1745 
  1746 * Pure: name spaces have been refined, with significant changes of the
  1747 internal interfaces -- INCOMPATIBILITY.  Renamed cond_extern(_table)
  1748 to extern(_table).  The plain name entry path is superceded by a
  1749 general 'naming' context, which also includes the 'policy' to produce
  1750 a fully qualified name and external accesses of a fully qualified
  1751 name; NameSpace.extend is superceded by context dependent
  1752 Sign.declare_name.  Several theory and proof context operations modify
  1753 the naming context.  Especially note Theory.restore_naming and
  1754 ProofContext.restore_naming to get back to a sane state; note that
  1755 Theory.add_path is no longer sufficient to recover from
  1756 Theory.absolute_path in particular.
  1757 
  1758 * Pure: new flags short_names (default false) and unique_names
  1759 (default true) for controlling output of qualified names.  If
  1760 short_names is set, names are printed unqualified.  If unique_names is
  1761 reset, the name prefix is reduced to the minimum required to achieve
  1762 the original result when interning again, even if there is an overlap
  1763 with earlier declarations.
  1764 
  1765 * Pure/TheoryDataFun: change of the argument structure; 'prep_ext' is
  1766 now 'extend', and 'merge' gets an additional Pretty.pp argument
  1767 (useful for printing error messages).  INCOMPATIBILITY.
  1768 
  1769 * Pure: major reorganization of the theory context.  Type Sign.sg and
  1770 Theory.theory are now identified, referring to the universal
  1771 Context.theory (see Pure/context.ML).  Actual signature and theory
  1772 content is managed as theory data.  The old code and interfaces were
  1773 spread over many files and structures; the new arrangement introduces
  1774 considerable INCOMPATIBILITY to gain more clarity:
  1775 
  1776   Context -- theory management operations (name, identity, inclusion,
  1777     parents, ancestors, merge, etc.), plus generic theory data;
  1778 
  1779   Sign -- logical signature and syntax operations (declaring consts,
  1780     types, etc.), plus certify/read for common entities;
  1781 
  1782   Theory -- logical theory operations (stating axioms, definitions,
  1783     oracles), plus a copy of logical signature operations (consts,
  1784     types, etc.); also a few basic management operations (Theory.copy,
  1785     Theory.merge, etc.)
  1786 
  1787 The most basic sign_of operations (Theory.sign_of, Thm.sign_of_thm
  1788 etc.) as well as the sign field in Thm.rep_thm etc. have been retained
  1789 for convenience -- they merely return the theory.
  1790 
  1791 * Pure: type Type.tsig is superceded by theory in most interfaces.
  1792 
  1793 * Pure: the Isar proof context type is already defined early in Pure
  1794 as Context.proof (note that ProofContext.context and Proof.context are
  1795 aliases, where the latter is the preferred name).  This enables other
  1796 Isabelle components to refer to that type even before Isar is present.
  1797 
  1798 * Pure/sign/theory: discontinued named name spaces (i.e. classK,
  1799 typeK, constK, axiomK, oracleK), but provide explicit operations for
  1800 any of these kinds.  For example, Sign.intern typeK is now
  1801 Sign.intern_type, Theory.hide_space Sign.typeK is now
  1802 Theory.hide_types.  Also note that former
  1803 Theory.hide_classes/types/consts are now
  1804 Theory.hide_classes_i/types_i/consts_i, while the non '_i' versions
  1805 internalize their arguments!  INCOMPATIBILITY.
  1806 
  1807 * Pure: get_thm interface (of PureThy and ProofContext) expects
  1808 datatype thmref (with constructors Name and NameSelection) instead of
  1809 plain string -- INCOMPATIBILITY;
  1810 
  1811 * Pure: cases produced by proof methods specify options, where NONE
  1812 means to remove case bindings -- INCOMPATIBILITY in
  1813 (RAW_)METHOD_CASES.
  1814 
  1815 * Pure: the following operations retrieve axioms or theorems from a
  1816 theory node or theory hierarchy, respectively:
  1817 
  1818   Theory.axioms_of: theory -> (string * term) list
  1819   Theory.all_axioms_of: theory -> (string * term) list
  1820   PureThy.thms_of: theory -> (string * thm) list
  1821   PureThy.all_thms_of: theory -> (string * thm) list
  1822 
  1823 * Pure: print_tac now outputs the goal through the trace channel.
  1824 
  1825 * Isar toplevel: improved diagnostics, mostly for Poly/ML only.
  1826 Reference Toplevel.debug (default false) controls detailed printing
  1827 and tracing of low-level exceptions; Toplevel.profiling (default 0)
  1828 controls execution profiling -- set to 1 for time and 2 for space
  1829 (both increase the runtime).
  1830 
  1831 * Isar session: The initial use of ROOT.ML is now always timed,
  1832 i.e. the log will show the actual process times, in contrast to the
  1833 elapsed wall-clock time that the outer shell wrapper produces.
  1834 
  1835 * Simplifier: improved handling of bound variables (nameless
  1836 representation, avoid allocating new strings).  Simprocs that invoke
  1837 the Simplifier recursively should use Simplifier.inherit_bounds to
  1838 avoid local name clashes.  Failure to do so produces warnings
  1839 "Simplifier: renamed bound variable ..."; set Simplifier.debug_bounds
  1840 for further details.
  1841 
  1842 * ML functions legacy_bindings and use_legacy_bindings produce ML fact
  1843 bindings for all theorems stored within a given theory; this may help
  1844 in porting non-Isar theories to Isar ones, while keeping ML proof
  1845 scripts for the time being.
  1846 
  1847 * ML operator HTML.with_charset specifies the charset begin used for
  1848 generated HTML files.  For example:
  1849 
  1850   HTML.with_charset "utf-8" use_thy "Hebrew";
  1851   HTML.with_charset "utf-8" use_thy "Chinese";
  1852 
  1853 
  1854 *** System ***
  1855 
  1856 * Allow symlinks to all proper Isabelle executables (Isabelle,
  1857 isabelle, isatool etc.).
  1858 
  1859 * ISABELLE_DOC_FORMAT setting specifies preferred document format (for
  1860 isatool doc, isatool mkdir, display_drafts etc.).
  1861 
  1862 * isatool usedir: option -f allows specification of the ML file to be
  1863 used by Isabelle; default is ROOT.ML.
  1864 
  1865 * New isatool version outputs the version identifier of the Isabelle
  1866 distribution being used.
  1867 
  1868 * HOL: new isatool dimacs2hol converts files in DIMACS CNF format
  1869 (containing Boolean satisfiability problems) into Isabelle/HOL
  1870 theories.
  1871 
  1872 
  1873 
  1874 New in Isabelle2004 (April 2004)
  1875 --------------------------------
  1876 
  1877 *** General ***
  1878 
  1879 * Provers/order.ML:  new efficient reasoner for partial and linear orders.
  1880   Replaces linorder.ML.
  1881 
  1882 * Pure: Greek letters (except small lambda, \<lambda>), as well as Gothic
  1883   (\<aa>...\<zz>\<AA>...\<ZZ>), calligraphic (\<A>...\<Z>), and Euler
  1884   (\<a>...\<z>), are now considered normal letters, and can therefore
  1885   be used anywhere where an ASCII letter (a...zA...Z) has until
  1886   now. COMPATIBILITY: This obviously changes the parsing of some
  1887   terms, especially where a symbol has been used as a binder, say
  1888   '\<Pi>x. ...', which is now a type error since \<Pi>x will be parsed
  1889   as an identifier.  Fix it by inserting a space around former
  1890   symbols.  Call 'isatool fixgreek' to try to fix parsing errors in
  1891   existing theory and ML files.
  1892 
  1893 * Pure: Macintosh and Windows line-breaks are now allowed in theory files.
  1894 
  1895 * Pure: single letter sub/superscripts (\<^isub> and \<^isup>) are now
  1896   allowed in identifiers. Similar to Greek letters \<^isub> is now considered
  1897   a normal (but invisible) letter. For multiple letter subscripts repeat
  1898   \<^isub> like this: x\<^isub>1\<^isub>2.
  1899 
  1900 * Pure: There are now sub-/superscripts that can span more than one
  1901   character. Text between \<^bsub> and \<^esub> is set in subscript in
  1902   ProofGeneral and LaTeX, text between \<^bsup> and \<^esup> in
  1903   superscript. The new control characters are not identifier parts.
  1904 
  1905 * Pure: Control-symbols of the form \<^raw:...> will literally print the
  1906   content of "..." to the latex file instead of \isacntrl... . The "..."
  1907   may consist of any printable characters excluding the end bracket >.
  1908 
  1909 * Pure: Using new Isar command "finalconsts" (or the ML functions
  1910   Theory.add_finals or Theory.add_finals_i) it is now possible to
  1911   declare constants "final", which prevents their being given a definition
  1912   later.  It is useful for constants whose behaviour is fixed axiomatically
  1913   rather than definitionally, such as the meta-logic connectives.
  1914 
  1915 * Pure: 'instance' now handles general arities with general sorts
  1916   (i.e. intersections of classes),
  1917 
  1918 * Presentation: generated HTML now uses a CSS style sheet to make layout
  1919   (somewhat) independent of content. It is copied from lib/html/isabelle.css.
  1920   It can be changed to alter the colors/layout of generated pages.
  1921 
  1922 
  1923 *** Isar ***
  1924 
  1925 * Tactic emulation methods rule_tac, erule_tac, drule_tac, frule_tac,
  1926   cut_tac, subgoal_tac and thin_tac:
  1927   - Now understand static (Isar) contexts.  As a consequence, users of Isar
  1928     locales are no longer forced to write Isar proof scripts.
  1929     For details see Isar Reference Manual, paragraph 4.3.2: Further tactic
  1930     emulations.
  1931   - INCOMPATIBILITY: names of variables to be instantiated may no
  1932     longer be enclosed in quotes.  Instead, precede variable name with `?'.
  1933     This is consistent with the instantiation attribute "where".
  1934 
  1935 * Attributes "where" and "of":
  1936   - Now take type variables of instantiated theorem into account when reading
  1937     the instantiation string.  This fixes a bug that caused instantiated
  1938     theorems to have too special types in some circumstances.
  1939   - "where" permits explicit instantiations of type variables.
  1940 
  1941 * Calculation commands "moreover" and "also" no longer interfere with
  1942   current facts ("this"), admitting arbitrary combinations with "then"
  1943   and derived forms.
  1944 
  1945 * Locales:
  1946   - Goal statements involving the context element "includes" no longer
  1947     generate theorems with internal delta predicates (those ending on
  1948     "_axioms") in the premise.
  1949     Resolve particular premise with <locale>.intro to obtain old form.
  1950   - Fixed bug in type inference ("unify_frozen") that prevented mix of target
  1951     specification and "includes" elements in goal statement.
  1952   - Rule sets <locale>.intro and <locale>.axioms no longer declared as
  1953     [intro?] and [elim?] (respectively) by default.
  1954   - Experimental command for instantiation of locales in proof contexts:
  1955         instantiate <label>[<attrs>]: <loc>
  1956     Instantiates locale <loc> and adds all its theorems to the current context
  1957     taking into account their attributes.  Label and attrs are optional
  1958     modifiers, like in theorem declarations.  If present, names of
  1959     instantiated theorems are qualified with <label>, and the attributes
  1960     <attrs> are applied after any attributes these theorems might have already.
  1961       If the locale has assumptions, a chained fact of the form
  1962     "<loc> t1 ... tn" is expected from which instantiations of the parameters
  1963     are derived.  The command does not support old-style locales declared
  1964     with "locale (open)".
  1965       A few (very simple) examples can be found in FOL/ex/LocaleInst.thy.
  1966 
  1967 * HOL: Tactic emulation methods induct_tac and case_tac understand static
  1968   (Isar) contexts.
  1969 
  1970 
  1971 *** HOL ***
  1972 
  1973 * Proof import: new image HOL4 contains the imported library from
  1974   the HOL4 system with about 2500 theorems. It is imported by
  1975   replaying proof terms produced by HOL4 in Isabelle. The HOL4 image
  1976   can be used like any other Isabelle image.  See
  1977   HOL/Import/HOL/README for more information.
  1978 
  1979 * Simplifier:
  1980   - Much improved handling of linear and partial orders.
  1981     Reasoners for linear and partial orders are set up for type classes
  1982     "linorder" and "order" respectively, and are added to the default simpset
  1983     as solvers.  This means that the simplifier can build transitivity chains
  1984     to solve goals from the assumptions.
  1985   - INCOMPATIBILITY: old proofs break occasionally.  Typically, applications
  1986     of blast or auto after simplification become unnecessary because the goal
  1987     is solved by simplification already.
  1988 
  1989 * Numerics: new theory Ring_and_Field contains over 250 basic numerical laws,
  1990     all proved in axiomatic type classes for semirings, rings and fields.
  1991 
  1992 * Numerics:
  1993   - Numeric types (nat, int, and in HOL-Complex rat, real, complex, etc.) are
  1994     now formalized using the Ring_and_Field theory mentioned above.
  1995   - INCOMPATIBILITY: simplification and arithmetic behaves somewhat differently
  1996     than before, because now they are set up once in a generic manner.
  1997   - INCOMPATIBILITY: many type-specific arithmetic laws have gone.
  1998     Look for the general versions in Ring_and_Field (and Power if they concern
  1999     exponentiation).
  2000 
  2001 * Type "rat" of the rational numbers is now available in HOL-Complex.
  2002 
  2003 * Records:
  2004   - Record types are now by default printed with their type abbreviation
  2005     instead of the list of all field types. This can be configured via
  2006     the reference "print_record_type_abbr".
  2007   - Simproc "record_upd_simproc" for simplification of multiple updates added
  2008     (not enabled by default).
  2009   - Simproc "record_ex_sel_eq_simproc" to simplify EX x. sel r = x resp.
  2010     EX x. x = sel r to True (not enabled by default).
  2011   - Tactic "record_split_simp_tac" to split and simplify records added.
  2012 
  2013 * 'specification' command added, allowing for definition by
  2014   specification.  There is also an 'ax_specification' command that
  2015   introduces the new constants axiomatically.
  2016 
  2017 * arith(_tac) is now able to generate counterexamples for reals as well.
  2018 
  2019 * HOL-Algebra: new locale "ring" for non-commutative rings.
  2020 
  2021 * HOL-ex: InductiveInvariant_examples illustrates advanced recursive function
  2022   definitions, thanks to Sava Krsti\'{c} and John Matthews.
  2023 
  2024 * HOL-Matrix: a first theory for matrices in HOL with an application of
  2025   matrix theory to linear programming.
  2026 
  2027 * Unions and Intersections:
  2028   The latex output syntax of UN and INT has been changed
  2029   from "\Union x \in A. B" to "\Union_{x \in A} B"
  2030   i.e. the index formulae has become a subscript.
  2031   Similarly for "\Union x. B", and for \Inter instead of \Union.
  2032 
  2033 * Unions and Intersections over Intervals:
  2034   There is new short syntax "UN i<=n. A" for "UN i:{0..n}. A". There is
  2035   also an x-symbol version with subscripts "\<Union>\<^bsub>i <= n\<^esub>. A"
  2036   like in normal math, and corresponding versions for < and for intersection.
  2037 
  2038 * HOL/List: Ordering "lexico" is renamed "lenlex" and the standard
  2039   lexicographic dictonary ordering has been added as "lexord".
  2040 
  2041 * ML: the legacy theory structures Int and List have been removed. They had
  2042   conflicted with ML Basis Library structures having the same names.
  2043 
  2044 * 'refute' command added to search for (finite) countermodels.  Only works
  2045   for a fragment of HOL.  The installation of an external SAT solver is
  2046   highly recommended.  See "HOL/Refute.thy" for details.
  2047 
  2048 * 'quickcheck' command: Allows to find counterexamples by evaluating
  2049   formulae under an assignment of free variables to random values.
  2050   In contrast to 'refute', it can deal with inductive datatypes,
  2051   but cannot handle quantifiers. See "HOL/ex/Quickcheck_Examples.thy"
  2052   for examples.
  2053 
  2054 
  2055 *** HOLCF ***
  2056 
  2057 * Streams now come with concatenation and are part of the HOLCF image
  2058 
  2059 
  2060 
  2061 New in Isabelle2003 (May 2003)
  2062 ------------------------------
  2063 
  2064 *** General ***
  2065 
  2066 * Provers/simplifier:
  2067 
  2068   - Completely reimplemented method simp (ML: Asm_full_simp_tac):
  2069     Assumptions are now subject to complete mutual simplification,
  2070     not just from left to right. The simplifier now preserves
  2071     the order of assumptions.
  2072 
  2073     Potential INCOMPATIBILITY:
  2074 
  2075     -- simp sometimes diverges where the old version did
  2076        not, e.g. invoking simp on the goal
  2077 
  2078         [| P (f x); y = x; f x = f y |] ==> Q
  2079 
  2080        now gives rise to the infinite reduction sequence
  2081 
  2082         P(f x) --(f x = f y)--> P(f y) --(y = x)--> P(f x) --(f x = f y)--> ...
  2083 
  2084        Using "simp (asm_lr)" (ML: Asm_lr_simp_tac) instead often solves this
  2085        kind of problem.
  2086 
  2087     -- Tactics combining classical reasoner and simplification (such as auto)
  2088        are also affected by this change, because many of them rely on
  2089        simp. They may sometimes diverge as well or yield a different numbers
  2090        of subgoals. Try to use e.g. force, fastsimp, or safe instead of auto
  2091        in case of problems. Sometimes subsequent calls to the classical
  2092        reasoner will fail because a preceeding call to the simplifier too
  2093        eagerly simplified the goal, e.g. deleted redundant premises.
  2094 
  2095   - The simplifier trace now shows the names of the applied rewrite rules
  2096 
  2097   - You can limit the number of recursive invocations of the simplifier
  2098     during conditional rewriting (where the simplifie tries to solve the
  2099     conditions before applying the rewrite rule):
  2100     ML "simp_depth_limit := n"
  2101     where n is an integer. Thus you can force termination where previously
  2102     the simplifier would diverge.
  2103 
  2104   - Accepts free variables as head terms in congruence rules.  Useful in Isar.
  2105 
  2106   - No longer aborts on failed congruence proof.  Instead, the
  2107     congruence is ignored.
  2108 
  2109 * Pure: New generic framework for extracting programs from constructive
  2110   proofs. See HOL/Extraction.thy for an example instantiation, as well
  2111   as HOL/Extraction for some case studies.
  2112 
  2113 * Pure: The main goal of the proof state is no longer shown by default, only
  2114 the subgoals. This behaviour is controlled by a new flag.
  2115    PG menu: Isabelle/Isar -> Settings -> Show Main Goal
  2116 (ML: Proof.show_main_goal).
  2117 
  2118 * Pure: You can find all matching introduction rules for subgoal 1, i.e. all
  2119 rules whose conclusion matches subgoal 1:
  2120       PG menu: Isabelle/Isar -> Show me -> matching rules
  2121 The rules are ordered by how closely they match the subgoal.
  2122 In particular, rules that solve a subgoal outright are displayed first
  2123 (or rather last, the way they are printed).
  2124 (ML: ProofGeneral.print_intros())
  2125 
  2126 * Pure: New flag trace_unify_fail causes unification to print
  2127 diagnostic information (PG: in trace buffer) when it fails. This is
  2128 useful for figuring out why single step proofs like rule, erule or
  2129 assumption failed.
  2130 
  2131 * Pure: Locale specifications now produce predicate definitions
  2132 according to the body of text (covering assumptions modulo local
  2133 definitions); predicate "loc_axioms" covers newly introduced text,
  2134 while "loc" is cumulative wrt. all included locale expressions; the
  2135 latter view is presented only on export into the global theory
  2136 context; potential INCOMPATIBILITY, use "(open)" option to fall back
  2137 on the old view without predicates;
  2138 
  2139 * Pure: predefined locales "var" and "struct" are useful for sharing
  2140 parameters (as in CASL, for example); just specify something like
  2141 ``var x + var y + struct M'' as import;
  2142 
  2143 * Pure: improved thms_containing: proper indexing of facts instead of
  2144 raw theorems; check validity of results wrt. current name space;
  2145 include local facts of proof configuration (also covers active
  2146 locales), cover fixed variables in index; may use "_" in term
  2147 specification; an optional limit for the number of printed facts may
  2148 be given (the default is 40);
  2149 
  2150 * Pure: disallow duplicate fact bindings within new-style theory files
  2151 (batch-mode only);
  2152 
  2153 * Provers: improved induct method: assumptions introduced by case
  2154 "foo" are split into "foo.hyps" (from the rule) and "foo.prems" (from
  2155 the goal statement); "foo" still refers to all facts collectively;
  2156 
  2157 * Provers: the function blast.overloaded has been removed: all constants
  2158 are regarded as potentially overloaded, which improves robustness in exchange
  2159 for slight decrease in efficiency;
  2160 
  2161 * Provers/linorder: New generic prover for transitivity reasoning over
  2162 linear orders.  Note: this prover is not efficient!
  2163 
  2164 * Isar: preview of problems to finish 'show' now produce an error
  2165 rather than just a warning (in interactive mode);
  2166 
  2167 
  2168 *** HOL ***
  2169 
  2170 * arith(_tac)
  2171 
  2172  - Produces a counter example if it cannot prove a goal.
  2173    Note that the counter example may be spurious if the goal is not a formula
  2174    of quantifier-free linear arithmetic.
  2175    In ProofGeneral the counter example appears in the trace buffer.
  2176 
  2177  - Knows about div k and mod k where k is a numeral of type nat or int.
  2178 
  2179  - Calls full Presburger arithmetic (by Amine Chaieb) if quantifier-free
  2180    linear arithmetic fails. This takes account of quantifiers and divisibility.
  2181    Presburger arithmetic can also be called explicitly via presburger(_tac).
  2182 
  2183 * simp's arithmetic capabilities have been enhanced a bit: it now
  2184 takes ~= in premises into account (by performing a case split);
  2185 
  2186 * simp reduces "m*(n div m) + n mod m" to n, even if the two summands
  2187 are distributed over a sum of terms;
  2188 
  2189 * New tactic "trans_tac" and method "trans" instantiate
  2190 Provers/linorder.ML for axclasses "order" and "linorder" (predicates
  2191 "<=", "<" and "=").
  2192 
  2193 * function INCOMPATIBILITIES: Pi-sets have been redefined and moved from main
  2194 HOL to Library/FuncSet; constant "Fun.op o" is now called "Fun.comp";
  2195 
  2196 * 'typedef' command has new option "open" to suppress the set
  2197 definition;
  2198 
  2199 * functions Min and Max on finite sets have been introduced (theory
  2200 Finite_Set);
  2201 
  2202 * attribute [symmetric] now works for relations as well; it turns
  2203 (x,y) : R^-1 into (y,x) : R, and vice versa;
  2204 
  2205 * induct over a !!-quantified statement (say !!x1..xn):
  2206   each "case" automatically performs "fix x1 .. xn" with exactly those names.
  2207 
  2208 * Map: `empty' is no longer a constant but a syntactic abbreviation for
  2209 %x. None. Warning: empty_def now refers to the previously hidden definition
  2210 of the empty set.
  2211 
  2212 * Algebra: formalization of classical algebra.  Intended as base for
  2213 any algebraic development in Isabelle.  Currently covers group theory
  2214 (up to Sylow's theorem) and ring theory (Universal Property of
  2215 Univariate Polynomials).  Contributions welcome;
  2216 
  2217 * GroupTheory: deleted, since its material has been moved to Algebra;
  2218 
  2219 * Complex: new directory of the complex numbers with numeric constants,
  2220 nonstandard complex numbers, and some complex analysis, standard and
  2221 nonstandard (Jacques Fleuriot);
  2222 
  2223 * HOL-Complex: new image for analysis, replacing HOL-Real and HOL-Hyperreal;
  2224 
  2225 * Hyperreal: introduced Gauge integration and hyperreal logarithms (Jacques
  2226 Fleuriot);
  2227 
  2228 * Real/HahnBanach: updated and adapted to locales;
  2229 
  2230 * NumberTheory: added Gauss's law of quadratic reciprocity (by Avigad,
  2231 Gray and Kramer);
  2232 
  2233 * UNITY: added the Meier-Sanders theory of progress sets;
  2234 
  2235 * MicroJava: bytecode verifier and lightweight bytecode verifier
  2236 as abstract algorithms, instantiated to the JVM;
  2237 
  2238 * Bali: Java source language formalization. Type system, operational
  2239 semantics, axiomatic semantics. Supported language features:
  2240 classes, interfaces, objects,virtual methods, static methods,
  2241 static/instance fields, arrays, access modifiers, definite
  2242 assignment, exceptions.
  2243 
  2244 
  2245 *** ZF ***
  2246 
  2247 * ZF/Constructible: consistency proof for AC (Gdel's constructible
  2248 universe, etc.);
  2249 
  2250 * Main ZF: virtually all theories converted to new-style format;
  2251 
  2252 
  2253 *** ML ***
  2254 
  2255 * Pure: Tactic.prove provides sane interface for internal proofs;
  2256 omits the infamous "standard" operation, so this is more appropriate
  2257 than prove_goalw_cterm in many situations (e.g. in simprocs);
  2258 
  2259 * Pure: improved error reporting of simprocs;
  2260 
  2261 * Provers: Simplifier.simproc(_i) provides sane interface for setting
  2262 up simprocs;
  2263 
  2264 
  2265 *** Document preparation ***
  2266 
  2267 * uses \par instead of \\ for line breaks in theory text. This may
  2268 shift some page breaks in large documents. To get the old behaviour
  2269 use \renewcommand{\isanewline}{\mbox{}\\\mbox{}} in root.tex.
  2270 
  2271 * minimized dependencies of isabelle.sty and isabellesym.sty on
  2272 other packages
  2273 
  2274 * \<euro> now needs package babel/greek instead of marvosym (which
  2275 broke \Rightarrow)
  2276 
  2277 * normal size for \<zero>...\<nine> (uses \mathbf instead of
  2278 textcomp package)
  2279 
  2280 
  2281 
  2282 New in Isabelle2002 (March 2002)
  2283 --------------------------------
  2284 
  2285 *** Document preparation ***
  2286 
  2287 * greatly simplified document preparation setup, including more
  2288 graceful interpretation of isatool usedir -i/-d/-D options, and more
  2289 instructive isatool mkdir; users should basically be able to get
  2290 started with "isatool mkdir HOL Test && isatool make"; alternatively,
  2291 users may run a separate document processing stage manually like this:
  2292 "isatool usedir -D output HOL Test && isatool document Test/output";
  2293 
  2294 * theory dependency graph may now be incorporated into documents;
  2295 isatool usedir -g true will produce session_graph.eps/.pdf for use
  2296 with \includegraphics of LaTeX;
  2297 
  2298 * proper spacing of consecutive markup elements, especially text
  2299 blocks after section headings;
  2300 
  2301 * support bold style (for single symbols only), input syntax is like
  2302 this: "\<^bold>\<alpha>" or "\<^bold>A";
  2303 
  2304 * \<bullet> is now output as bold \cdot by default, which looks much
  2305 better in printed text;
  2306 
  2307 * added default LaTeX bindings for \<tturnstile> and \<TTurnstile>;
  2308 note that these symbols are currently unavailable in Proof General /
  2309 X-Symbol; new symbols \<zero>, \<one>, ..., \<nine>, and \<euro>;
  2310 
  2311 * isatool latex no longer depends on changed TEXINPUTS, instead
  2312 isatool document copies the Isabelle style files to the target
  2313 location;
  2314 
  2315 
  2316 *** Isar ***
  2317 
  2318 * Pure/Provers: improved proof by cases and induction;
  2319   - 'case' command admits impromptu naming of parameters (such as
  2320     "case (Suc n)");
  2321   - 'induct' method divinates rule instantiation from the inductive
  2322     claim; no longer requires excessive ?P bindings for proper
  2323     instantiation of cases;
  2324   - 'induct' method properly enumerates all possibilities of set/type
  2325     rules; as a consequence facts may be also passed through *type*
  2326     rules without further ado;
  2327   - 'induct' method now derives symbolic cases from the *rulified*
  2328     rule (before it used to rulify cases stemming from the internal
  2329     atomized version); this means that the context of a non-atomic
  2330     statement becomes is included in the hypothesis, avoiding the
  2331     slightly cumbersome show "PROP ?case" form;
  2332   - 'induct' may now use elim-style induction rules without chaining
  2333     facts, using ``missing'' premises from the goal state; this allows
  2334     rules stemming from inductive sets to be applied in unstructured
  2335     scripts, while still benefitting from proper handling of non-atomic
  2336     statements; NB: major inductive premises need to be put first, all
  2337     the rest of the goal is passed through the induction;
  2338   - 'induct' proper support for mutual induction involving non-atomic
  2339     rule statements (uses the new concept of simultaneous goals, see
  2340     below);
  2341   - append all possible rule selections, but only use the first
  2342     success (no backtracking);
  2343   - removed obsolete "(simplified)" and "(stripped)" options of methods;
  2344   - undeclared rule case names default to numbers 1, 2, 3, ...;
  2345   - added 'print_induct_rules' (covered by help item in recent Proof
  2346     General versions);
  2347   - moved induct/cases attributes to Pure, methods to Provers;
  2348   - generic method setup instantiated for FOL and HOL;
  2349 
  2350 * Pure: support multiple simultaneous goal statements, for example
  2351 "have a: A and b: B" (same for 'theorem' etc.); being a pure
  2352 meta-level mechanism, this acts as if several individual goals had
  2353 been stated separately; in particular common proof methods need to be
  2354 repeated in order to cover all claims; note that a single elimination
  2355 step is *not* sufficient to establish the two conjunctions, so this
  2356 fails:
  2357 
  2358   assume "A & B" then have A and B ..   (*".." fails*)
  2359 
  2360 better use "obtain" in situations as above; alternative refer to
  2361 multi-step methods like 'auto', 'simp_all', 'blast+' etc.;
  2362 
  2363 * Pure: proper integration with ``locales''; unlike the original
  2364 version by Florian Kammller, Isar locales package high-level proof
  2365 contexts rather than raw logical ones (e.g. we admit to include
  2366 attributes everywhere); operations on locales include merge and
  2367 rename; support for implicit arguments (``structures''); simultaneous
  2368 type-inference over imports and text; see also HOL/ex/Locales.thy for
  2369 some examples;
  2370 
  2371 * Pure: the following commands have been ``localized'', supporting a
  2372 target locale specification "(in name)": 'lemma', 'theorem',
  2373 'corollary', 'lemmas', 'theorems', 'declare'; the results will be
  2374 stored both within the locale and at the theory level (exported and
  2375 qualified by the locale name);
  2376 
  2377 * Pure: theory goals may now be specified in ``long'' form, with
  2378 ad-hoc contexts consisting of arbitrary locale elements. for example
  2379 ``lemma foo: fixes x assumes "A x" shows "B x"'' (local syntax and
  2380 definitions may be given, too); the result is a meta-level rule with
  2381 the context elements being discharged in the obvious way;
  2382 
  2383 * Pure: new proof command 'using' allows to augment currently used
  2384 facts after a goal statement ('using' is syntactically analogous to
  2385 'apply', but acts on the goal's facts only); this allows chained facts
  2386 to be separated into parts given before and after a claim, as in
  2387 ``from a and b have C using d and e <proof>'';
  2388 
  2389 * Pure: renamed "antecedent" case to "rule_context";
  2390 
  2391 * Pure: new 'judgment' command records explicit information about the
  2392 object-logic embedding (used by several tools internally); no longer
  2393 use hard-wired "Trueprop";
  2394 
  2395 * Pure: added 'corollary' command;
  2396 
  2397 * Pure: fixed 'token_translation' command;
  2398 
  2399 * Pure: removed obsolete 'exported' attribute;
  2400 
  2401 * Pure: dummy pattern "_" in is/let is now automatically lifted over
  2402 bound variables: "ALL x. P x --> Q x" (is "ALL x. _ --> ?C x")
  2403 supersedes more cumbersome ... (is "ALL x. _ x --> ?C x");
  2404 
  2405 * Pure: method 'atomize' presents local goal premises as object-level
  2406 statements (atomic meta-level propositions); setup controlled via
  2407 rewrite rules declarations of 'atomize' attribute; example
  2408 application: 'induct' method with proper rule statements in improper
  2409 proof *scripts*;
  2410 
  2411 * Pure: emulation of instantiation tactics (rule_tac, cut_tac, etc.)
  2412 now consider the syntactic context of assumptions, giving a better
  2413 chance to get type-inference of the arguments right (this is
  2414 especially important for locales);
  2415 
  2416 * Pure: "sorry" no longer requires quick_and_dirty in interactive
  2417 mode;
  2418 
  2419 * Pure/obtain: the formal conclusion "thesis", being marked as
  2420 ``internal'', may no longer be reference directly in the text;
  2421 potential INCOMPATIBILITY, may need to use "?thesis" in rare
  2422 situations;
  2423 
  2424 * Pure: generic 'sym' attribute which declares a rule both as pure
  2425 'elim?' and for the 'symmetric' operation;
  2426 
  2427 * Pure: marginal comments ``--'' may now occur just anywhere in the
  2428 text; the fixed correlation with particular command syntax has been
  2429 discontinued;
  2430 
  2431 * Pure: new method 'rules' is particularly well-suited for proof
  2432 search in intuitionistic logic; a bit slower than 'blast' or 'fast',
  2433 but often produces more compact proof terms with less detours;
  2434 
  2435 * Pure/Provers/classical: simplified integration with pure rule
  2436 attributes and methods; the classical "intro?/elim?/dest?"
  2437 declarations coincide with the pure ones; the "rule" method no longer
  2438 includes classically swapped intros; "intro" and "elim" methods no
  2439 longer pick rules from the context; also got rid of ML declarations
  2440 AddXIs/AddXEs/AddXDs; all of this has some potential for
  2441 INCOMPATIBILITY;
  2442 
  2443 * Provers/classical: attribute 'swapped' produces classical inversions
  2444 of introduction rules;
  2445 
  2446 * Provers/simplifier: 'simplified' attribute may refer to explicit
  2447 rules instead of full simplifier context; 'iff' attribute handles
  2448 conditional rules;
  2449 
  2450 * HOL: 'typedef' now allows alternative names for Rep/Abs morphisms;
  2451 
  2452 * HOL: 'recdef' now fails on unfinished automated proofs, use
  2453 "(permissive)" option to recover old behavior;
  2454 
  2455 * HOL: 'inductive' no longer features separate (collective) attributes
  2456 for 'intros' (was found too confusing);
  2457 
  2458 * HOL: properly declared induction rules less_induct and
  2459 wf_induct_rule;
  2460 
  2461 
  2462 *** HOL ***
  2463 
  2464 * HOL: moved over to sane numeral syntax; the new policy is as
  2465 follows:
  2466 
  2467   - 0 and 1 are polymorphic constants, which are defined on any
  2468   numeric type (nat, int, real etc.);
  2469 
  2470   - 2, 3, 4, ... and -1, -2, -3, ... are polymorphic numerals, based
  2471   binary representation internally;
  2472 
  2473   - type nat has special constructor Suc, and generally prefers Suc 0
  2474   over 1::nat and Suc (Suc 0) over 2::nat;
  2475 
  2476 This change may cause significant problems of INCOMPATIBILITY; here
  2477 are some hints on converting existing sources:
  2478 
  2479   - due to the new "num" token, "-0" and "-1" etc. are now atomic
  2480   entities, so expressions involving "-" (unary or binary minus) need
  2481   to be spaced properly;
  2482 
  2483   - existing occurrences of "1" may need to be constraint "1::nat" or
  2484   even replaced by Suc 0; similar for old "2";
  2485 
  2486   - replace "#nnn" by "nnn", and "#-nnn" by "-nnn";
  2487 
  2488   - remove all special provisions on numerals in proofs;
  2489 
  2490 * HOL: simp rules nat_number expand numerals on nat to Suc/0
  2491 representation (depends on bin_arith_simps in the default context);
  2492 
  2493 * HOL: symbolic syntax for x^2 (numeral 2);
  2494 
  2495 * HOL: the class of all HOL types is now called "type" rather than
  2496 "term"; INCOMPATIBILITY, need to adapt references to this type class
  2497 in axclass/classes, instance/arities, and (usually rare) occurrences
  2498 in typings (of consts etc.); internally the class is called
  2499 "HOL.type", ML programs should refer to HOLogic.typeS;
  2500 
  2501 * HOL/record package improvements:
  2502   - new derived operations "fields" to build a partial record section,
  2503     "extend" to promote a fixed record to a record scheme, and
  2504     "truncate" for the reverse; cf. theorems "xxx.defs", which are *not*
  2505     declared as simp by default;
  2506   - shared operations ("more", "fields", etc.) now need to be always
  2507     qualified) --- potential INCOMPATIBILITY;
  2508   - removed "make_scheme" operations (use "make" with "extend") --
  2509     INCOMPATIBILITY;
  2510   - removed "more" class (simply use "term") -- INCOMPATIBILITY;
  2511   - provides cases/induct rules for use with corresponding Isar
  2512     methods (for concrete records, record schemes, concrete more
  2513     parts, and schematic more parts -- in that order);
  2514   - internal definitions directly based on a light-weight abstract
  2515     theory of product types over typedef rather than datatype;
  2516 
  2517 * HOL: generic code generator for generating executable ML code from
  2518 specifications; specific support for HOL constructs such as inductive
  2519 datatypes and sets, as well as recursive functions; can be invoked
  2520 via 'generate_code' theory section;
  2521 
  2522 * HOL: canonical cases/induct rules for n-tuples (n = 3..7);
  2523 
  2524 * HOL: consolidated and renamed several theories.  In particular:
  2525         Ord.thy has been absorbed into HOL.thy
  2526         String.thy has been absorbed into List.thy
  2527 
  2528 * HOL: concrete setsum syntax "\<Sum>i:A. b" == "setsum (%i. b) A"
  2529 (beware of argument permutation!);
  2530 
  2531 * HOL: linorder_less_split superseded by linorder_cases;
  2532 
  2533 * HOL/List: "nodups" renamed to "distinct";
  2534 
  2535 * HOL: added "The" definite description operator; move Hilbert's "Eps"
  2536 to peripheral theory "Hilbert_Choice"; some INCOMPATIBILITIES:
  2537   - Ex_def has changed, now need to use some_eq_ex
  2538 
  2539 * HOL: made split_all_tac safe; EXISTING PROOFS MAY FAIL OR LOOP, so
  2540 in this (rare) case use:
  2541 
  2542   delSWrapper "split_all_tac"
  2543   addSbefore ("unsafe_split_all_tac", unsafe_split_all_tac)
  2544 
  2545 * HOL: added safe wrapper "split_conv_tac" to claset; EXISTING PROOFS
  2546 MAY FAIL;
  2547 
  2548 * HOL: introduced f^n = f o ... o f; warning: due to the limits of
  2549 Isabelle's type classes, ^ on functions and relations has too general
  2550 a domain, namely ('a * 'b) set and 'a => 'b; this means that it may be
  2551 necessary to attach explicit type constraints;
  2552 
  2553 * HOL/Relation: the prefix name of the infix "O" has been changed from
  2554 "comp" to "rel_comp"; INCOMPATIBILITY: a few theorems have been
  2555 renamed accordingly (eg "compI" -> "rel_compI").
  2556 
  2557 * HOL: syntax translations now work properly with numerals and records
  2558 expressions;
  2559 
  2560 * HOL: bounded abstraction now uses syntax "%" / "\<lambda>" instead
  2561 of "lam" -- INCOMPATIBILITY;
  2562 
  2563 * HOL: got rid of some global declarations (potential INCOMPATIBILITY
  2564 for ML tools): const "()" renamed "Product_Type.Unity", type "unit"
  2565 renamed "Product_Type.unit";
  2566 
  2567 * HOL: renamed rtrancl_into_rtrancl2 to converse_rtrancl_into_rtrancl
  2568 
  2569 * HOL: removed obsolete theorem "optionE" (use "option.exhaust", or
  2570 the "cases" method);
  2571 
  2572 * HOL/GroupTheory: group theory examples including Sylow's theorem (by
  2573 Florian Kammller);
  2574 
  2575 * HOL/IMP: updated and converted to new-style theory format; several
  2576 parts turned into readable document, with proper Isar proof texts and
  2577 some explanations (by Gerwin Klein);
  2578 
  2579 * HOL-Real: added Complex_Numbers (by Gertrud Bauer);
  2580 
  2581 * HOL-Hyperreal is now a logic image;
  2582 
  2583 
  2584 *** HOLCF ***
  2585 
  2586 * Isar: consts/constdefs supports mixfix syntax for continuous
  2587 operations;
  2588 
  2589 * Isar: domain package adapted to new-style theory format, e.g. see
  2590 HOLCF/ex/Dnat.thy;
  2591 
  2592 * theory Lift: proper use of rep_datatype lift instead of ML hacks --
  2593 potential INCOMPATIBILITY; now use plain induct_tac instead of former
  2594 lift.induct_tac, always use UU instead of Undef;
  2595 
  2596 * HOLCF/IMP: updated and converted to new-style theory;
  2597 
  2598 
  2599 *** ZF ***
  2600 
  2601 * Isar: proper integration of logic-specific tools and packages,
  2602 including theory commands '(co)inductive', '(co)datatype',
  2603 'rep_datatype', 'inductive_cases', as well as methods 'ind_cases',
  2604 'induct_tac', 'case_tac', and 'typecheck' (with attribute 'TC');
  2605 
  2606 * theory Main no longer includes AC; for the Axiom of Choice, base
  2607 your theory on Main_ZFC;
  2608 
  2609 * the integer library now covers quotients and remainders, with many
  2610 laws relating division to addition, multiplication, etc.;
  2611 
  2612 * ZF/UNITY: Chandy and Misra's UNITY is now available in ZF, giving a
  2613 typeless version of the formalism;
  2614 
  2615 * ZF/AC, Coind, IMP, Resid: updated and converted to new-style theory
  2616 format;
  2617 
  2618 * ZF/Induct: new directory for examples of inductive definitions,
  2619 including theory Multiset for multiset orderings; converted to
  2620 new-style theory format;
  2621 
  2622 * ZF: many new theorems about lists, ordinals, etc.;
  2623 
  2624 
  2625 *** General ***
  2626 
  2627 * Pure/kernel: meta-level proof terms (by Stefan Berghofer); reference
  2628 variable proof controls level of detail: 0 = no proofs (only oracle
  2629 dependencies), 1 = lemma dependencies, 2 = compact proof terms; see
  2630 also ref manual for further ML interfaces;
  2631 
  2632 * Pure/axclass: removed obsolete ML interface
  2633 goal_subclass/goal_arity;
  2634 
  2635 * Pure/syntax: new token syntax "num" for plain numerals (without "#"
  2636 of "xnum"); potential INCOMPATIBILITY, since -0, -1 etc. are now
  2637 separate tokens, so expressions involving minus need to be spaced
  2638 properly;
  2639 
  2640 * Pure/syntax: support non-oriented infixes, using keyword "infix"
  2641 rather than "infixl" or "infixr";
  2642 
  2643 * Pure/syntax: concrete syntax for dummy type variables admits genuine
  2644 sort constraint specifications in type inference; e.g. "x::_::foo"
  2645 ensures that the type of "x" is of sort "foo" (but not necessarily a
  2646 type variable);
  2647 
  2648 * Pure/syntax: print modes "type_brackets" and "no_type_brackets"
  2649 control output of nested => (types); the default behavior is
  2650 "type_brackets";
  2651 
  2652 * Pure/syntax: builtin parse translation for "_constify" turns valued
  2653 tokens into AST constants;
  2654 
  2655 * Pure/syntax: prefer later declarations of translations and print
  2656 translation functions; potential INCOMPATIBILITY: need to reverse
  2657 multiple declarations for same syntax element constant;
  2658 
  2659 * Pure/show_hyps reset by default (in accordance to existing Isar
  2660 practice);
  2661 
  2662 * Provers/classical: renamed addaltern to addafter, addSaltern to
  2663 addSafter;
  2664 
  2665 * Provers/clasimp: ``iff'' declarations now handle conditional rules
  2666 as well;
  2667 
  2668 * system: tested support for MacOS X; should be able to get Isabelle +
  2669 Proof General to work in a plain Terminal after installing Poly/ML
  2670 (e.g. from the Isabelle distribution area) and GNU bash alone
  2671 (e.g. from http://www.apple.com); full X11, XEmacs and X-Symbol
  2672 support requires further installations, e.g. from
  2673 http://fink.sourceforge.net/);
  2674 
  2675 * system: support Poly/ML 4.1.1 (able to manage larger heaps);
  2676 
  2677 * system: reduced base memory usage by Poly/ML (approx. 20 MB instead
  2678 of 40 MB), cf. ML_OPTIONS;
  2679 
  2680 * system: Proof General keywords specification is now part of the
  2681 Isabelle distribution (see etc/isar-keywords.el);
  2682 
  2683 * system: support for persistent Proof General sessions (refrain from
  2684 outdating all loaded theories on startup); user may create writable
  2685 logic images like this: ``isabelle -q HOL Test'';
  2686 
  2687 * system: smart selection of Isabelle process versus Isabelle
  2688 interface, accommodates case-insensitive file systems (e.g. HFS+); may
  2689 run both "isabelle" and "Isabelle" even if file names are badly
  2690 damaged (executable inspects the case of the first letter of its own
  2691 name); added separate "isabelle-process" and "isabelle-interface";
  2692 
  2693 * system: refrain from any attempt at filtering input streams; no
  2694 longer support ``8bit'' encoding of old isabelle font, instead proper
  2695 iso-latin characters may now be used; the related isatools
  2696 "symbolinput" and "nonascii" have disappeared as well;
  2697 
  2698 * system: removed old "xterm" interface (the print modes "xterm" and
  2699 "xterm_color" are still available for direct use in a suitable
  2700 terminal);
  2701 
  2702 
  2703 
  2704 New in Isabelle99-2 (February 2001)
  2705 -----------------------------------
  2706 
  2707 *** Overview of INCOMPATIBILITIES ***
  2708 
  2709 * HOL: please note that theories in the Library and elsewhere often use the
  2710 new-style (Isar) format; to refer to their theorems in an ML script you must
  2711 bind them to ML identifers by e.g.      val thm_name = thm "thm_name";
  2712 
  2713 * HOL: inductive package no longer splits induction rule aggressively,
  2714 but only as far as specified by the introductions given; the old
  2715 format may be recovered via ML function complete_split_rule or attribute
  2716 'split_rule (complete)';
  2717 
  2718 * HOL: induct renamed to lfp_induct, lfp_Tarski to lfp_unfold,
  2719 gfp_Tarski to gfp_unfold;
  2720 
  2721 * HOL: contrapos, contrapos2 renamed to contrapos_nn, contrapos_pp;
  2722 
  2723 * HOL: infix "dvd" now has priority 50 rather than 70 (because it is a
  2724 relation); infix "^^" has been renamed "``"; infix "``" has been
  2725 renamed "`"; "univalent" has been renamed "single_valued";
  2726 
  2727 * HOL/Real: "rinv" and "hrinv" replaced by overloaded "inverse"
  2728 operation;
  2729 
  2730 * HOLCF: infix "`" has been renamed "$"; the symbol syntax is \<cdot>;
  2731 
  2732 * Isar: 'obtain' no longer declares "that" fact as simp/intro;
  2733 
  2734 * Isar/HOL: method 'induct' now handles non-atomic goals; as a
  2735 consequence, it is no longer monotonic wrt. the local goal context
  2736 (which is now passed through the inductive cases);
  2737 
  2738 * Document preparation: renamed standard symbols \<ll> to \<lless> and
  2739 \<gg> to \<ggreater>;
  2740 
  2741 
  2742 *** Document preparation ***
  2743 
  2744 * \isabellestyle{NAME} selects version of Isabelle output (currently
  2745 available: are "it" for near math-mode best-style output, "sl" for
  2746 slanted text style, and "tt" for plain type-writer; if no
  2747 \isabellestyle command is given, output is according to slanted
  2748 type-writer);
  2749 
  2750 * support sub/super scripts (for single symbols only), input syntax is
  2751 like this: "A\<^sup>*" or "A\<^sup>\<star>";
  2752 
  2753 * some more standard symbols; see Appendix A of the system manual for
  2754 the complete list of symbols defined in isabellesym.sty;
  2755 
  2756 * improved isabelle style files; more abstract symbol implementation
  2757 (should now use \isamath{...} and \isatext{...} in custom symbol
  2758 definitions);
  2759 
  2760 * antiquotation @{goals} and @{subgoals} for output of *dynamic* goals
  2761 state; Note that presentation of goal states does not conform to
  2762 actual human-readable proof documents.  Please do not include goal
  2763 states into document output unless you really know what you are doing!
  2764 
  2765 * proper indentation of antiquoted output with proportional LaTeX
  2766 fonts;
  2767 
  2768 * no_document ML operator temporarily disables LaTeX document
  2769 generation;
  2770 
  2771 * isatool unsymbolize tunes sources for plain ASCII communication;
  2772 
  2773 
  2774 *** Isar ***
  2775 
  2776 * Pure: Isar now suffers initial goal statements to contain unbound
  2777 schematic variables (this does not conform to actual readable proof
  2778 documents, due to unpredictable outcome and non-compositional proof
  2779 checking); users who know what they are doing may use schematic goals
  2780 for Prolog-style synthesis of proven results;
  2781 
  2782 * Pure: assumption method (an implicit finishing) now handles actual
  2783 rules as well;
  2784 
  2785 * Pure: improved 'obtain' --- moved to Pure, insert "that" into
  2786 initial goal, declare "that" only as Pure intro (only for single
  2787 steps); the "that" rule assumption may now be involved in implicit
  2788 finishing, thus ".." becomes a feasible for trivial obtains;
  2789 
  2790 * Pure: default proof step now includes 'intro_classes'; thus trivial
  2791 instance proofs may be performed by "..";
  2792 
  2793 * Pure: ?thesis / ?this / "..." now work for pure meta-level
  2794 statements as well;
  2795 
  2796 * Pure: more robust selection of calculational rules;
  2797 
  2798 * Pure: the builtin notion of 'finished' goal now includes the ==-refl
  2799 rule (as well as the assumption rule);
  2800 
  2801 * Pure: 'thm_deps' command visualizes dependencies of theorems and
  2802 lemmas, using the graph browser tool;
  2803 
  2804 * Pure: predict failure of "show" in interactive mode;
  2805 
  2806 * Pure: 'thms_containing' now takes actual terms as arguments;
  2807 
  2808 * HOL: improved method 'induct' --- now handles non-atomic goals
  2809 (potential INCOMPATIBILITY); tuned error handling;
  2810 
  2811 * HOL: cases and induct rules now provide explicit hints about the
  2812 number of facts to be consumed (0 for "type" and 1 for "set" rules);
  2813 any remaining facts are inserted into the goal verbatim;
  2814 
  2815 * HOL: local contexts (aka cases) may now contain term bindings as
  2816 well; the 'cases' and 'induct' methods new provide a ?case binding for
  2817 the result to be shown in each case;
  2818 
  2819 * HOL: added 'recdef_tc' command;
  2820 
  2821 * isatool convert assists in eliminating legacy ML scripts;
  2822 
  2823 
  2824 *** HOL ***
  2825 
  2826 * HOL/Library: a collection of generic theories to be used together
  2827 with main HOL; the theory loader path already includes this directory
  2828 by default; the following existing theories have been moved here:
  2829 HOL/Induct/Multiset, HOL/Induct/Acc (as Accessible_Part), HOL/While
  2830 (as While_Combinator), HOL/Lex/Prefix (as List_Prefix);
  2831 
  2832 * HOL/Unix: "Some aspects of Unix file-system security", a typical
  2833 modelling and verification task performed in Isabelle/HOL +
  2834 Isabelle/Isar + Isabelle document preparation (by Markus Wenzel).
  2835 
  2836 * HOL/Algebra: special summation operator SUM no longer exists, it has
  2837 been replaced by setsum; infix 'assoc' now has priority 50 (like
  2838 'dvd'); axiom 'one_not_zero' has been moved from axclass 'ring' to
  2839 'domain', this makes the theory consistent with mathematical
  2840 literature;
  2841 
  2842 * HOL basics: added overloaded operations "inverse" and "divide"
  2843 (infix "/"), syntax for generic "abs" operation, generic summation
  2844 operator \<Sum>;
  2845 
  2846 * HOL/typedef: simplified package, provide more useful rules (see also
  2847 HOL/subset.thy);
  2848 
  2849 * HOL/datatype: induction rule for arbitrarily branching datatypes is
  2850 now expressed as a proper nested rule (old-style tactic scripts may
  2851 require atomize_strip_tac to cope with non-atomic premises);
  2852 
  2853 * HOL: renamed theory "Prod" to "Product_Type", renamed "split" rule
  2854 to "split_conv" (old name still available for compatibility);
  2855 
  2856 * HOL: improved concrete syntax for strings (e.g. allows translation
  2857 rules with string literals);
  2858 
  2859 * HOL-Real-Hyperreal: this extends HOL-Real with the hyperreals
  2860  and Fleuriot's mechanization of analysis, including the transcendental
  2861  functions for the reals;
  2862 
  2863 * HOL/Real, HOL/Hyperreal: improved arithmetic simplification;
  2864 
  2865 
  2866 *** CTT ***
  2867 
  2868 * CTT: x-symbol support for Pi, Sigma, -->, : (membership); note that
  2869 "lam" is displayed as TWO lambda-symbols
  2870 
  2871 * CTT: theory Main now available, containing everything (that is, Bool
  2872 and Arith);
  2873 
  2874 
  2875 *** General ***
  2876 
  2877 * Pure: the Simplifier has been implemented properly as a derived rule
  2878 outside of the actual kernel (at last!); the overall performance
  2879 penalty in practical applications is about 50%, while reliability of
  2880 the Isabelle inference kernel has been greatly improved;
  2881 
  2882 * print modes "brackets" and "no_brackets" control output of nested =>
  2883 (types) and ==> (props); the default behaviour is "brackets";
  2884 
  2885 * Provers: fast_tac (and friends) now handle actual object-logic rules
  2886 as assumptions as well;
  2887 
  2888 * system: support Poly/ML 4.0;
  2889 
  2890 * system: isatool install handles KDE version 1 or 2;
  2891 
  2892 
  2893 
  2894 New in Isabelle99-1 (October 2000)
  2895 ----------------------------------
  2896 
  2897 *** Overview of INCOMPATIBILITIES ***
  2898 
  2899 * HOL: simplification of natural numbers is much changed; to partly
  2900 recover the old behaviour (e.g. to prevent n+n rewriting to #2*n)
  2901 issue the following ML commands:
  2902 
  2903   Delsimprocs Nat_Numeral_Simprocs.cancel_numerals;
  2904   Delsimprocs [Nat_Numeral_Simprocs.combine_numerals];
  2905 
  2906 * HOL: simplification no longer dives into case-expressions; this is
  2907 controlled by "t.weak_case_cong" for each datatype t;
  2908 
  2909 * HOL: nat_less_induct renamed to less_induct;
  2910 
  2911 * HOL: systematic renaming of the SOME (Eps) rules, may use isatool
  2912 fixsome to patch .thy and .ML sources automatically;
  2913 
  2914   select_equality  -> some_equality
  2915   select_eq_Ex     -> some_eq_ex
  2916   selectI2EX       -> someI2_ex
  2917   selectI2         -> someI2
  2918   selectI          -> someI
  2919   select1_equality -> some1_equality
  2920   Eps_sym_eq       -> some_sym_eq_trivial
  2921   Eps_eq           -> some_eq_trivial
  2922 
  2923 * HOL: exhaust_tac on datatypes superceded by new generic case_tac;
  2924 
  2925 * HOL: removed obsolete theorem binding expand_if (refer to split_if
  2926 instead);
  2927 
  2928 * HOL: the recursion equations generated by 'recdef' are now called
  2929 f.simps instead of f.rules;
  2930 
  2931 * HOL: qed_spec_mp now also handles bounded ALL as well;
  2932 
  2933 * HOL: 0 is now overloaded, so the type constraint ":: nat" may
  2934 sometimes be needed;
  2935 
  2936 * HOL: the constant for "f``x" is now "image" rather than "op ``";
  2937 
  2938 * HOL: the constant for "f-``x" is now "vimage" rather than "op -``";
  2939 
  2940 * HOL: the disjoint sum is now "<+>" instead of "Plus"; the cartesian
  2941 product is now "<*>" instead of "Times"; the lexicographic product is
  2942 now "<*lex*>" instead of "**";
  2943 
  2944 * HOL: theory Sexp is now in HOL/Induct examples (it used to be part
  2945 of main HOL, but was unused); better use HOL's datatype package;
  2946 
  2947 * HOL: removed "symbols" syntax for constant "override" of theory Map;
  2948 the old syntax may be recovered as follows:
  2949 
  2950   syntax (symbols)
  2951     override  :: "('a ~=> 'b) => ('a ~=> 'b) => ('a ~=> 'b)"
  2952       (infixl "\\<oplus>" 100)
  2953 
  2954 * HOL/Real: "rabs" replaced by overloaded "abs" function;
  2955 
  2956 * HOL/ML: even fewer consts are declared as global (see theories Ord,
  2957 Lfp, Gfp, WF); this only affects ML packages that refer to const names
  2958 internally;
  2959 
  2960 * HOL and ZF: syntax for quotienting wrt an equivalence relation
  2961 changed from A/r to A//r;
  2962 
  2963 * ZF: new treatment of arithmetic (nat & int) may break some old
  2964 proofs;
  2965 
  2966 * Isar: renamed some attributes (RS -> THEN, simplify -> simplified,
  2967 rulify -> rule_format, elimify -> elim_format, ...);
  2968 
  2969 * Isar/Provers: intro/elim/dest attributes changed; renamed
  2970 intro/intro!/intro!! flags to intro!/intro/intro? (in most cases, one
  2971 should have to change intro!! to intro? only); replaced "delrule" by
  2972 "rule del";
  2973 
  2974 * Isar/HOL: renamed "intrs" to "intros" in inductive definitions;
  2975 
  2976 * Provers: strengthened force_tac by using new first_best_tac;
  2977 
  2978 * LaTeX document preparation: several changes of isabelle.sty (see
  2979 lib/texinputs);
  2980 
  2981 
  2982 *** Document preparation ***
  2983 
  2984 * formal comments (text blocks etc.) in new-style theories may now
  2985 contain antiquotations of thm/prop/term/typ/text to be presented
  2986 according to latex print mode; concrete syntax is like this:
  2987 @{term[show_types] "f(x) = a + x"};
  2988 
  2989 * isatool mkdir provides easy setup of Isabelle session directories,
  2990 including proper document sources;
  2991 
  2992 * generated LaTeX sources are now deleted after successful run
  2993 (isatool document -c); may retain a copy somewhere else via -D option
  2994 of isatool usedir;
  2995 
  2996 * isatool usedir -D now lets isatool latex -o sty update the Isabelle
  2997 style files, achieving self-contained LaTeX sources and simplifying
  2998 LaTeX debugging;
  2999 
  3000 * old-style theories now produce (crude) LaTeX output as well;
  3001 
  3002 * browser info session directories are now self-contained (may be put
  3003 on WWW server seperately); improved graphs of nested sessions; removed
  3004 graph for 'all sessions';
  3005 
  3006 * several improvements in isabelle style files; \isabellestyle{it}
  3007 produces fake math mode output; \isamarkupheader is now \section by
  3008 default; see lib/texinputs/isabelle.sty etc.;
  3009 
  3010 
  3011 *** Isar ***
  3012 
  3013 * Isar/Pure: local results and corresponding term bindings are now
  3014 subject to Hindley-Milner polymorphism (similar to ML); this
  3015 accommodates incremental type-inference very nicely;
  3016 
  3017 * Isar/Pure: new derived language element 'obtain' supports
  3018 generalized existence reasoning;
  3019 
  3020 * Isar/Pure: new calculational elements 'moreover' and 'ultimately'
  3021 support accumulation of results, without applying any rules yet;
  3022 useful to collect intermediate results without explicit name
  3023 references, and for use with transitivity rules with more than 2
  3024 premises;
  3025 
  3026 * Isar/Pure: scalable support for case-analysis type proofs: new
  3027 'case' language element refers to local contexts symbolically, as
  3028 produced by certain proof methods; internally, case names are attached
  3029 to theorems as "tags";
  3030 
  3031 * Isar/Pure: theory command 'hide' removes declarations from
  3032 class/type/const name spaces;
  3033 
  3034 * Isar/Pure: theory command 'defs' supports option "(overloaded)" to
  3035 indicate potential overloading;
  3036 
  3037 * Isar/Pure: changed syntax of local blocks from {{ }} to { };
  3038 
  3039 * Isar/Pure: syntax of sorts made 'inner', i.e. have to write
  3040 "{a,b,c}" instead of {a,b,c};
  3041 
  3042 * Isar/Pure now provides its own version of intro/elim/dest
  3043 attributes; useful for building new logics, but beware of confusion
  3044 with the version in Provers/classical;
  3045 
  3046 * Isar/Pure: the local context of (non-atomic) goals is provided via
  3047 case name 'antecedent';
  3048 
  3049 * Isar/Pure: removed obsolete 'transfer' attribute (transfer of thms
  3050 to the current context is now done automatically);
  3051 
  3052 * Isar/Pure: theory command 'method_setup' provides a simple interface
  3053 for definining proof methods in ML;
  3054 
  3055 * Isar/Provers: intro/elim/dest attributes changed; renamed
  3056 intro/intro!/intro!! flags to intro!/intro/intro? (INCOMPATIBILITY, in
  3057 most cases, one should have to change intro!! to intro? only);
  3058 replaced "delrule" by "rule del";
  3059 
  3060 * Isar/Provers: new 'hypsubst' method, plain 'subst' method and
  3061 'symmetric' attribute (the latter supercedes [RS sym]);
  3062 
  3063 * Isar/Provers: splitter support (via 'split' attribute and 'simp'
  3064 method modifier); 'simp' method: 'only:' modifier removes loopers as
  3065 well (including splits);
  3066 
  3067 * Isar/Provers: Simplifier and Classical methods now support all kind
  3068 of modifiers used in the past, including 'cong', 'iff', etc.
  3069 
  3070 * Isar/Provers: added 'fastsimp' and 'clarsimp' methods (combination
  3071 of Simplifier and Classical reasoner);
  3072 
  3073 * Isar/HOL: new proof method 'cases' and improved version of 'induct'
  3074 now support named cases; major packages (inductive, datatype, primrec,
  3075 recdef) support case names and properly name parameters;
  3076 
  3077 * Isar/HOL: new transitivity rules for substitution in inequalities --
  3078 monotonicity conditions are extracted to be proven at end of
  3079 calculations;
  3080 
  3081 * Isar/HOL: removed 'case_split' thm binding, should use 'cases' proof
  3082 method anyway;
  3083 
  3084 * Isar/HOL: removed old expand_if = split_if; theorems if_splits =
  3085 split_if split_if_asm; datatype package provides theorems foo.splits =
  3086 foo.split foo.split_asm for each datatype;
  3087 
  3088 * Isar/HOL: tuned inductive package, rename "intrs" to "intros"
  3089 (potential INCOMPATIBILITY), emulation of mk_cases feature for proof
  3090 scripts: new 'inductive_cases' command and 'ind_cases' method; (Note:
  3091 use "(cases (simplified))" method in proper proof texts);
  3092 
  3093 * Isar/HOL: added global 'arith_split' attribute for 'arith' method;
  3094 
  3095 * Isar: names of theorems etc. may be natural numbers as well;
  3096 
  3097 * Isar: 'pr' command: optional arguments for goals_limit and
  3098 ProofContext.prems_limit; no longer prints theory contexts, but only
  3099 proof states;
  3100 
  3101 * Isar: diagnostic commands 'pr', 'thm', 'prop', 'term', 'typ' admit
  3102 additional print modes to be specified; e.g. "pr(latex)" will print
  3103 proof state according to the Isabelle LaTeX style;
  3104 
  3105 * Isar: improved support for emulating tactic scripts, including proof
  3106 methods 'rule_tac' etc., 'cut_tac', 'thin_tac', 'subgoal_tac',
  3107 'rename_tac', 'rotate_tac', 'tactic', and 'case_tac' / 'induct_tac'
  3108 (for HOL datatypes);
  3109 
  3110 * Isar: simplified (more robust) goal selection of proof methods: 1st
  3111 goal, all goals, or explicit goal specifier (tactic emulation); thus
  3112 'proof method scripts' have to be in depth-first order;
  3113 
  3114 * Isar: tuned 'let' syntax: replaced 'as' keyword by 'and';
  3115 
  3116 * Isar: removed 'help' command, which hasn't been too helpful anyway;
  3117 should instead use individual commands for printing items
  3118 (print_commands, print_methods etc.);
  3119 
  3120 * Isar: added 'nothing' --- the empty list of theorems;
  3121 
  3122 
  3123 *** HOL ***
  3124 
  3125 * HOL/MicroJava: formalization of a fragment of Java, together with a
  3126 corresponding virtual machine and a specification of its bytecode
  3127 verifier and a lightweight bytecode verifier, including proofs of
  3128 type-safety; by Gerwin Klein, Tobias Nipkow, David von Oheimb, and
  3129 Cornelia Pusch (see also the homepage of project Bali at
  3130 http://isabelle.in.tum.de/Bali/);
  3131 
  3132 * HOL/Algebra: new theory of rings and univariate polynomials, by
  3133 Clemens Ballarin;
  3134 
  3135 * HOL/NumberTheory: fundamental Theorem of Arithmetic, Chinese
  3136 Remainder Theorem, Fermat/Euler Theorem, Wilson's Theorem, by Thomas M
  3137 Rasmussen;
  3138 
  3139 * HOL/Lattice: fundamental concepts of lattice theory and order
  3140 structures, including duals, properties of bounds versus algebraic
  3141 laws, lattice operations versus set-theoretic ones, the Knaster-Tarski
  3142 Theorem for complete lattices etc.; may also serve as a demonstration
  3143 for abstract algebraic reasoning using axiomatic type classes, and
  3144 mathematics-style proof in Isabelle/Isar; by Markus Wenzel;
  3145 
  3146 * HOL/Prolog: a (bare-bones) implementation of Lambda-Prolog, by David
  3147 von Oheimb;
  3148 
  3149 * HOL/IMPP: extension of IMP with local variables and mutually
  3150 recursive procedures, by David von Oheimb;
  3151 
  3152 * HOL/Lambda: converted into new-style theory and document;
  3153 
  3154 * HOL/ex/Multiquote: example of multiple nested quotations and
  3155 anti-quotations -- basically a generalized version of de-Bruijn
  3156 representation; very useful in avoiding lifting of operations;
  3157 
  3158 * HOL/record: added general record equality rule to simpset; fixed
  3159 select-update simplification procedure to handle extended records as
  3160 well; admit "r" as field name;
  3161 
  3162 * HOL: 0 is now overloaded over the new sort "zero", allowing its use with
  3163 other numeric types and also as the identity of groups, rings, etc.;
  3164 
  3165 * HOL: new axclass plus_ac0 for addition with the AC-laws and 0 as identity.
  3166 Types nat and int belong to this axclass;
  3167 
  3168 * HOL: greatly improved simplification involving numerals of type nat, int, real:
  3169    (i + #8 + j) = Suc k simplifies to  #7 + (i + j) = k
  3170    i*j + k + j*#3*i     simplifies to  #4*(i*j) + k
  3171   two terms #m*u and #n*u are replaced by #(m+n)*u
  3172     (where #m, #n and u can implicitly be 1; this is simproc combine_numerals)
  3173   and the term/formula #m*u+x ~~ #n*u+y simplifies simplifies to #(m-n)+x ~~ y
  3174     or x ~~ #(n-m)+y, where ~~ is one of = < <= or - (simproc cancel_numerals);
  3175 
  3176 * HOL: meson_tac is available (previously in ex/meson.ML); it is a
  3177 powerful prover for predicate logic but knows nothing of clasets; see
  3178 ex/mesontest.ML and ex/mesontest2.ML for example applications;
  3179 
  3180 * HOL: new version of "case_tac" subsumes both boolean case split and
  3181 "exhaust_tac" on datatypes; INCOMPATIBILITY: exhaust_tac no longer
  3182 exists, may define val exhaust_tac = case_tac for ad-hoc portability;
  3183 
  3184 * HOL: simplification no longer dives into case-expressions: only the
  3185 selector expression is simplified, but not the remaining arms; to
  3186 enable full simplification of case-expressions for datatype t, you may
  3187 remove t.weak_case_cong from the simpset, either globally (Delcongs
  3188 [thm"t.weak_case_cong"];) or locally (delcongs [...]).
  3189 
  3190 * HOL/recdef: the recursion equations generated by 'recdef' for
  3191 function 'f' are now called f.simps instead of f.rules; if all
  3192 termination conditions are proved automatically, these simplification
  3193 rules are added to the simpset, as in primrec; rules may be named
  3194 individually as well, resulting in a separate list of theorems for
  3195 each equation;
  3196 
  3197 * HOL/While is a new theory that provides a while-combinator. It
  3198 permits the definition of tail-recursive functions without the
  3199 provision of a termination measure. The latter is necessary once the
  3200 invariant proof rule for while is applied.
  3201 
  3202 * HOL: new (overloaded) notation for the set of elements below/above
  3203 some element: {..u}, {..u(}, {l..}, {)l..}. See theory SetInterval.
  3204 
  3205 * HOL: theorems impI, allI, ballI bound as "strip";
  3206 
  3207 * HOL: new tactic induct_thm_tac: thm -> string -> int -> tactic
  3208 induct_tac th "x1 ... xn" expects th to have a conclusion of the form
  3209 P v1 ... vn and abbreviates res_inst_tac [("v1","x1"),...,("vn","xn")] th;
  3210 
  3211 * HOL/Real: "rabs" replaced by overloaded "abs" function;
  3212 
  3213 * HOL: theory Sexp now in HOL/Induct examples (it used to be part of
  3214 main HOL, but was unused);
  3215 
  3216 * HOL: fewer consts declared as global (e.g. have to refer to
  3217 "Lfp.lfp" instead of "lfp" internally; affects ML packages only);
  3218 
  3219 * HOL: tuned AST representation of nested pairs, avoiding bogus output
  3220 in case of overlap with user translations (e.g. judgements over
  3221 tuples); (note that the underlying logical represenation is still
  3222 bogus);
  3223 
  3224 
  3225 *** ZF ***
  3226 
  3227 * ZF: simplification automatically cancels common terms in arithmetic
  3228 expressions over nat and int;
  3229 
  3230 * ZF: new treatment of nat to minimize type-checking: all operators
  3231 coerce their operands to a natural number using the function natify,
  3232 making the algebraic laws unconditional;
  3233 
  3234 * ZF: as above, for int: operators coerce their operands to an integer
  3235 using the function intify;
  3236 
  3237 * ZF: the integer library now contains many of the usual laws for the
  3238 orderings, including $<=, and monotonicity laws for $+ and $*;
  3239 
  3240 * ZF: new example ZF/ex/NatSum to demonstrate integer arithmetic
  3241 simplification;
  3242 
  3243 * FOL and ZF: AddIffs now available, giving theorems of the form P<->Q
  3244 to the simplifier and classical reasoner simultaneously;
  3245 
  3246 
  3247 *** General ***
  3248 
  3249 * Provers: blast_tac now handles actual object-logic rules as
  3250 assumptions; note that auto_tac uses blast_tac internally as well;
  3251 
  3252 * Provers: new functions rulify/rulify_no_asm: thm -> thm for turning
  3253 outer -->/All/Ball into ==>/!!; qed_spec_mp now uses rulify_no_asm;
  3254 
  3255 * Provers: delrules now handles destruct rules as well (no longer need
  3256 explicit make_elim);
  3257 
  3258 * Provers: Blast_tac now warns of and ignores "weak elimination rules" e.g.
  3259   [| inj ?f;          ?f ?x = ?f ?y; ?x = ?y ==> ?W |] ==> ?W
  3260 use instead the strong form,
  3261   [| inj ?f; ~ ?W ==> ?f ?x = ?f ?y; ?x = ?y ==> ?W |] ==> ?W
  3262 in HOL, FOL and ZF the function cla_make_elim will create such rules
  3263 from destruct-rules;
  3264 
  3265 * Provers: Simplifier.easy_setup provides a fast path to basic
  3266 Simplifier setup for new object-logics;
  3267 
  3268 * Pure: AST translation rules no longer require constant head on LHS;
  3269 
  3270 * Pure: improved name spaces: ambiguous output is qualified; support
  3271 for hiding of names;
  3272 
  3273 * system: smart setup of canonical ML_HOME, ISABELLE_INTERFACE, and
  3274 XSYMBOL_HOME; no longer need to do manual configuration in most
  3275 situations;
  3276 
  3277 * system: compression of ML heaps images may now be controlled via -c
  3278 option of isabelle and isatool usedir (currently only observed by
  3279 Poly/ML);
  3280 
  3281 * system: isatool installfonts may handle X-Symbol fonts as well (very
  3282 useful for remote X11);
  3283 
  3284 * system: provide TAGS file for Isabelle sources;
  3285 
  3286 * ML: infix 'OF' is a version of 'MRS' with more appropriate argument
  3287 order;
  3288 
  3289 * ML: renamed flags Syntax.trace_norm_ast to Syntax.trace_ast; global
  3290 timing flag supersedes proof_timing and Toplevel.trace;
  3291 
  3292 * ML: new combinators |>> and |>>> for incremental transformations
  3293 with secondary results (e.g. certain theory extensions):
  3294 
  3295 * ML: PureThy.add_defs gets additional argument to indicate potential
  3296 overloading (usually false);
  3297 
  3298 * ML: PureThy.add_thms/add_axioms/add_defs now return theorems as
  3299 results;
  3300 
  3301 
  3302 
  3303 New in Isabelle99 (October 1999)
  3304 --------------------------------
  3305 
  3306 *** Overview of INCOMPATIBILITIES (see below for more details) ***
  3307 
  3308 * HOL: The THEN and ELSE parts of conditional expressions (if P then x else y)
  3309 are no longer simplified.  (This allows the simplifier to unfold recursive
  3310 functional programs.)  To restore the old behaviour, declare
  3311 
  3312     Delcongs [if_weak_cong];
  3313 
  3314 * HOL: Removed the obsolete syntax "Compl A"; use -A for set
  3315 complement;
  3316 
  3317 * HOL: the predicate "inj" is now defined by translation to "inj_on";
  3318 
  3319 * HOL/datatype: mutual_induct_tac no longer exists --
  3320   use induct_tac "x_1 ... x_n" instead of mutual_induct_tac ["x_1", ..., "x_n"]
  3321 
  3322 * HOL/typedef: fixed type inference for representing set; type
  3323 arguments now have to occur explicitly on the rhs as type constraints;
  3324 
  3325 * ZF: The con_defs part of an inductive definition may no longer refer
  3326 to constants declared in the same theory;
  3327 
  3328 * HOL, ZF: the function mk_cases, generated by the inductive
  3329 definition package, has lost an argument.  To simplify its result, it
  3330 uses the default simpset instead of a supplied list of theorems.
  3331 
  3332 * HOL/List: the constructors of type list are now Nil and Cons;
  3333 
  3334 * Simplifier: the type of the infix ML functions
  3335         setSSolver addSSolver setSolver addSolver
  3336 is now  simpset * solver -> simpset  where `solver' is a new abstract type
  3337 for packaging solvers. A solver is created via
  3338         mk_solver: string -> (thm list -> int -> tactic) -> solver
  3339 where the string argument is only a comment.
  3340 
  3341 
  3342 *** Proof tools ***
  3343 
  3344 * Provers/Arith/fast_lin_arith.ML contains a functor for creating a
  3345 decision procedure for linear arithmetic. Currently it is used for
  3346 types `nat', `int', and `real' in HOL (see below); it can, should and
  3347 will be instantiated for other types and logics as well.
  3348 
  3349 * The simplifier now accepts rewrite rules with flexible heads, eg
  3350      hom ?f ==> ?f(?x+?y) = ?f ?x + ?f ?y
  3351   They are applied like any rule with a non-pattern lhs, i.e. by first-order
  3352   matching.
  3353 
  3354 
  3355 *** General ***
  3356 
  3357 * New Isabelle/Isar subsystem provides an alternative to traditional
  3358 tactical theorem proving; together with the ProofGeneral/isar user
  3359 interface it offers an interactive environment for developing human
  3360 readable proof documents (Isar == Intelligible semi-automated
  3361 reasoning); for further information see isatool doc isar-ref,
  3362 src/HOL/Isar_examples and http://isabelle.in.tum.de/Isar/
  3363 
  3364 * improved and simplified presentation of theories: better HTML markup
  3365 (including colors), graph views in several sizes; isatool usedir now
  3366 provides a proper interface for user theories (via -P option); actual
  3367 document preparation based on (PDF)LaTeX is available as well (for
  3368 new-style theories only); see isatool doc system for more information;
  3369 
  3370 * native support for Proof General, both for classic Isabelle and
  3371 Isabelle/Isar;
  3372 
  3373 * ML function thm_deps visualizes dependencies of theorems and lemmas,
  3374 using the graph browser tool;
  3375 
  3376 * Isabelle manuals now also available as PDF;
  3377 
  3378 * theory loader rewritten from scratch (may not be fully
  3379 bug-compatible); old loadpath variable has been replaced by show_path,
  3380 add_path, del_path, reset_path functions; new operations such as
  3381 update_thy, touch_thy, remove_thy, use/update_thy_only (see also
  3382 isatool doc ref);
  3383 
  3384 * improved isatool install: option -k creates KDE application icon,
  3385 option -p DIR installs standalone binaries;
  3386 
  3387 * added ML_PLATFORM setting (useful for cross-platform installations);
  3388 more robust handling of platform specific ML images for SML/NJ;
  3389 
  3390 * the settings environment is now statically scoped, i.e. it is never
  3391 created again in sub-processes invoked from isabelle, isatool, or
  3392 Isabelle;
  3393 
  3394 * path element specification '~~' refers to '$ISABELLE_HOME';
  3395 
  3396 * in locales, the "assumes" and "defines" parts may be omitted if
  3397 empty;
  3398 
  3399 * new print_mode "xsymbols" for extended symbol support (e.g. genuine
  3400 long arrows);
  3401 
  3402 * new print_mode "HTML";
  3403 
  3404 * new flag show_tags controls display of tags of theorems (which are
  3405 basically just comments that may be attached by some tools);
  3406 
  3407 * Isamode 2.6 requires patch to accomodate change of Isabelle font
  3408 mode and goal output format:
  3409 
  3410 diff -r Isamode-2.6/elisp/isa-load.el Isamode/elisp/isa-load.el
  3411 244c244
  3412 <       (list (isa-getenv "ISABELLE") "-msymbols" logic-name)
  3413 ---
  3414 >       (list (isa-getenv "ISABELLE") "-misabelle_font" "-msymbols" logic-name)
  3415 diff -r Isabelle-2.6/elisp/isa-proofstate.el Isamode/elisp/isa-proofstate.el
  3416 181c181
  3417 < (defconst proofstate-proofstart-regexp "^Level [0-9]+$"
  3418 ---
  3419 > (defconst proofstate-proofstart-regexp "^Level [0-9]+"
  3420 
  3421 * function bind_thms stores lists of theorems (cf. bind_thm);
  3422 
  3423 * new shorthand tactics ftac, eatac, datac, fatac;
  3424 
  3425 * qed (and friends) now accept "" as result name; in that case the
  3426 theorem is not stored, but proper checks and presentation of the
  3427 result still apply;
  3428 
  3429 * theorem database now also indexes constants "Trueprop", "all",
  3430 "==>", "=="; thus thms_containing, findI etc. may retrieve more rules;
  3431 
  3432 
  3433 *** HOL ***
  3434 
  3435 ** HOL arithmetic **
  3436 
  3437 * There are now decision procedures for linear arithmetic over nat and
  3438 int:
  3439 
  3440 1. arith_tac copes with arbitrary formulae involving `=', `<', `<=',
  3441 `+', `-', `Suc', `min', `max' and numerical constants; other subterms
  3442 are treated as atomic; subformulae not involving type `nat' or `int'
  3443 are ignored; quantified subformulae are ignored unless they are
  3444 positive universal or negative existential. The tactic has to be
  3445 invoked by hand and can be a little bit slow. In particular, the
  3446 running time is exponential in the number of occurrences of `min' and
  3447 `max', and `-' on `nat'.
  3448 
  3449 2. fast_arith_tac is a cut-down version of arith_tac: it only takes
  3450 (negated) (in)equalities among the premises and the conclusion into
  3451 account (i.e. no compound formulae) and does not know about `min' and
  3452 `max', and `-' on `nat'. It is fast and is used automatically by the
  3453 simplifier.
  3454 
  3455 NB: At the moment, these decision procedures do not cope with mixed
  3456 nat/int formulae where the two parts interact, such as `m < n ==>
  3457 int(m) < int(n)'.
  3458 
  3459 * HOL/Numeral provides a generic theory of numerals (encoded
  3460 efficiently as bit strings); setup for types nat/int/real is in place;
  3461 INCOMPATIBILITY: since numeral syntax is now polymorphic, rather than
  3462 int, existing theories and proof scripts may require a few additional
  3463 type constraints;
  3464 
  3465 * integer division and remainder can now be performed on constant
  3466 arguments;
  3467 
  3468 * many properties of integer multiplication, division and remainder
  3469 are now available;
  3470 
  3471 * An interface to the Stanford Validity Checker (SVC) is available through the
  3472 tactic svc_tac.  Propositional tautologies and theorems of linear arithmetic
  3473 are proved automatically.  SVC must be installed separately, and its results
  3474 must be TAKEN ON TRUST (Isabelle does not check the proofs, but tags any
  3475 invocation of the underlying oracle).  For SVC see
  3476   http://verify.stanford.edu/SVC
  3477 
  3478 * IsaMakefile: the HOL-Real target now builds an actual image;
  3479 
  3480 
  3481 ** HOL misc **
  3482 
  3483 * HOL/Real/HahnBanach: the Hahn-Banach theorem for real vector spaces
  3484 (in Isabelle/Isar) -- by Gertrud Bauer;
  3485 
  3486 * HOL/BCV: generic model of bytecode verification, i.e. data-flow
  3487 analysis for assembly languages with subtypes;
  3488 
  3489 * HOL/TLA (Lamport's Temporal Logic of Actions): major reorganization
  3490 -- avoids syntactic ambiguities and treats state, transition, and
  3491 temporal levels more uniformly; introduces INCOMPATIBILITIES due to
  3492 changed syntax and (many) tactics;
  3493 
  3494 * HOL/inductive: Now also handles more general introduction rules such
  3495   as "ALL y. (y, x) : r --> y : acc r ==> x : acc r"; monotonicity
  3496   theorems are now maintained within the theory (maintained via the
  3497   "mono" attribute);
  3498 
  3499 * HOL/datatype: Now also handles arbitrarily branching datatypes
  3500   (using function types) such as
  3501 
  3502   datatype 'a tree = Atom 'a | Branch "nat => 'a tree"
  3503 
  3504 * HOL/record: record_simproc (part of the default simpset) takes care
  3505 of selectors applied to updated records; record_split_tac is no longer
  3506 part of the default claset; update_defs may now be removed from the
  3507 simpset in many cases; COMPATIBILITY: old behavior achieved by
  3508 
  3509   claset_ref () := claset() addSWrapper record_split_wrapper;
  3510   Delsimprocs [record_simproc]
  3511 
  3512 * HOL/typedef: fixed type inference for representing set; type
  3513 arguments now have to occur explicitly on the rhs as type constraints;
  3514 
  3515 * HOL/recdef (TFL): 'congs' syntax now expects comma separated list of theorem
  3516 names rather than an ML expression;
  3517 
  3518 * HOL/defer_recdef (TFL): like recdef but the well-founded relation can be
  3519 supplied later.  Program schemes can be defined, such as
  3520     "While B C s = (if B s then While B C (C s) else s)"
  3521 where the well-founded relation can be chosen after B and C have been given.
  3522 
  3523 * HOL/List: the constructors of type list are now Nil and Cons;
  3524 INCOMPATIBILITY: while [] and infix # syntax is still there, of
  3525 course, ML tools referring to List.list.op # etc. have to be adapted;
  3526 
  3527 * HOL_quantifiers flag superseded by "HOL" print mode, which is
  3528 disabled by default; run isabelle with option -m HOL to get back to
  3529 the original Gordon/HOL-style output;
  3530 
  3531 * HOL/Ord.thy: new bounded quantifier syntax (input only): ALL x<y. P,
  3532 ALL x<=y. P, EX x<y. P, EX x<=y. P;
  3533 
  3534 * HOL basic syntax simplified (more orthogonal): all variants of
  3535 All/Ex now support plain / symbolic / HOL notation; plain syntax for
  3536 Eps operator is provided as well: "SOME x. P[x]";
  3537 
  3538 * HOL/Sum.thy: sum_case has been moved to HOL/Datatype;
  3539 
  3540 * HOL/Univ.thy: infix syntax <*>, <+>, <**>, <+> eliminated and made
  3541 thus available for user theories;
  3542 
  3543 * HOLCF/IOA/Sequents: renamed 'Cons' to 'Consq' to avoid clash with
  3544 HOL/List; hardly an INCOMPATIBILITY since '>>' syntax is used all the
  3545 time;
  3546 
  3547 * HOL: new tactic smp_tac: int -> int -> tactic, which applies spec
  3548 several times and then mp;
  3549 
  3550 
  3551 *** LK ***
  3552 
  3553 * the notation <<...>> is now available as a notation for sequences of
  3554 formulas;
  3555 
  3556 * the simplifier is now installed
  3557 
  3558 * the axiom system has been generalized (thanks to Soren Heilmann)
  3559 
  3560 * the classical reasoner now has a default rule database
  3561 
  3562 
  3563 *** ZF ***
  3564 
  3565 * new primrec section allows primitive recursive functions to be given
  3566 directly (as in HOL) over datatypes and the natural numbers;
  3567 
  3568 * new tactics induct_tac and exhaust_tac for induction (or case
  3569 analysis) over datatypes and the natural numbers;
  3570 
  3571 * the datatype declaration of type T now defines the recursor T_rec;
  3572 
  3573 * simplification automatically does freeness reasoning for datatype
  3574 constructors;
  3575 
  3576 * automatic type-inference, with AddTCs command to insert new
  3577 type-checking rules;
  3578 
  3579 * datatype introduction rules are now added as Safe Introduction rules
  3580 to the claset;
  3581 
  3582 * the syntax "if P then x else y" is now available in addition to
  3583 if(P,x,y);
  3584 
  3585 
  3586 *** Internal programming interfaces ***
  3587 
  3588 * tuned simplifier trace output; new flag debug_simp;
  3589 
  3590 * structures Vartab / Termtab (instances of TableFun) offer efficient
  3591 tables indexed by indexname_ord / term_ord (compatible with aconv);
  3592 
  3593 * AxClass.axclass_tac lost the theory argument;
  3594 
  3595 * tuned current_goals_markers semantics: begin / end goal avoids
  3596 printing empty lines;
  3597 
  3598 * removed prs and prs_fn hook, which was broken because it did not
  3599 include \n in its semantics, forcing writeln to add one
  3600 uncoditionally; replaced prs_fn by writeln_fn; consider std_output:
  3601 string -> unit if you really want to output text without newline;
  3602 
  3603 * Symbol.output subject to print mode; INCOMPATIBILITY: defaults to
  3604 plain output, interface builders may have to enable 'isabelle_font'
  3605 mode to get Isabelle font glyphs as before;
  3606 
  3607 * refined token_translation interface; INCOMPATIBILITY: output length
  3608 now of type real instead of int;
  3609 
  3610 * theory loader actions may be traced via new ThyInfo.add_hook
  3611 interface (see src/Pure/Thy/thy_info.ML); example application: keep
  3612 your own database of information attached to *whole* theories -- as
  3613 opposed to intra-theory data slots offered via TheoryDataFun;
  3614 
  3615 * proper handling of dangling sort hypotheses (at last!);
  3616 Thm.strip_shyps and Drule.strip_shyps_warning take care of removing
  3617 extra sort hypotheses that can be witnessed from the type signature;
  3618 the force_strip_shyps flag is gone, any remaining shyps are simply
  3619 left in the theorem (with a warning issued by strip_shyps_warning);
  3620 
  3621 
  3622 
  3623 New in Isabelle98-1 (October 1998)
  3624 ----------------------------------
  3625 
  3626 *** Overview of INCOMPATIBILITIES (see below for more details) ***
  3627 
  3628 * several changes of automated proof tools;
  3629 
  3630 * HOL: major changes to the inductive and datatype packages, including
  3631 some minor incompatibilities of theory syntax;
  3632 
  3633 * HOL: renamed r^-1 to 'converse' from 'inverse'; 'inj_onto' is now
  3634 called `inj_on';
  3635 
  3636 * HOL: removed duplicate thms in Arith:
  3637   less_imp_add_less  should be replaced by  trans_less_add1
  3638   le_imp_add_le      should be replaced by  trans_le_add1
  3639 
  3640 * HOL: unary minus is now overloaded (new type constraints may be
  3641 required);
  3642 
  3643 * HOL and ZF: unary minus for integers is now #- instead of #~.  In
  3644 ZF, expressions such as n#-1 must be changed to n#- 1, since #-1 is
  3645 now taken as an integer constant.
  3646 
  3647 * Pure: ML function 'theory_of' renamed to 'theory';
  3648 
  3649 
  3650 *** Proof tools ***
  3651 
  3652 * Simplifier:
  3653   1. Asm_full_simp_tac is now more aggressive.
  3654      1. It will sometimes reorient premises if that increases their power to
  3655         simplify.
  3656      2. It does no longer proceed strictly from left to right but may also
  3657         rotate premises to achieve further simplification.
  3658      For compatibility reasons there is now Asm_lr_simp_tac which is like the
  3659      old Asm_full_simp_tac in that it does not rotate premises.
  3660   2. The simplifier now knows a little bit about nat-arithmetic.
  3661 
  3662 * Classical reasoner: wrapper mechanism for the classical reasoner now
  3663 allows for selected deletion of wrappers, by introduction of names for
  3664 wrapper functionals.  This implies that addbefore, addSbefore,
  3665 addaltern, and addSaltern now take a pair (name, tactic) as argument,
  3666 and that adding two tactics with the same name overwrites the first
  3667 one (emitting a warning).
  3668   type wrapper = (int -> tactic) -> (int -> tactic)
  3669   setWrapper, setSWrapper, compWrapper and compSWrapper are replaced by
  3670   addWrapper, addSWrapper: claset * (string * wrapper) -> claset
  3671   delWrapper, delSWrapper: claset *  string            -> claset
  3672   getWrapper is renamed to appWrappers, getSWrapper to appSWrappers;
  3673 
  3674 * Classical reasoner: addbefore/addSbefore now have APPEND/ORELSE
  3675 semantics; addbefore now affects only the unsafe part of step_tac
  3676 etc.; this affects addss/auto_tac/force_tac, so EXISTING PROOFS MAY
  3677 FAIL, but proofs should be fixable easily, e.g. by replacing Auto_tac
  3678 by Force_tac;
  3679 
  3680 * Classical reasoner: setwrapper to setWrapper and compwrapper to
  3681 compWrapper; added safe wrapper (and access functions for it);
  3682 
  3683 * HOL/split_all_tac is now much faster and fails if there is nothing
  3684 to split.  Some EXISTING PROOFS MAY REQUIRE ADAPTION because the order
  3685 and the names of the automatically generated variables have changed.
  3686 split_all_tac has moved within claset() from unsafe wrappers to safe
  3687 wrappers, which means that !!-bound variables are split much more
  3688 aggressively, and safe_tac and clarify_tac now split such variables.
  3689 If this splitting is not appropriate, use delSWrapper "split_all_tac".
  3690 Note: the same holds for record_split_tac, which does the job of
  3691 split_all_tac for record fields.
  3692 
  3693 * HOL/Simplifier: Rewrite rules for case distinctions can now be added
  3694 permanently to the default simpset using Addsplits just like
  3695 Addsimps. They can be removed via Delsplits just like
  3696 Delsimps. Lower-case versions are also available.
  3697 
  3698 * HOL/Simplifier: The rule split_if is now part of the default
  3699 simpset. This means that the simplifier will eliminate all occurrences
  3700 of if-then-else in the conclusion of a goal. To prevent this, you can
  3701 either remove split_if completely from the default simpset by
  3702 `Delsplits [split_if]' or remove it in a specific call of the
  3703 simplifier using `... delsplits [split_if]'.  You can also add/delete
  3704 other case splitting rules to/from the default simpset: every datatype
  3705 generates suitable rules `split_t_case' and `split_t_case_asm' (where
  3706 t is the name of the datatype).
  3707 
  3708 * Classical reasoner / Simplifier combination: new force_tac (and
  3709 derivatives Force_tac, force) combines rewriting and classical
  3710 reasoning (and whatever other tools) similarly to auto_tac, but is
  3711 aimed to solve the given subgoal completely.
  3712 
  3713 
  3714 *** General ***
  3715 
  3716 * new top-level commands `Goal' and `Goalw' that improve upon `goal'
  3717 and `goalw': the theory is no longer needed as an explicit argument -
  3718 the current theory context is used; assumptions are no longer returned
  3719 at the ML-level unless one of them starts with ==> or !!; it is
  3720 recommended to convert to these new commands using isatool fixgoal
  3721 (backup your sources first!);
  3722 
  3723 * new top-level commands 'thm' and 'thms' for retrieving theorems from
  3724 the current theory context, and 'theory' to lookup stored theories;
  3725 
  3726 * new theory section 'locale' for declaring constants, assumptions and
  3727 definitions that have local scope;
  3728 
  3729 * new theory section 'nonterminals' for purely syntactic types;
  3730 
  3731 * new theory section 'setup' for generic ML setup functions
  3732 (e.g. package initialization);
  3733 
  3734 * the distribution now includes Isabelle icons: see
  3735 lib/logo/isabelle-{small,tiny}.xpm;
  3736 
  3737 * isatool install - install binaries with absolute references to
  3738 ISABELLE_HOME/bin;
  3739 
  3740 * isatool logo -- create instances of the Isabelle logo (as EPS);
  3741 
  3742 * print mode 'emacs' reserved for Isamode;
  3743 
  3744 * support multiple print (ast) translations per constant name;
  3745 
  3746 * theorems involving oracles are now printed with a suffixed [!];
  3747 
  3748 
  3749 *** HOL ***
  3750 
  3751 * there is now a tutorial on Isabelle/HOL (do 'isatool doc tutorial');
  3752 
  3753 * HOL/inductive package reorganized and improved: now supports mutual
  3754 definitions such as
  3755 
  3756   inductive EVEN ODD
  3757     intrs
  3758       null "0 : EVEN"
  3759       oddI "n : EVEN ==> Suc n : ODD"
  3760       evenI "n : ODD ==> Suc n : EVEN"
  3761 
  3762 new theorem list "elims" contains an elimination rule for each of the
  3763 recursive sets; inductive definitions now handle disjunctive premises
  3764 correctly (also ZF);
  3765 
  3766 INCOMPATIBILITIES: requires Inductive as an ancestor; component
  3767 "mutual_induct" no longer exists - the induction rule is always
  3768 contained in "induct";
  3769 
  3770 
  3771 * HOL/datatype package re-implemented and greatly improved: now
  3772 supports mutually recursive datatypes such as
  3773 
  3774   datatype
  3775     'a aexp = IF_THEN_ELSE ('a bexp) ('a aexp) ('a aexp)
  3776             | SUM ('a aexp) ('a aexp)
  3777             | DIFF ('a aexp) ('a aexp)
  3778             | NUM 'a
  3779   and
  3780     'a bexp = LESS ('a aexp) ('a aexp)
  3781             | AND ('a bexp) ('a bexp)
  3782             | OR ('a bexp) ('a bexp)
  3783 
  3784 as well as indirectly recursive datatypes such as
  3785 
  3786   datatype
  3787     ('a, 'b) term = Var 'a
  3788                   | App 'b ((('a, 'b) term) list)
  3789 
  3790 The new tactic  mutual_induct_tac [<var_1>, ..., <var_n>] i  performs
  3791 induction on mutually / indirectly recursive datatypes.
  3792 
  3793 Primrec equations are now stored in theory and can be accessed via
  3794 <function_name>.simps.
  3795 
  3796 INCOMPATIBILITIES:
  3797 
  3798   - Theories using datatypes must now have theory Datatype as an
  3799     ancestor.
  3800   - The specific <typename>.induct_tac no longer exists - use the
  3801     generic induct_tac instead.
  3802   - natE has been renamed to nat.exhaust - use exhaust_tac
  3803     instead of res_inst_tac ... natE. Note that the variable
  3804     names in nat.exhaust differ from the names in natE, this
  3805     may cause some "fragile" proofs to fail.
  3806   - The theorems split_<typename>_case and split_<typename>_case_asm
  3807     have been renamed to <typename>.split and <typename>.split_asm.
  3808   - Since default sorts of type variables are now handled correctly,
  3809     some datatype definitions may have to be annotated with explicit
  3810     sort constraints.
  3811   - Primrec definitions no longer require function name and type
  3812     of recursive argument.
  3813 
  3814 Consider using isatool fixdatatype to adapt your theories and proof
  3815 scripts to the new package (backup your sources first!).
  3816 
  3817 
  3818 * HOL/record package: considerably improved implementation; now
  3819 includes concrete syntax for record types, terms, updates; theorems
  3820 for surjective pairing and splitting !!-bound record variables; proof
  3821 support is as follows:
  3822 
  3823   1) standard conversions (selectors or updates applied to record
  3824 constructor terms) are part of the standard simpset;
  3825 
  3826   2) inject equations of the form ((x, y) = (x', y')) == x=x' & y=y' are
  3827 made part of standard simpset and claset via addIffs;
  3828 
  3829   3) a tactic for record field splitting (record_split_tac) is part of
  3830 the standard claset (addSWrapper);
  3831 
  3832 To get a better idea about these rules you may retrieve them via
  3833 something like 'thms "foo.simps"' or 'thms "foo.iffs"', where "foo" is
  3834 the name of your record type.
  3835 
  3836 The split tactic 3) conceptually simplifies by the following rule:
  3837 
  3838   "(!!x. PROP ?P x) == (!!a b. PROP ?P (a, b))"
  3839 
  3840 Thus any record variable that is bound by meta-all will automatically
  3841 blow up into some record constructor term, consequently the
  3842 simplifications of 1), 2) apply.  Thus force_tac, auto_tac etc. shall
  3843 solve record problems automatically.
  3844 
  3845 
  3846 * reorganized the main HOL image: HOL/Integ and String loaded by
  3847 default; theory Main includes everything;
  3848 
  3849 * automatic simplification of integer sums and comparisons, using cancellation;
  3850 
  3851 * added option_map_eq_Some and not_Some_eq to the default simpset and claset;
  3852 
  3853 * added disj_not1 = "(~P | Q) = (P --> Q)" to the default simpset;
  3854 
  3855 * many new identities for unions, intersections, set difference, etc.;
  3856 
  3857 * expand_if, expand_split, expand_sum_case and expand_nat_case are now
  3858 called split_if, split_split, split_sum_case and split_nat_case (to go
  3859 with add/delsplits);
  3860 
  3861 * HOL/Prod introduces simplification procedure unit_eq_proc rewriting
  3862 (?x::unit) = (); this is made part of the default simpset, which COULD
  3863 MAKE EXISTING PROOFS FAIL under rare circumstances (consider
  3864 'Delsimprocs [unit_eq_proc];' as last resort); also note that
  3865 unit_abs_eta_conv is added in order to counter the effect of
  3866 unit_eq_proc on (%u::unit. f u), replacing it by f rather than by
  3867 %u.f();
  3868 
  3869 * HOL/Fun INCOMPATIBILITY: `inj_onto' is now called `inj_on' (which
  3870 makes more sense);
  3871 
  3872 * HOL/Set INCOMPATIBILITY: rule `equals0D' is now a well-formed destruct rule;
  3873   It and 'sym RS equals0D' are now in the default  claset, giving automatic
  3874   disjointness reasoning but breaking a few old proofs.
  3875 
  3876 * HOL/Relation INCOMPATIBILITY: renamed the relational operator r^-1
  3877 to 'converse' from 'inverse' (for compatibility with ZF and some
  3878 literature);
  3879 
  3880 * HOL/recdef can now declare non-recursive functions, with {} supplied as
  3881 the well-founded relation;
  3882 
  3883 * HOL/Set INCOMPATIBILITY: the complement of set A is now written -A instead of
  3884     Compl A.  The "Compl" syntax remains available as input syntax for this
  3885     release ONLY.
  3886 
  3887 * HOL/Update: new theory of function updates:
  3888     f(a:=b) == %x. if x=a then b else f x
  3889 may also be iterated as in f(a:=b,c:=d,...);
  3890 
  3891 * HOL/Vimage: new theory for inverse image of a function, syntax f-``B;
  3892 
  3893 * HOL/List:
  3894   - new function list_update written xs[i:=v] that updates the i-th
  3895     list position. May also be iterated as in xs[i:=a,j:=b,...].
  3896   - new function `upt' written [i..j(] which generates the list
  3897     [i,i+1,...,j-1], i.e. the upper bound is excluded. To include the upper
  3898     bound write [i..j], which is a shorthand for [i..j+1(].
  3899   - new lexicographic orderings and corresponding wellfoundedness theorems.
  3900 
  3901 * HOL/Arith:
  3902   - removed 'pred' (predecessor) function;
  3903   - generalized some theorems about n-1;
  3904   - many new laws about "div" and "mod";
  3905   - new laws about greatest common divisors (see theory ex/Primes);
  3906 
  3907 * HOL/Relation: renamed the relational operator r^-1 "converse"
  3908 instead of "inverse";
  3909 
  3910 * HOL/Induct/Multiset: a theory of multisets, including the wellfoundedness
  3911   of the multiset ordering;
  3912 
  3913 * directory HOL/Real: a construction of the reals using Dedekind cuts
  3914   (not included by default);
  3915 
  3916 * directory HOL/UNITY: Chandy and Misra's UNITY formalism;
  3917 
  3918 * directory HOL/Hoare: a new version of Hoare logic which permits many-sorted
  3919   programs, i.e. different program variables may have different types.
  3920 
  3921 * calling (stac rew i) now fails if "rew" has no effect on the goal
  3922   [previously, this check worked only if the rewrite rule was unconditional]
  3923   Now rew can involve either definitions or equalities (either == or =).
  3924 
  3925 
  3926 *** ZF ***
  3927 
  3928 * theory Main includes everything; INCOMPATIBILITY: theory ZF.thy contains
  3929   only the theorems proved on ZF.ML;
  3930 
  3931 * ZF INCOMPATIBILITY: rule `equals0D' is now a well-formed destruct rule;
  3932   It and 'sym RS equals0D' are now in the default  claset, giving automatic
  3933   disjointness reasoning but breaking a few old proofs.
  3934 
  3935 * ZF/Update: new theory of function updates
  3936     with default rewrite rule  f(x:=y) ` z = if(z=x, y, f`z)
  3937   may also be iterated as in f(a:=b,c:=d,...);
  3938 
  3939 * in  let x=t in u(x), neither t nor u(x) has to be an FOL term.
  3940 
  3941 * calling (stac rew i) now fails if "rew" has no effect on the goal
  3942   [previously, this check worked only if the rewrite rule was unconditional]
  3943   Now rew can involve either definitions or equalities (either == or =).
  3944 
  3945 * case_tac provided for compatibility with HOL
  3946     (like the old excluded_middle_tac, but with subgoals swapped)
  3947 
  3948 
  3949 *** Internal programming interfaces ***
  3950 
  3951 * Pure: several new basic modules made available for general use, see
  3952 also src/Pure/README;
  3953 
  3954 * improved the theory data mechanism to support encapsulation (data
  3955 kind name replaced by private Object.kind, acting as authorization
  3956 key); new type-safe user interface via functor TheoryDataFun; generic
  3957 print_data function becomes basically useless;
  3958 
  3959 * removed global_names compatibility flag -- all theory declarations
  3960 are qualified by default;
  3961 
  3962 * module Pure/Syntax now offers quote / antiquote translation
  3963 functions (useful for Hoare logic etc. with implicit dependencies);
  3964 see HOL/ex/Antiquote for an example use;
  3965 
  3966 * Simplifier now offers conversions (asm_)(full_)rewrite: simpset ->
  3967 cterm -> thm;
  3968 
  3969 * new tactical CHANGED_GOAL for checking that a tactic modifies a
  3970 subgoal;
  3971 
  3972 * Display.print_goals function moved to Locale.print_goals;
  3973 
  3974 * standard print function for goals supports current_goals_markers
  3975 variable for marking begin of proof, end of proof, start of goal; the
  3976 default is ("", "", ""); setting current_goals_markers := ("<proof>",
  3977 "</proof>", "<goal>") causes SGML like tagged proof state printing,
  3978 for example;
  3979 
  3980 
  3981 
  3982 New in Isabelle98 (January 1998)
  3983 --------------------------------
  3984 
  3985 *** Overview of INCOMPATIBILITIES (see below for more details) ***
  3986 
  3987 * changed lexical syntax of terms / types: dots made part of long
  3988 identifiers, e.g. "%x.x" no longer possible, should be "%x. x";
  3989 
  3990 * simpset (and claset) reference variable replaced by functions
  3991 simpset / simpset_ref;
  3992 
  3993 * no longer supports theory aliases (via merge) and non-trivial
  3994 implicit merge of thms' signatures;
  3995 
  3996 * most internal names of constants changed due to qualified names;
  3997 
  3998 * changed Pure/Sequence interface (see Pure/seq.ML);
  3999 
  4000 
  4001 *** General Changes ***
  4002 
  4003 * hierachically structured name spaces (for consts, types, axms, thms
  4004 etc.); new lexical class 'longid' (e.g. Foo.bar.x) may render much of
  4005 old input syntactically incorrect (e.g. "%x.x"); COMPATIBILITY:
  4006 isatool fixdots ensures space after dots (e.g. "%x. x"); set
  4007 long_names for fully qualified output names; NOTE: ML programs
  4008 (special tactics, packages etc.) referring to internal names may have
  4009 to be adapted to cope with fully qualified names; in case of severe
  4010 backward campatibility problems try setting 'global_names' at compile
  4011 time to have enrything declared within a flat name space; one may also
  4012 fine tune name declarations in theories via the 'global' and 'local'
  4013 section;
  4014 
  4015 * reimplemented the implicit simpset and claset using the new anytype
  4016 data filed in signatures; references simpset:simpset ref etc. are
  4017 replaced by functions simpset:unit->simpset and
  4018 simpset_ref:unit->simpset ref; COMPATIBILITY: use isatool fixclasimp
  4019 to patch your ML files accordingly;
  4020 
  4021 * HTML output now includes theory graph data for display with Java
  4022 applet or isatool browser; data generated automatically via isatool
  4023 usedir (see -i option, ISABELLE_USEDIR_OPTIONS);
  4024 
  4025 * defs may now be conditional; improved rewrite_goals_tac to handle
  4026 conditional equations;
  4027 
  4028 * defs now admits additional type arguments, using TYPE('a) syntax;
  4029 
  4030 * theory aliases via merge (e.g. M=A+B+C) no longer supported, always
  4031 creates a new theory node; implicit merge of thms' signatures is
  4032 restricted to 'trivial' ones; COMPATIBILITY: one may have to use
  4033 transfer:theory->thm->thm in (rare) cases;
  4034 
  4035 * improved handling of draft signatures / theories; draft thms (and
  4036 ctyps, cterms) are automatically promoted to real ones;
  4037 
  4038 * slightly changed interfaces for oracles: admit many per theory, named
  4039 (e.g. oracle foo = mlfun), additional name argument for invoke_oracle;
  4040 
  4041 * print_goals: optional output of const types (set show_consts and
  4042 show_types);
  4043 
  4044 * improved output of warnings (###) and errors (***);
  4045 
  4046 * subgoal_tac displays a warning if the new subgoal has type variables;
  4047 
  4048 * removed old README and Makefiles;
  4049 
  4050 * replaced print_goals_ref hook by print_current_goals_fn and result_error_fn;
  4051 
  4052 * removed obsolete init_pps and init_database;
  4053 
  4054 * deleted the obsolete tactical STATE, which was declared by
  4055     fun STATE tacfun st = tacfun st st;
  4056 
  4057 * cd and use now support path variables, e.g. $ISABELLE_HOME, or ~
  4058 (which abbreviates $HOME);
  4059 
  4060 * changed Pure/Sequence interface (see Pure/seq.ML); COMPATIBILITY:
  4061 use isatool fixseq to adapt your ML programs (this works for fully
  4062 qualified references to the Sequence structure only!);
  4063 
  4064 * use_thy no longer requires writable current directory; it always
  4065 reloads .ML *and* .thy file, if either one is out of date;
  4066 
  4067 
  4068 *** Classical Reasoner ***
  4069 
  4070 * Clarify_tac, clarify_tac, clarify_step_tac, Clarify_step_tac: new
  4071 tactics that use classical reasoning to simplify a subgoal without
  4072 splitting it into several subgoals;
  4073 
  4074 * Safe_tac: like safe_tac but uses the default claset;
  4075 
  4076 
  4077 *** Simplifier ***
  4078 
  4079 * added simplification meta rules:
  4080     (asm_)(full_)simplify: simpset -> thm -> thm;
  4081 
  4082 * simplifier.ML no longer part of Pure -- has to be loaded by object
  4083 logics (again);
  4084 
  4085 * added prems argument to simplification procedures;
  4086 
  4087 * HOL, FOL, ZF: added infix function `addsplits':
  4088   instead of `<simpset> setloop (split_tac <thms>)'
  4089   you can simply write `<simpset> addsplits <thms>'
  4090 
  4091 
  4092 *** Syntax ***
  4093 
  4094 * TYPE('a) syntax for type reflection terms;
  4095 
  4096 * no longer handles consts with name "" -- declare as 'syntax' instead;
  4097 
  4098 * pretty printer: changed order of mixfix annotation preference (again!);
  4099 
  4100 * Pure: fixed idt/idts vs. pttrn/pttrns syntactic categories;
  4101 
  4102 
  4103 *** HOL ***
  4104 
  4105 * HOL: there is a new splitter `split_asm_tac' that can be used e.g.
  4106   with `addloop' of the simplifier to faciliate case splitting in premises.
  4107 
  4108 * HOL/TLA: Stephan Merz's formalization of Lamport's Temporal Logic of Actions;
  4109 
  4110 * HOL/Auth: new protocol proofs including some for the Internet
  4111   protocol TLS;
  4112 
  4113 * HOL/Map: new theory of `maps' a la VDM;
  4114 
  4115 * HOL/simplifier: simplification procedures nat_cancel_sums for
  4116 cancelling out common nat summands from =, <, <= (in)equalities, or
  4117 differences; simplification procedures nat_cancel_factor for
  4118 cancelling common factor from =, <, <= (in)equalities over natural
  4119 sums; nat_cancel contains both kinds of procedures, it is installed by
  4120 default in Arith.thy -- this COULD MAKE EXISTING PROOFS FAIL;
  4121 
  4122 * HOL/simplifier: terms of the form
  4123   `? x. P1(x) & ... & Pn(x) & x=t & Q1(x) & ... Qn(x)'  (or t=x)
  4124   are rewritten to
  4125   `P1(t) & ... & Pn(t) & Q1(t) & ... Qn(t)',
  4126   and those of the form
  4127   `! x. P1(x) & ... & Pn(x) & x=t & Q1(x) & ... Qn(x) --> R(x)'  (or t=x)
  4128   are rewritten to
  4129   `P1(t) & ... & Pn(t) & Q1(t) & ... Qn(t) --> R(t)',
  4130 
  4131 * HOL/datatype
  4132   Each datatype `t' now comes with a theorem `split_t_case' of the form
  4133 
  4134   P(t_case f1 ... fn x) =
  4135      ( (!y1 ... ym1. x = C1 y1 ... ym1 --> P(f1 y1 ... ym1)) &
  4136         ...
  4137        (!y1 ... ymn. x = Cn y1 ... ymn --> P(f1 y1 ... ymn))
  4138      )
  4139 
  4140   and a theorem `split_t_case_asm' of the form
  4141 
  4142   P(t_case f1 ... fn x) =
  4143     ~( (? y1 ... ym1. x = C1 y1 ... ym1 & ~P(f1 y1 ... ym1)) |
  4144         ...
  4145        (? y1 ... ymn. x = Cn y1 ... ymn & ~P(f1 y1 ... ymn))
  4146      )
  4147   which can be added to a simpset via `addsplits'. The existing theorems
  4148   expand_list_case and expand_option_case have been renamed to
  4149   split_list_case and split_option_case.
  4150 
  4151 * HOL/Arithmetic:
  4152   - `pred n' is automatically converted to `n-1'.
  4153     Users are strongly encouraged not to use `pred' any longer,
  4154     because it will disappear altogether at some point.
  4155   - Users are strongly encouraged to write "0 < n" rather than
  4156     "n ~= 0". Theorems and proof tools have been modified towards this
  4157     `standard'.
  4158 
  4159 * HOL/Lists:
  4160   the function "set_of_list" has been renamed "set" (and its theorems too);
  4161   the function "nth" now takes its arguments in the reverse order and
  4162   has acquired the infix notation "!" as in "xs!n".
  4163 
  4164 * HOL/Set: UNIV is now a constant and is no longer translated to Compl{};
  4165 
  4166 * HOL/Set: The operator (UN x.B x) now abbreviates (UN x:UNIV. B x) and its
  4167   specialist theorems (like UN1_I) are gone.  Similarly for (INT x.B x);
  4168 
  4169 * HOL/record: extensible records with schematic structural subtyping
  4170 (single inheritance); EXPERIMENTAL version demonstrating the encoding,
  4171 still lacks various theorems and concrete record syntax;
  4172 
  4173 
  4174 *** HOLCF ***
  4175 
  4176 * removed "axioms" and "generated by" sections;
  4177 
  4178 * replaced "ops" section by extended "consts" section, which is capable of
  4179   handling the continuous function space "->" directly;
  4180 
  4181 * domain package:
  4182   . proves theorems immediately and stores them in the theory,
  4183   . creates hierachical name space,
  4184   . now uses normal mixfix annotations (instead of cinfix...),
  4185   . minor changes to some names and values (for consistency),
  4186   . e.g. cases -> casedist, dists_eq -> dist_eqs, [take_lemma] -> take_lemmas,
  4187   . separator between mutual domain defs: changed "," to "and",
  4188   . improved handling of sort constraints;  now they have to
  4189     appear on the left-hand side of the equations only;
  4190 
  4191 * fixed LAM <x,y,zs>.b syntax;
  4192 
  4193 * added extended adm_tac to simplifier in HOLCF -- can now discharge
  4194 adm (%x. P (t x)), where P is chainfinite and t continuous;
  4195 
  4196 
  4197 *** FOL and ZF ***
  4198 
  4199 * FOL: there is a new splitter `split_asm_tac' that can be used e.g.
  4200   with `addloop' of the simplifier to faciliate case splitting in premises.
  4201 
  4202 * qed_spec_mp, qed_goal_spec_mp, qed_goalw_spec_mp are available, as
  4203 in HOL, they strip ALL and --> from proved theorems;
  4204 
  4205 
  4206 
  4207 New in Isabelle94-8 (May 1997)
  4208 ------------------------------
  4209 
  4210 *** General Changes ***
  4211 
  4212 * new utilities to build / run / maintain Isabelle etc. (in parts
  4213 still somewhat experimental); old Makefiles etc. still functional;
  4214 
  4215 * new 'Isabelle System Manual';
  4216 
  4217 * INSTALL text, together with ./configure and ./build scripts;
  4218 
  4219 * reimplemented type inference for greater efficiency, better error
  4220 messages and clean internal interface;
  4221 
  4222 * prlim command for dealing with lots of subgoals (an easier way of
  4223 setting goals_limit);
  4224 
  4225 
  4226 *** Syntax ***
  4227 
  4228 * supports alternative (named) syntax tables (parser and pretty
  4229 printer); internal interface is provided by add_modesyntax(_i);
  4230 
  4231 * Pure, FOL, ZF, HOL, HOLCF now support symbolic input and output; to
  4232 be used in conjunction with the Isabelle symbol font; uses the
  4233 "symbols" syntax table;
  4234 
  4235 * added token_translation interface (may translate name tokens in
  4236 arbitrary ways, dependent on their type (free, bound, tfree, ...) and
  4237 the current print_mode); IMPORTANT: user print translation functions
  4238 are responsible for marking newly introduced bounds
  4239 (Syntax.mark_boundT);
  4240 
  4241 * token translations for modes "xterm" and "xterm_color" that display
  4242 names in bold, underline etc. or colors (which requires a color
  4243 version of xterm);
  4244 
  4245 * infixes may now be declared with names independent of their syntax;
  4246 
  4247 * added typed_print_translation (like print_translation, but may
  4248 access type of constant);
  4249 
  4250 
  4251 *** Classical Reasoner ***
  4252 
  4253 Blast_tac: a new tactic!  It is often more powerful than fast_tac, but has
  4254 some limitations.  Blast_tac...
  4255   + ignores addss, addbefore, addafter; this restriction is intrinsic
  4256   + ignores elimination rules that don't have the correct format
  4257         (the conclusion MUST be a formula variable)
  4258   + ignores types, which can make HOL proofs fail
  4259   + rules must not require higher-order unification, e.g. apply_type in ZF
  4260     [message "Function Var's argument not a bound variable" relates to this]
  4261   + its proof strategy is more general but can actually be slower
  4262 
  4263 * substitution with equality assumptions no longer permutes other
  4264 assumptions;
  4265 
  4266 * minor changes in semantics of addafter (now called addaltern); renamed
  4267 setwrapper to setWrapper and compwrapper to compWrapper; added safe wrapper
  4268 (and access functions for it);
  4269 
  4270 * improved combination of classical reasoner and simplifier:
  4271   + functions for handling clasimpsets
  4272   + improvement of addss: now the simplifier is called _after_ the
  4273     safe steps.
  4274   + safe variant of addss called addSss: uses safe simplifications
  4275     _during_ the safe steps. It is more complete as it allows multiple
  4276     instantiations of unknowns (e.g. with slow_tac).
  4277 
  4278 *** Simplifier ***
  4279 
  4280 * added interface for simplification procedures (functions that
  4281 produce *proven* rewrite rules on the fly, depending on current
  4282 redex);
  4283 
  4284 * ordering on terms as parameter (used for ordered rewriting);
  4285 
  4286 * new functions delcongs, deleqcongs, and Delcongs. richer rep_ss;
  4287 
  4288 * the solver is now split into a safe and an unsafe part.
  4289 This should be invisible for the normal user, except that the
  4290 functions setsolver and addsolver have been renamed to setSolver and
  4291 addSolver; added safe_asm_full_simp_tac;
  4292 
  4293 
  4294 *** HOL ***
  4295 
  4296 * a generic induction tactic `induct_tac' which works for all datatypes and
  4297 also for type `nat';
  4298 
  4299 * a generic case distinction tactic `exhaust_tac' which works for all
  4300 datatypes and also for type `nat';
  4301 
  4302 * each datatype comes with a function `size';
  4303 
  4304 * patterns in case expressions allow tuple patterns as arguments to
  4305 constructors, for example `case x of [] => ... | (x,y,z)#ps => ...';
  4306 
  4307 * primrec now also works with type nat;
  4308 
  4309 * recdef: a new declaration form, allows general recursive functions to be
  4310 defined in theory files.  See HOL/ex/Fib, HOL/ex/Primes, HOL/Subst/Unify.
  4311 
  4312 * the constant for negation has been renamed from "not" to "Not" to
  4313 harmonize with FOL, ZF, LK, etc.;
  4314 
  4315 * HOL/ex/LFilter theory of a corecursive "filter" functional for
  4316 infinite lists;
  4317 
  4318 * HOL/Modelcheck demonstrates invocation of model checker oracle;
  4319 
  4320 * HOL/ex/Ring.thy declares cring_simp, which solves equational
  4321 problems in commutative rings, using axiomatic type classes for + and *;
  4322 
  4323 * more examples in HOL/MiniML and HOL/Auth;
  4324 
  4325 * more default rewrite rules for quantifiers, union/intersection;
  4326 
  4327 * a new constant `arbitrary == @x.False';
  4328 
  4329 * HOLCF/IOA replaces old HOL/IOA;
  4330 
  4331 * HOLCF changes: derived all rules and arities
  4332   + axiomatic type classes instead of classes
  4333   + typedef instead of faking type definitions
  4334   + eliminated the internal constants less_fun, less_cfun, UU_fun, UU_cfun etc.
  4335   + new axclasses cpo, chfin, flat with flat < chfin < pcpo < cpo < po
  4336   + eliminated the types void, one, tr
  4337   + use unit lift and bool lift (with translations) instead of one and tr
  4338   + eliminated blift from Lift3.thy (use Def instead of blift)
  4339   all eliminated rules are derived as theorems --> no visible changes ;
  4340 
  4341 
  4342 *** ZF ***
  4343 
  4344 * ZF now has Fast_tac, Simp_tac and Auto_tac.  Union_iff is a now a default
  4345 rewrite rule; this may affect some proofs.  eq_cs is gone but can be put back
  4346 as ZF_cs addSIs [equalityI];
  4347 
  4348 
  4349 
  4350 New in Isabelle94-7 (November 96)
  4351 ---------------------------------
  4352 
  4353 * allowing negative levels (as offsets) in prlev and choplev;
  4354 
  4355 * super-linear speedup for large simplifications;
  4356 
  4357 * FOL, ZF and HOL now use miniscoping: rewriting pushes
  4358 quantifications in as far as possible (COULD MAKE EXISTING PROOFS
  4359 FAIL); can suppress it using the command Delsimps (ex_simps @
  4360 all_simps); De Morgan laws are also now included, by default;
  4361 
  4362 * improved printing of ==>  :  ~:
  4363 
  4364 * new object-logic "Sequents" adds linear logic, while replacing LK
  4365 and Modal (thanks to Sara Kalvala);
  4366 
  4367 * HOL/Auth: correctness proofs for authentication protocols;
  4368 
  4369 * HOL: new auto_tac combines rewriting and classical reasoning (many
  4370 examples on HOL/Auth);
  4371 
  4372 * HOL: new command AddIffs for declaring theorems of the form P=Q to
  4373 the rewriter and classical reasoner simultaneously;
  4374 
  4375 * function uresult no longer returns theorems in "standard" format;
  4376 regain previous version by: val uresult = standard o uresult;
  4377 
  4378 
  4379 
  4380 New in Isabelle94-6
  4381 -------------------
  4382 
  4383 * oracles -- these establish an interface between Isabelle and trusted
  4384 external reasoners, which may deliver results as theorems;
  4385 
  4386 * proof objects (in particular record all uses of oracles);
  4387 
  4388 * Simp_tac, Fast_tac, etc. that refer to implicit simpset / claset;
  4389 
  4390 * "constdefs" section in theory files;
  4391 
  4392 * "primrec" section (HOL) no longer requires names;
  4393 
  4394 * internal type "tactic" now simply "thm -> thm Sequence.seq";
  4395 
  4396 
  4397 
  4398 New in Isabelle94-5
  4399 -------------------
  4400 
  4401 * reduced space requirements;
  4402 
  4403 * automatic HTML generation from theories;
  4404 
  4405 * theory files no longer require "..." (quotes) around most types;
  4406 
  4407 * new examples, including two proofs of the Church-Rosser theorem;
  4408 
  4409 * non-curried (1994) version of HOL is no longer distributed;
  4410 
  4411 
  4412 
  4413 New in Isabelle94-4
  4414 -------------------
  4415 
  4416 * greatly reduced space requirements;
  4417 
  4418 * theory files (.thy) no longer require \...\ escapes at line breaks;
  4419 
  4420 * searchable theorem database (see the section "Retrieving theorems" on
  4421 page 8 of the Reference Manual);
  4422 
  4423 * new examples, including Grabczewski's monumental case study of the
  4424 Axiom of Choice;
  4425 
  4426 * The previous version of HOL renamed to Old_HOL;
  4427 
  4428 * The new version of HOL (previously called CHOL) uses a curried syntax
  4429 for functions.  Application looks like f a b instead of f(a,b);
  4430 
  4431 * Mutually recursive inductive definitions finally work in HOL;
  4432 
  4433 * In ZF, pattern-matching on tuples is now available in all abstractions and
  4434 translates to the operator "split";
  4435 
  4436 
  4437 
  4438 New in Isabelle94-3
  4439 -------------------
  4440 
  4441 * new infix operator, addss, allowing the classical reasoner to
  4442 perform simplification at each step of its search.  Example:
  4443         fast_tac (cs addss ss)
  4444 
  4445 * a new logic, CHOL, the same as HOL, but with a curried syntax
  4446 for functions.  Application looks like f a b instead of f(a,b).  Also pairs
  4447 look like (a,b) instead of <a,b>;
  4448 
  4449 * PLEASE NOTE: CHOL will eventually replace HOL!
  4450 
  4451 * In CHOL, pattern-matching on tuples is now available in all abstractions.
  4452 It translates to the operator "split".  A new theory of integers is available;
  4453 
  4454 * In ZF, integer numerals now denote two's-complement binary integers.
  4455 Arithmetic operations can be performed by rewriting.  See ZF/ex/Bin.ML;
  4456 
  4457 * Many new examples: I/O automata, Church-Rosser theorem, equivalents
  4458 of the Axiom of Choice;
  4459 
  4460 
  4461 
  4462 New in Isabelle94-2
  4463 -------------------
  4464 
  4465 * Significantly faster resolution;
  4466 
  4467 * the different sections in a .thy file can now be mixed and repeated
  4468 freely;
  4469 
  4470 * Database of theorems for FOL, HOL and ZF.  New
  4471 commands including qed, qed_goal and bind_thm store theorems in the database.
  4472 
  4473 * Simple database queries: return a named theorem (get_thm) or all theorems of
  4474 a given theory (thms_of), or find out what theory a theorem was proved in
  4475 (theory_of_thm);
  4476 
  4477 * Bugs fixed in the inductive definition and datatype packages;
  4478 
  4479 * The classical reasoner provides deepen_tac and depth_tac, making FOL_dup_cs
  4480 and HOL_dup_cs obsolete;
  4481 
  4482 * Syntactic ambiguities caused by the new treatment of syntax in Isabelle94-1
  4483 have been removed;
  4484 
  4485 * Simpler definition of function space in ZF;
  4486 
  4487 * new results about cardinal and ordinal arithmetic in ZF;
  4488 
  4489 * 'subtype' facility in HOL for introducing new types as subsets of existing
  4490 types;
  4491 
  4492 
  4493 $Id$