NEWS
author wenzelm
Tue Sep 19 23:15:28 2006 +0200 (2006-09-19)
changeset 20622 e1a573146be5
parent 20620 8b26f58c5646
child 20712 b3cd1233167f
permissions -rw-r--r--
tuned proofs;
     1 Isabelle NEWS -- history user-relevant changes
     2 ==============================================
     3 
     4 New in this Isabelle release
     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") are now
    19 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 achieve
    48 Haskell-like type classes in Isabelle. See HOL/ex/Classpackage.thy for examples.
    49 
    50 * Yet another code generator framework allows to generate executable code
    51 for ML and Haskell (including "class"es). A short usage sketch:
    52 
    53     internal compilation:
    54         code_gen <list of constants (term syntax)> (SML -)
    55     writing SML code to a file:
    56         code_gen <list of constants (term syntax)> (SML <filename>)
    57     writing Haskell code to a bunch of files:
    58         code_gen <list of constants (term syntax)> (Haskell <filename>)
    59 
    60 Reasonable default setup of framework in HOL/Main.
    61 
    62 See HOL/ex/Code*.thy for examples.
    63 
    64 Theorem attributs for selecting and transforming function equations theorems:
    65 
    66     [code fun]:       select a theorem as function equation for a specific constant
    67     [code nofun]:     deselect a theorem as function equation for a specific constant
    68     [code inline]:    select an equation theorem for unfolding (inlining) in place
    69     [code noinline]:  deselect an equation theorem for unfolding (inlining) in place
    70 
    71 User-defined serializations (target in {SML, Haskell}):
    72 
    73     code_const <and-list of constants (term syntax)>
    74       {(target) <and-list of const target syntax>}+
    75 
    76     code_type <and-list of type constructors>
    77       {(target) <and-list of type target syntax>}+
    78 
    79     code_instance <and-list of instances>
    80       {(target)}+
    81         where instance ::= <type constructor> :: <class>
    82 
    83     code_class <and_list of classes>
    84       {(target) <and-list of class target syntax>}+
    85         where class target syntax ::= <class name> {where {<classop> == <target syntax>}+}?
    86 
    87 For code_instance and code_class, target SML is silently ignored.
    88 
    89 See HOL theories and HOL/ex/Code*.thy for usage examples.
    90 
    91 * Command 'no_translations' removes translation rules from theory
    92 syntax.
    93 
    94 * Overloaded definitions are now actually checked for acyclic
    95 dependencies.  The overloading scheme is slightly more general than
    96 that of Haskell98, although Isabelle does not demand an exact
    97 correspondence to type class and instance declarations.
    98 INCOMPATIBILITY, use ``defs (unchecked overloaded)'' to admit more
    99 exotic versions of overloading -- at the discretion of the user!
   100 
   101 Polymorphic constants are represented via type arguments, i.e. the
   102 instantiation that matches an instance against the most general
   103 declaration given in the signature.  For example, with the declaration
   104 c :: 'a => 'a => 'a, an instance c :: nat => nat => nat is represented
   105 as c(nat).  Overloading is essentially simultaneous structural
   106 recursion over such type arguments.  Incomplete specification patterns
   107 impose global constraints on all occurrences, e.g. c('a * 'a) on the
   108 LHS means that more general c('a * 'b) will be disallowed on any RHS.
   109 Command 'print_theory' outputs the normalized system of recursive
   110 equations, see section "definitions".
   111 
   112 * Isar: improper proof element 'guess' is like 'obtain', but derives
   113 the obtained context from the course of reasoning!  For example:
   114 
   115   assume "EX x y. A x & B y"   -- "any previous fact"
   116   then guess x and y by clarify
   117 
   118 This technique is potentially adventurous, depending on the facts and
   119 proof tools being involved here.
   120 
   121 * Isar: known facts from the proof context may be specified as literal
   122 propositions, using ASCII back-quote syntax.  This works wherever
   123 named facts used to be allowed so far, in proof commands, proof
   124 methods, attributes etc.  Literal facts are retrieved from the context
   125 according to unification of type and term parameters.  For example,
   126 provided that "A" and "A ==> B" and "!!x. P x ==> Q x" are known
   127 theorems in the current context, then these are valid literal facts:
   128 `A` and `A ==> B` and `!!x. P x ==> Q x" as well as `P a ==> Q a` etc.
   129 
   130 There is also a proof method "fact" which does the same composition
   131 for explicit goal states, e.g. the following proof texts coincide with
   132 certain special cases of literal facts:
   133 
   134   have "A" by fact                 ==  note `A`
   135   have "A ==> B" by fact           ==  note `A ==> B`
   136   have "!!x. P x ==> Q x" by fact  ==  note `!!x. P x ==> Q x`
   137   have "P a ==> Q a" by fact       ==  note `P a ==> Q a`
   138 
   139 * Isar: ":" (colon) is no longer a symbolic identifier character in
   140 outer syntax.  Thus symbolic identifiers may be used without
   141 additional white space in declarations like this: ``assume *: A''.
   142 
   143 * Isar: 'print_facts' prints all local facts of the current context,
   144 both named and unnamed ones.
   145 
   146 * Isar: 'def' now admits simultaneous definitions, e.g.:
   147 
   148   def x == "t" and y == "u"
   149 
   150 * Isar: added command 'unfolding', which is structurally similar to
   151 'using', but affects both the goal state and facts by unfolding given
   152 rewrite rules.  Thus many occurrences of the 'unfold' method or
   153 'unfolded' attribute may be replaced by first-class proof text.
   154 
   155 * Isar: methods 'unfold' / 'fold', attributes 'unfolded' / 'folded',
   156 and command 'unfolding' now all support object-level equalities
   157 (potentially conditional).  The underlying notion of rewrite rule is
   158 analogous to the 'rule_format' attribute, but *not* that of the
   159 Simplifier (which is usually more generous).
   160 
   161 * Isar: the goal restriction operator [N] (default N = 1) evaluates a
   162 method expression within a sandbox consisting of the first N
   163 sub-goals, which need to exist.  For example, ``simp_all [3]''
   164 simplifies the first three sub-goals, while (rule foo, simp_all)[]
   165 simplifies all new goals that emerge from applying rule foo to the
   166 originally first one.
   167 
   168 * Isar: schematic goals are no longer restricted to higher-order
   169 patterns; e.g. ``lemma "?P(?x)" by (rule TrueI)'' now works as
   170 expected.
   171 
   172 * Isar: the conclusion of a long theorem statement is now either
   173 'shows' (a simultaneous conjunction, as before), or 'obtains'
   174 (essentially a disjunction of cases with local parameters and
   175 assumptions).  The latter allows to express general elimination rules
   176 adequately; in this notation common elimination rules look like this:
   177 
   178   lemma exE:    -- "EX x. P x ==> (!!x. P x ==> thesis) ==> thesis"
   179     assumes "EX x. P x"
   180     obtains x where "P x"
   181 
   182   lemma conjE:  -- "A & B ==> (A ==> B ==> thesis) ==> thesis"
   183     assumes "A & B"
   184     obtains A and B
   185 
   186   lemma disjE:  -- "A | B ==> (A ==> thesis) ==> (B ==> thesis) ==> thesis"
   187     assumes "A | B"
   188     obtains
   189       A
   190     | B
   191 
   192 The subsequent classical rules even refer to the formal "thesis"
   193 explicitly:
   194 
   195   lemma classical:     -- "(~ thesis ==> thesis) ==> thesis"
   196     obtains "~ thesis"
   197 
   198   lemma Peirce's_Law:  -- "((thesis ==> something) ==> thesis) ==> thesis"
   199     obtains "thesis ==> something"
   200 
   201 The actual proof of an 'obtains' statement is analogous to that of the
   202 Isar proof element 'obtain', only that there may be several cases.
   203 Optional case names may be specified in parentheses; these will be
   204 available both in the present proof and as annotations in the
   205 resulting rule, for later use with the 'cases' method (cf. attribute
   206 case_names).
   207 
   208 * Isar: 'print_statement' prints theorems from the current theory or
   209 proof context in long statement form, according to the syntax of a
   210 top-level lemma.
   211 
   212 * Isar: 'obtain' takes an optional case name for the local context
   213 introduction rule (default "that").
   214 
   215 * Isar: removed obsolete 'concl is' patterns.  INCOMPATIBILITY, use
   216 explicit (is "_ ==> ?foo") in the rare cases where this still happens
   217 to occur.
   218 
   219 * Pure: syntax "CONST name" produces a fully internalized constant
   220 according to the current context.  This is particularly useful for
   221 syntax translations that should refer to internal constant
   222 representations independently of name spaces.
   223 
   224 * Isar/locales: 'const_syntax' provides a robust interface to the
   225 'syntax' primitive that also works in a locale context.  Type
   226 declaration and internal syntactic representation of given constants
   227 retrieved from the context.
   228 
   229 * Isar/locales: new derived specification elements 'axiomatization',
   230 'definition', 'abbreviation', which support type-inference, admit
   231 object-level specifications (equality, equivalence).  See also the
   232 isar-ref manual.  Examples:
   233 
   234   axiomatization
   235     eq  (infix "===" 50)
   236     where eq_refl: "x === x" and eq_subst: "x === y ==> P x ==> P y"
   237 
   238   definition
   239     "f x y = x + y + 1"
   240     "g x = f x x"
   241 
   242   abbreviation
   243     neq  (infix "=!=" 50)
   244     "x =!= y == ~ (x === y)"
   245 
   246 These specifications may be also used in a locale context.  Then the
   247 constants being introduced depend on certain fixed parameters, and the
   248 constant name is qualified by the locale base name.  An internal
   249 abbreviation takes care for convenient input and output, making the
   250 parameters implicit and using the original short name.  See also
   251 HOL/ex/Abstract_NAT.thy for an example of deriving polymorphic
   252 entities from a monomorphic theory.
   253 
   254 Presently, abbreviations are only available 'in' a target locale, but
   255 not inherited by general import expressions.  Also note that
   256 'abbreviation' may be used as a type-safe replacement for 'syntax' +
   257 'translations' in common applications.
   258 
   259 Concrete syntax is attached to specified constants in internal form,
   260 independently of name spaces.  The parse tree representation is
   261 slightly different -- use 'const_syntax' instead of raw 'syntax', and
   262 'translations' with explicit "CONST" markup to accommodate this.
   263 
   264 * Isar/locales: improved parameter handling:
   265 - use of locales "var" and "struct" no longer necessary;
   266 - parameter renamings are no longer required to be injective.
   267   This enables, for example, to define a locale for endomorphisms thus:
   268   locale endom = homom mult mult h.
   269 
   270 * Isar/locales: changed the way locales with predicates are defined.
   271 Instead of accumulating the specification, the imported expression is
   272 now an interpretation.
   273 INCOMPATIBILITY: different normal form of locale expressions.
   274 In particular, in interpretations of locales with predicates,
   275 goals repesenting already interpreted fragments are not removed
   276 automatically.  Use methods `intro_locales' and `unfold_locales'; see below.
   277 
   278 * Isar/locales: new methods `intro_locales' and `unfold_locales' provide
   279 backward reasoning on locales predicates.  The methods are aware of
   280 interpretations and discharge corresponding goals.  `intro_locales' is
   281 less aggressive then `unfold_locales' and does not unfold predicates to
   282 assumptions.
   283 
   284 * Isar/locales: the order in which locale fragments are accumulated
   285 has changed.  This enables to override declarations from fragments
   286 due to interpretations -- for example, unwanted simp rules.
   287 
   288 * Provers/induct: improved internal context management to support
   289 local fixes and defines on-the-fly.  Thus explicit meta-level
   290 connectives !! and ==> are rarely required anymore in inductive goals
   291 (using object-logic connectives for this purpose has been long
   292 obsolete anyway).  The subsequent proof patterns illustrate advanced
   293 techniques of natural induction; general datatypes and inductive sets
   294 work analogously (see also src/HOL/Lambda for realistic examples).
   295 
   296 (1) This is how to ``strengthen'' an inductive goal wrt. certain
   297 parameters:
   298 
   299   lemma
   300     fixes n :: nat and x :: 'a
   301     assumes a: "A n x"
   302     shows "P n x"
   303     using a                     -- {* make induct insert fact a *}
   304   proof (induct n arbitrary: x) -- {* generalize goal to "!!x. A n x ==> P n x" *}
   305     case 0
   306     show ?case sorry
   307   next
   308     case (Suc n)
   309     note `!!x. A n x ==> P n x` -- {* induction hypothesis, according to induction rule *}
   310     note `A (Suc n) x`          -- {* induction premise, stemming from fact a *}
   311     show ?case sorry
   312   qed
   313 
   314 (2) This is how to perform induction over ``expressions of a certain
   315 form'', using a locally defined inductive parameter n == "a x"
   316 together with strengthening (the latter is usually required to get
   317 sufficiently flexible induction hypotheses):
   318 
   319   lemma
   320     fixes a :: "'a => nat"
   321     assumes a: "A (a x)"
   322     shows "P (a x)"
   323     using a
   324   proof (induct n == "a x" arbitrary: x)
   325     ...
   326 
   327 See also HOL/Isar_examples/Puzzle.thy for an application of the this
   328 particular technique.
   329 
   330 (3) This is how to perform existential reasoning ('obtains' or
   331 'obtain') by induction, while avoiding explicit object-logic
   332 encodings:
   333 
   334   lemma
   335     fixes n :: nat
   336     obtains x :: 'a where "P n x" and "Q n x"
   337   proof (induct n arbitrary: thesis)
   338     case 0
   339     obtain x where "P 0 x" and "Q 0 x" sorry
   340     then show thesis by (rule 0)
   341   next
   342     case (Suc n)
   343     obtain x where "P n x" and "Q n x" by (rule Suc.hyps)
   344     obtain x where "P (Suc n) x" and "Q (Suc n) x" sorry
   345     then show thesis by (rule Suc.prems)
   346   qed
   347 
   348 Here the 'arbitrary: thesis' specification essentially modifies the
   349 scope of the formal thesis parameter, in order to the get the whole
   350 existence statement through the induction as expected.
   351 
   352 * Provers/induct: mutual induction rules are now specified as a list
   353 of rule sharing the same induction cases.  HOL packages usually
   354 provide foo_bar.inducts for mutually defined items foo and bar
   355 (e.g. inductive sets or datatypes).  INCOMPATIBILITY, users need to
   356 specify mutual induction rules differently, i.e. like this:
   357 
   358   (induct rule: foo_bar.inducts)
   359   (induct set: foo bar)
   360   (induct type: foo bar)
   361 
   362 The ML function ProjectRule.projections turns old-style rules into the
   363 new format.
   364 
   365 * Provers/induct: improved handling of simultaneous goals.  Instead of
   366 introducing object-level conjunction, the statement is now split into
   367 several conclusions, while the corresponding symbolic cases are
   368 nested accordingly.  INCOMPATIBILITY, proofs need to be structured
   369 explicitly.  For example:
   370 
   371   lemma
   372     fixes n :: nat
   373     shows "P n" and "Q n"
   374   proof (induct n)
   375     case 0 case 1
   376     show "P 0" sorry
   377   next
   378     case 0 case 2
   379     show "Q 0" sorry
   380   next
   381     case (Suc n) case 1
   382     note `P n` and `Q n`
   383     show "P (Suc n)" sorry
   384   next
   385     case (Suc n) case 2
   386     note `P n` and `Q n`
   387     show "Q (Suc n)" sorry
   388   qed
   389 
   390 The split into subcases may be deferred as follows -- this is
   391 particularly relevant for goal statements with local premises.
   392 
   393   lemma
   394     fixes n :: nat
   395     shows "A n ==> P n" and "B n ==> Q n"
   396   proof (induct n)
   397     case 0
   398     {
   399       case 1
   400       note `A 0`
   401       show "P 0" sorry
   402     next
   403       case 2
   404       note `B 0`
   405       show "Q 0" sorry
   406     }
   407   next
   408     case (Suc n)
   409     note `A n ==> P n` and `B n ==> Q n`
   410     {
   411       case 1
   412       note `A (Suc n)`
   413       show "P (Suc n)" sorry
   414     next
   415       case 2
   416       note `B (Suc n)`
   417       show "Q (Suc n)" sorry
   418     }
   419   qed
   420 
   421 If simultaneous goals are to be used with mutual rules, the statement
   422 needs to be structured carefully as a two-level conjunction, using
   423 lists of propositions separated by 'and':
   424 
   425   lemma
   426     shows "a : A ==> P1 a"
   427           "a : A ==> P2 a"
   428       and "b : B ==> Q1 b"
   429           "b : B ==> Q2 b"
   430           "b : B ==> Q3 b"
   431   proof (induct set: A B)
   432 
   433 * Provers/induct: support coinduction as well.  See
   434 src/HOL/Library/Coinductive_List.thy for various examples.
   435 
   436 * Simplifier: by default the simplifier trace only shows top level rewrites
   437 now. That is, trace_simp_depth_limit is set to 1 by default. Thus there is
   438 less danger of being flooded by the trace. The trace indicates where parts
   439 have been suppressed.
   440   
   441 * Provers/classical: removed obsolete classical version of elim_format
   442 attribute; classical elim/dest rules are now treated uniformly when
   443 manipulating the claset.
   444 
   445 * Provers/classical: stricter checks to ensure that supplied intro,
   446 dest and elim rules are well-formed; dest and elim rules must have at
   447 least one premise.
   448 
   449 * Provers/classical: attributes dest/elim/intro take an optional
   450 weight argument for the rule (just as the Pure versions).  Weights are
   451 ignored by automated tools, but determine the search order of single
   452 rule steps.
   453 
   454 * Syntax: input syntax now supports dummy variable binding "%_. b",
   455 where the body does not mention the bound variable.  Note that dummy
   456 patterns implicitly depend on their context of bounds, which makes
   457 "{_. _}" match any set comprehension as expected.  Potential
   458 INCOMPATIBILITY -- parse translations need to cope with syntactic
   459 constant "_idtdummy" in the binding position.
   460 
   461 * Syntax: removed obsolete syntactic constant "_K" and its associated
   462 parse translation.  INCOMPATIBILITY -- use dummy abstraction instead,
   463 for example "A -> B" => "Pi A (%_. B)".
   464 
   465 * Pure: 'class_deps' command visualizes the subclass relation, using
   466 the graph browser tool.
   467 
   468 * Pure: 'print_theory' now suppresses entities with internal name
   469 (trailing "_") by default; use '!' option for full details.
   470 
   471 
   472 *** HOL ***
   473 
   474 * New theory OperationalEquality providing class 'eq' with constant 'eq',
   475 allowing for code generation with polymorphic equality.
   476 
   477 * Numeral syntax: type 'bin' which was a mere type copy of 'int' has been
   478 abandoned in favour of plain 'int'. INCOMPATIBILITY -- Significant changes
   479 for setting up numeral syntax for types:
   480 
   481   - new constants Numeral.pred and Numeral.succ instead
   482       of former Numeral.bin_pred and Numeral.bin_succ.
   483   - Use integer operations instead of bin_add, bin_mult and so on.
   484   - Numeral simplification theorems named Numeral.numeral_simps instead of Bin_simps.
   485   - ML structure Bin_Simprocs now named Int_Numeral_Base_Simprocs.
   486 
   487 See HOL/Integ/IntArith.thy for an example setup.
   488 
   489 * New top level command 'normal_form' computes the normal form of a term
   490 that may contain free variables. For example 'normal_form "rev[a,b,c]"'
   491 prints '[b,c,a]'. This command is suitable for heavy-duty computations
   492 because the functions are compiled to ML first.
   493 INCOMPATIBILITY: new keywords 'normal_form' must quoted when used as
   494 an identifier.
   495 
   496 * Alternative iff syntax "A <-> B" for equality on bool (with priority
   497 25 like -->); output depends on the "iff" print_mode, the default is
   498 "A = B" (with priority 50).
   499 
   500 * Renamed constants in HOL.thy and Orderings.thy:
   501     op +   ~> HOL.plus
   502     op -   ~> HOL.minus
   503     uminus ~> HOL.uminus
   504     op *   ~> HOL.times
   505     op <   ~> Orderings.less
   506     op <=  ~> Orderings.less_eq
   507 
   508 Adaptions may be required in the following cases:
   509 
   510 a) User-defined constants using any of the names "plus", "minus", "times",
   511 "less" or "less_eq". The standard syntax translations for "+", "-" and "*"
   512 may go wrong.
   513 INCOMPATIBILITY: use more specific names.
   514 
   515 b) Variables named "plus", "minus", "times", "less", "less_eq"
   516 INCOMPATIBILITY: use more specific names.
   517 
   518 c) Permutative equations (e.g. "a + b = b + a")
   519 Since the change of names also changes the order of terms, permutative
   520 rewrite rules may get applied in a different order. Experience shows that
   521 this is rarely the case (only two adaptions in the whole Isabelle
   522 distribution).
   523 INCOMPATIBILITY: rewrite proofs
   524 
   525 d) ML code directly refering to constant names
   526 This in general only affects hand-written proof tactics, simprocs and so on.
   527 INCOMPATIBILITY: grep your sourcecode and replace names.
   528 
   529 * "LEAST x:A. P" expands to "LEAST x. x:A & P" (input only).
   530 
   531 * The old set interval syntax "{m..n(}" (and relatives) has been removed.
   532   Use "{m..<n}" (and relatives) instead.
   533 
   534 * In the context of the assumption "~(s = t)" the Simplifier rewrites
   535 "t = s" to False (by simproc "neq_simproc").  For backward
   536 compatibility this can be disabled by ML "reset use_neq_simproc".
   537 
   538 * "m dvd n" where m and n are numbers is evaluated to True/False by simp.
   539 
   540 * Theorem Cons_eq_map_conv no longer has attribute `simp'.
   541 
   542 * Theorem setsum_mult renamed to setsum_right_distrib.
   543 
   544 * Prefer ex1I over ex_ex1I in single-step reasoning, e.g. by the
   545 'rule' method.
   546 
   547 * Tactics 'sat' and 'satx' reimplemented, several improvements: goals
   548 no longer need to be stated as "<prems> ==> False", equivalences (i.e.
   549 "=" on type bool) are handled, variable names of the form "lit_<n>"
   550 are no longer reserved, significant speedup.
   551 
   552 * Tactics 'sat' and 'satx' can now replay MiniSat proof traces.  zChaff is
   553 still supported as well.
   554 
   555 * inductive and datatype: provide projections of mutual rules, bundled
   556 as foo_bar.inducts;
   557 
   558 * Library: theory Accessible_Part has been move to main HOL.
   559 
   560 * Library: added theory Coinductive_List of potentially infinite lists
   561 as greatest fixed-point.
   562 
   563 * Library: added theory AssocList which implements (finite) maps as
   564 association lists.
   565 
   566 * New proof method "evaluation" for efficiently solving a goal
   567   (i.e. a boolean expression) by compiling it to ML. The goal is
   568   "proved" (via the oracle "Evaluation") if it evaluates to True.
   569 
   570 * Linear arithmetic now splits certain operators (e.g. min, max, abs) also
   571 when invoked by the simplifier.  This results in the simplifier being more
   572 powerful on arithmetic goals.
   573 INCOMPATIBILTY: rewrite proofs.  Set fast_arith_split_limit to 0 to obtain
   574 the old behavior.
   575 
   576 * Support for hex (0x20) and binary (0b1001) numerals. 
   577 
   578 * New method:
   579   reify eqs (t), where eqs are equations for an interpretation 
   580  I :: 'a list => 'b => 'c and t::'c is an optional parameter,
   581  computes a term s::'b and a list xs::'a list and proves the theorem
   582   I xs s = t. This is also known as reification or quoting. The resulting theorem is applied to the subgoal to substitute t with I xs s.
   583 If t is omitted, the subgoal itself is reified.
   584 
   585 * New method:
   586  reflection corr_thm eqs (t). The parameters eqs and (t) are as explained above. corr_thm is a theorem for 
   587 I vs (f t) = I vs t, where f is supposed to be a computable function (in the sense of code generattion). The method uses reify to compute s and xs as above then applies corr_thm and uses normalization by evaluation to "prove" f s = r and finally gets the theorem t = r, which is again applied to the subgoal. An Example is available in HOL/ex/ReflectionEx.thy.
   588 
   589 * Reflection: Automatic refification now handels binding, an example is available in HOL/ex/ReflectionEx.thy
   590 *** HOL-Algebra ***
   591 
   592 * Method algebra is now set up via an attribute.  For examples see CRing.thy.
   593   INCOMPATIBILITY: the method is now weaker on combinations of algebraic
   594   structures.
   595 
   596 * Formalisation of ideals and the quotient construction over rings, contributed
   597   by Stephan Hohe.
   598 
   599 * Renamed `CRing.thy' to `Ring.thy'.  INCOMPATIBILITY.
   600 
   601 *** HOL-Complex ***
   602 
   603 * Theory Real: new method ferrack implements quantifier elimination
   604 for linear arithmetic over the reals. The quantifier elimination
   605 feature is used only for decision, for compatibility with arith. This
   606 means a goal is either solved or left unchanged, no simplification.
   607 
   608 
   609 *** ML ***
   610 
   611 * Pure/General/susp.ML:
   612 
   613 New module for delayed evaluations.
   614 
   615 * Pure/library:
   616 
   617 Semantically identical functions "equal_list" and "eq_list" have been
   618 unified to "eq_list".
   619 
   620 * Pure/library:
   621 
   622   val burrow: ('a list -> 'b list) -> 'a list list -> 'b list list
   623   val fold_burrow: ('a list -> 'c -> 'b list * 'd) -> 'a list list -> 'c -> 'b list list * 'd
   624 
   625 The semantics of "burrow" is: "take a function with *simulatanously*
   626 transforms a list of value, and apply it *simulatanously* to a list of
   627 list of values of the appropriate type". Confer this with "map" which
   628 would *not* apply its argument function simulatanously but in
   629 sequence. "fold_burrow" has an additional context.
   630 
   631 Both actually avoid the usage of "unflat" since they hide away
   632 "unflat" from the user.
   633 
   634 * Pure/library: functions map2 and fold2 with curried syntax for
   635 simultanous mapping and folding:
   636 
   637     val map2: ('a -> 'b -> 'c) -> 'a list -> 'b list -> 'c list
   638     val fold2: ('a -> 'b -> 'c -> 'c) -> 'a list -> 'b list -> 'c -> 'c
   639 
   640 * Pure/library: indexed lists - some functions in the Isabelle library
   641 treating lists over 'a as finite mappings from [0...n] to 'a have been
   642 given more convenient names and signatures reminiscent of similar
   643 functions for alists, tables, etc:
   644 
   645   val nth: 'a list -> int -> 'a 
   646   val nth_update: int * 'a -> 'a list -> 'a list
   647   val nth_map: int -> ('a -> 'a) -> 'a list -> 'a list
   648   val fold_index: (int * 'a -> 'b -> 'b) -> 'a list -> 'b -> 'b
   649 
   650 Note that fold_index starts counting at index 0, not 1 like foldln
   651 used to.
   652 
   653 * Pure/library: general ``divide_and_conquer'' combinator on lists.
   654 
   655 * Pure/General/name_mangler.ML provides a functor for generic name
   656 mangling (bijective mapping from any expression values to strings).
   657 
   658 * Pure/General/rat.ML implements rational numbers.
   659 
   660 * Pure/General/table.ML: the join operations now works via exceptions
   661 DUP/SAME instead of type option.  This is simpler in simple cases, and
   662 admits slightly more efficient complex applications.
   663 
   664 * Pure: datatype Context.generic joins theory/Proof.context and
   665 provides some facilities for code that works in either kind of
   666 context, notably GenericDataFun for uniform theory and proof data.
   667 
   668 * Pure: 'advanced' translation functions (parse_translation etc.) now
   669 use Context.generic instead of just theory.
   670 
   671 * Pure: simplified internal attribute type, which is now always
   672 Context.generic * thm -> Context.generic * thm.  Global (theory)
   673 vs. local (Proof.context) attributes have been discontinued, while
   674 minimizing code duplication.  Thm.rule_attribute and
   675 Thm.declaration_attribute build canonical attributes; see also
   676 structure Context for further operations on Context.generic, notably
   677 GenericDataFun.  INCOMPATIBILITY, need to adapt attribute type
   678 declarations and definitions.
   679 
   680 * Pure/kernel: consts certification ignores sort constraints given in
   681 signature declarations.  (This information is not relevant to the
   682 logic, but only for type inference.)  IMPORTANT INTERNAL CHANGE.
   683 
   684 * Pure: axiomatic type classes are now purely definitional, with
   685 explicit proofs of class axioms and super class relations performed
   686 internally.  See Pure/axclass.ML for the main internal interfaces --
   687 notably AxClass.define_class supercedes AxClass.add_axclass, and
   688 AxClass.axiomatize_class/classrel/arity supercede
   689 Sign.add_classes/classrel/arities.
   690 
   691 * Pure/Isar: Args/Attrib parsers operate on Context.generic --
   692 global/local versions on theory vs. Proof.context have been
   693 discontinued; Attrib.syntax and Method.syntax have been adapted
   694 accordingly.  INCOMPATIBILITY, need to adapt parser expressions for
   695 attributes, methods, etc.
   696 
   697 * Pure: several functions of signature "... -> theory -> theory * ..."
   698 have been reoriented to "... -> theory -> ... * theory" in order to
   699 allow natural usage in combination with the ||>, ||>>, |-> and
   700 fold_map combinators.
   701 
   702 * Pure: primitive rule lift_rule now takes goal cterm instead of an
   703 actual goal state (thm).  Use Thm.lift_rule (Thm.cprem_of st i) to
   704 achieve the old behaviour.
   705 
   706 * Pure: the "Goal" constant is now called "prop", supporting a
   707 slightly more general idea of ``protecting'' meta-level rule
   708 statements.
   709 
   710 * Pure: Logic.(un)varify only works in a global context, which is now
   711 enforced instead of silently assumed.  INCOMPATIBILITY, may use
   712 Logic.legacy_(un)varify as temporary workaround.
   713 
   714 * Pure: structure Name provides scalable operations for generating
   715 internal variable names, notably Name.variants etc.  This replaces
   716 some popular functions from term.ML:
   717 
   718   Term.variant		->  Name.variant
   719   Term.variantlist	->  Name.variant_list  (*canonical argument order*)
   720   Term.invent_names	->  Name.invent_list
   721 
   722 Note that low-level renaming rarely occurs in new code -- operations
   723 from structure Variable are used instead (see below).
   724 
   725 * Pure: structure Variable provides fundamental operations for proper
   726 treatment of fixed/schematic variables in a context.  For example,
   727 Variable.import introduces fixes for schematics of given facts and
   728 Variable.export reverses the effect (up to renaming) -- this replaces
   729 various freeze_thaw operations.
   730 
   731 * Pure: structure Goal provides simple interfaces for
   732 init/conclude/finish and tactical prove operations (replacing former
   733 Tactic.prove).  Goal.prove is the canonical way to prove results
   734 within a given context; Goal.prove_global is a degraded version for
   735 theory level goals, including a global Drule.standard.  Note that
   736 OldGoals.prove_goalw_cterm has long been obsolete, since it is
   737 ill-behaved in a local proof context (e.g. with local fixes/assumes or
   738 in a locale context).
   739 
   740 * Isar: simplified treatment of user-level errors, using exception
   741 ERROR of string uniformly.  Function error now merely raises ERROR,
   742 without any side effect on output channels.  The Isar toplevel takes
   743 care of proper display of ERROR exceptions.  ML code may use plain
   744 handle/can/try; cat_error may be used to concatenate errors like this:
   745 
   746   ... handle ERROR msg => cat_error msg "..."
   747 
   748 Toplevel ML code (run directly or through the Isar toplevel) may be
   749 embedded into the Isar toplevel with exception display/debug like
   750 this:
   751 
   752   Isar.toplevel (fn () => ...)
   753 
   754 INCOMPATIBILITY, removed special transform_error facilities, removed
   755 obsolete variants of user-level exceptions (ERROR_MESSAGE,
   756 Context.PROOF, ProofContext.CONTEXT, Proof.STATE, ProofHistory.FAIL)
   757 -- use plain ERROR instead.
   758 
   759 * Isar: theory setup now has type (theory -> theory), instead of a
   760 list.  INCOMPATIBILITY, may use #> to compose setup functions.
   761 
   762 * Isar: installed ML toplevel pretty printer for type Proof.context,
   763 subject to ProofContext.debug/verbose flags.
   764 
   765 * Isar: Toplevel.theory_to_proof admits transactions that modify the
   766 theory before entering a proof state.  Transactions now always see a
   767 quasi-functional intermediate checkpoint, both in interactive and
   768 batch mode.
   769 
   770 * Simplifier: the simpset of a running simplification process now
   771 contains a proof context (cf. Simplifier.the_context), which is the
   772 very context that the initial simpset has been retrieved from (by
   773 simpset_of/local_simpset_of).  Consequently, all plug-in components
   774 (solver, looper etc.) may depend on arbitrary proof data.
   775 
   776 * Simplifier.inherit_context inherits the proof context (plus the
   777 local bounds) of the current simplification process; any simproc
   778 etc. that calls the Simplifier recursively should do this!  Removed
   779 former Simplifier.inherit_bounds, which is already included here --
   780 INCOMPATIBILITY.  Tools based on low-level rewriting may even have to
   781 specify an explicit context using Simplifier.context/theory_context.
   782 
   783 * Simplifier/Classical Reasoner: more abstract interfaces
   784 change_simpset/claset for modifying the simpset/claset reference of a
   785 theory; raw versions simpset/claset_ref etc. have been discontinued --
   786 INCOMPATIBILITY.
   787 
   788 * Provers: more generic wrt. syntax of object-logics, avoid hardwired
   789 "Trueprop" etc.
   790 
   791 
   792 
   793 New in Isabelle2005 (October 2005)
   794 ----------------------------------
   795 
   796 *** General ***
   797 
   798 * Theory headers: the new header syntax for Isar theories is
   799 
   800   theory <name>
   801   imports <theory1> ... <theoryN>
   802   uses <file1> ... <fileM>
   803   begin
   804 
   805 where the 'uses' part is optional.  The previous syntax
   806 
   807   theory <name> = <theory1> + ... + <theoryN>:
   808 
   809 will disappear in the next release.  Use isatool fixheaders to convert
   810 existing theory files.  Note that there is no change in ancient
   811 non-Isar theories now, but these will disappear soon.
   812 
   813 * Theory loader: parent theories can now also be referred to via
   814 relative and absolute paths.
   815 
   816 * Command 'find_theorems' searches for a list of criteria instead of a
   817 list of constants. Known criteria are: intro, elim, dest, name:string,
   818 simp:term, and any term. Criteria can be preceded by '-' to select
   819 theorems that do not match. Intro, elim, dest select theorems that
   820 match the current goal, name:s selects theorems whose fully qualified
   821 name contain s, and simp:term selects all simplification rules whose
   822 lhs match term.  Any other term is interpreted as pattern and selects
   823 all theorems matching the pattern. Available in ProofGeneral under
   824 'ProofGeneral -> Find Theorems' or C-c C-f.  Example:
   825 
   826   C-c C-f (100) "(_::nat) + _ + _" intro -name: "HOL."
   827 
   828 prints the last 100 theorems matching the pattern "(_::nat) + _ + _",
   829 matching the current goal as introduction rule and not having "HOL."
   830 in their name (i.e. not being defined in theory HOL).
   831 
   832 * Command 'thms_containing' has been discontinued in favour of
   833 'find_theorems'; INCOMPATIBILITY.
   834 
   835 * Communication with Proof General is now 8bit clean, which means that
   836 Unicode text in UTF-8 encoding may be used within theory texts (both
   837 formal and informal parts).  Cf. option -U of the Isabelle Proof
   838 General interface.  Here are some simple examples (cf. src/HOL/ex):
   839 
   840   http://isabelle.in.tum.de/library/HOL/ex/Hebrew.html
   841   http://isabelle.in.tum.de/library/HOL/ex/Chinese.html
   842 
   843 * Improved efficiency of the Simplifier and, to a lesser degree, the
   844 Classical Reasoner.  Typical big applications run around 2 times
   845 faster.
   846 
   847 
   848 *** Document preparation ***
   849 
   850 * Commands 'display_drafts' and 'print_drafts' perform simple output
   851 of raw sources.  Only those symbols that do not require additional
   852 LaTeX packages (depending on comments in isabellesym.sty) are
   853 displayed properly, everything else is left verbatim.  isatool display
   854 and isatool print are used as front ends (these are subject to the
   855 DVI/PDF_VIEWER and PRINT_COMMAND settings, respectively).
   856 
   857 * Command tags control specific markup of certain regions of text,
   858 notably folding and hiding.  Predefined tags include "theory" (for
   859 theory begin and end), "proof" for proof commands, and "ML" for
   860 commands involving ML code; the additional tags "visible" and
   861 "invisible" are unused by default.  Users may give explicit tag
   862 specifications in the text, e.g. ''by %invisible (auto)''.  The
   863 interpretation of tags is determined by the LaTeX job during document
   864 preparation: see option -V of isatool usedir, or options -n and -t of
   865 isatool document, or even the LaTeX macros \isakeeptag, \isafoldtag,
   866 \isadroptag.
   867 
   868 Several document versions may be produced at the same time via isatool
   869 usedir (the generated index.html will link all of them).  Typical
   870 specifications include ''-V document=theory,proof,ML'' to present
   871 theory/proof/ML parts faithfully, ''-V outline=/proof,/ML'' to fold
   872 proof and ML commands, and ''-V mutilated=-theory,-proof,-ML'' to omit
   873 these parts without any formal replacement text.  The Isabelle site
   874 default settings produce ''document'' and ''outline'' versions as
   875 specified above.
   876 
   877 * Several new antiquotations:
   878 
   879   @{term_type term} prints a term with its type annotated;
   880 
   881   @{typeof term} prints the type of a term;
   882 
   883   @{const const} is the same as @{term const}, but checks that the
   884   argument is a known logical constant;
   885 
   886   @{term_style style term} and @{thm_style style thm} print a term or
   887   theorem applying a "style" to it
   888 
   889   @{ML text}
   890 
   891 Predefined styles are 'lhs' and 'rhs' printing the lhs/rhs of
   892 definitions, equations, inequations etc., 'concl' printing only the
   893 conclusion of a meta-logical statement theorem, and 'prem1' .. 'prem19'
   894 to print the specified premise.  TermStyle.add_style provides an ML
   895 interface for introducing further styles.  See also the "LaTeX Sugar"
   896 document practical applications.  The ML antiquotation prints
   897 type-checked ML expressions verbatim.
   898 
   899 * Markup commands 'chapter', 'section', 'subsection', 'subsubsection',
   900 and 'text' support optional locale specification '(in loc)', which
   901 specifies the default context for interpreting antiquotations.  For
   902 example: 'text (in lattice) {* @{thm inf_assoc}*}'.
   903 
   904 * Option 'locale=NAME' of antiquotations specifies an alternative
   905 context interpreting the subsequent argument.  For example: @{thm
   906 [locale=lattice] inf_assoc}.
   907 
   908 * Proper output of proof terms (@{prf ...} and @{full_prf ...}) within
   909 a proof context.
   910 
   911 * Proper output of antiquotations for theory commands involving a
   912 proof context (such as 'locale' or 'theorem (in loc) ...').
   913 
   914 * Delimiters of outer tokens (string etc.) now produce separate LaTeX
   915 macros (\isachardoublequoteopen, isachardoublequoteclose etc.).
   916 
   917 * isatool usedir: new option -C (default true) controls whether option
   918 -D should include a copy of the original document directory; -C false
   919 prevents unwanted effects such as copying of administrative CVS data.
   920 
   921 
   922 *** Pure ***
   923 
   924 * Considerably improved version of 'constdefs' command.  Now performs
   925 automatic type-inference of declared constants; additional support for
   926 local structure declarations (cf. locales and HOL records), see also
   927 isar-ref manual.  Potential INCOMPATIBILITY: need to observe strictly
   928 sequential dependencies of definitions within a single 'constdefs'
   929 section; moreover, the declared name needs to be an identifier.  If
   930 all fails, consider to fall back on 'consts' and 'defs' separately.
   931 
   932 * Improved indexed syntax and implicit structures.  First of all,
   933 indexed syntax provides a notational device for subscripted
   934 application, using the new syntax \<^bsub>term\<^esub> for arbitrary
   935 expressions.  Secondly, in a local context with structure
   936 declarations, number indexes \<^sub>n or the empty index (default
   937 number 1) refer to a certain fixed variable implicitly; option
   938 show_structs controls printing of implicit structures.  Typical
   939 applications of these concepts involve record types and locales.
   940 
   941 * New command 'no_syntax' removes grammar declarations (and
   942 translations) resulting from the given syntax specification, which is
   943 interpreted in the same manner as for the 'syntax' command.
   944 
   945 * 'Advanced' translation functions (parse_translation etc.) may depend
   946 on the signature of the theory context being presently used for
   947 parsing/printing, see also isar-ref manual.
   948 
   949 * Improved 'oracle' command provides a type-safe interface to turn an
   950 ML expression of type theory -> T -> term into a primitive rule of
   951 type theory -> T -> thm (i.e. the functionality of Thm.invoke_oracle
   952 is already included here); see also FOL/ex/IffExample.thy;
   953 INCOMPATIBILITY.
   954 
   955 * axclass: name space prefix for class "c" is now "c_class" (was "c"
   956 before); "cI" is no longer bound, use "c.intro" instead.
   957 INCOMPATIBILITY.  This change avoids clashes of fact bindings for
   958 axclasses vs. locales.
   959 
   960 * Improved internal renaming of symbolic identifiers -- attach primes
   961 instead of base 26 numbers.
   962 
   963 * New flag show_question_marks controls printing of leading question
   964 marks in schematic variable names.
   965 
   966 * In schematic variable names, *any* symbol following \<^isub> or
   967 \<^isup> is now treated as part of the base name.  For example, the
   968 following works without printing of awkward ".0" indexes:
   969 
   970   lemma "x\<^isub>1 = x\<^isub>2 ==> x\<^isub>2 = x\<^isub>1"
   971     by simp
   972 
   973 * Inner syntax includes (*(*nested*) comments*).
   974 
   975 * Pretty printer now supports unbreakable blocks, specified in mixfix
   976 annotations as "(00...)".
   977 
   978 * Clear separation of logical types and nonterminals, where the latter
   979 may only occur in 'syntax' specifications or type abbreviations.
   980 Before that distinction was only partially implemented via type class
   981 "logic" vs. "{}".  Potential INCOMPATIBILITY in rare cases of improper
   982 use of 'types'/'consts' instead of 'nonterminals'/'syntax'.  Some very
   983 exotic syntax specifications may require further adaption
   984 (e.g. Cube/Cube.thy).
   985 
   986 * Removed obsolete type class "logic", use the top sort {} instead.
   987 Note that non-logical types should be declared as 'nonterminals'
   988 rather than 'types'.  INCOMPATIBILITY for new object-logic
   989 specifications.
   990 
   991 * Attributes 'induct' and 'cases': type or set names may now be
   992 locally fixed variables as well.
   993 
   994 * Simplifier: can now control the depth to which conditional rewriting
   995 is traced via the PG menu Isabelle -> Settings -> Trace Simp Depth
   996 Limit.
   997 
   998 * Simplifier: simplification procedures may now take the current
   999 simpset into account (cf. Simplifier.simproc(_i) / mk_simproc
  1000 interface), which is very useful for calling the Simplifier
  1001 recursively.  Minor INCOMPATIBILITY: the 'prems' argument of simprocs
  1002 is gone -- use prems_of_ss on the simpset instead.  Moreover, the
  1003 low-level mk_simproc no longer applies Logic.varify internally, to
  1004 allow for use in a context of fixed variables.
  1005 
  1006 * thin_tac now works even if the assumption being deleted contains !!
  1007 or ==>.  More generally, erule now works even if the major premise of
  1008 the elimination rule contains !! or ==>.
  1009 
  1010 * Method 'rules' has been renamed to 'iprover'. INCOMPATIBILITY.
  1011 
  1012 * Reorganized bootstrapping of the Pure theories; CPure is now derived
  1013 from Pure, which contains all common declarations already.  Both
  1014 theories are defined via plain Isabelle/Isar .thy files.
  1015 INCOMPATIBILITY: elements of CPure (such as the CPure.intro /
  1016 CPure.elim / CPure.dest attributes) now appear in the Pure name space;
  1017 use isatool fixcpure to adapt your theory and ML sources.
  1018 
  1019 * New syntax 'name(i-j, i-, i, ...)' for referring to specific
  1020 selections of theorems in named facts via index ranges.
  1021 
  1022 * 'print_theorems': in theory mode, really print the difference
  1023 wrt. the last state (works for interactive theory development only),
  1024 in proof mode print all local facts (cf. 'print_facts');
  1025 
  1026 * 'hide': option '(open)' hides only base names.
  1027 
  1028 * More efficient treatment of intermediate checkpoints in interactive
  1029 theory development.
  1030 
  1031 * Code generator is now invoked via code_module (incremental code
  1032 generation) and code_library (modular code generation, ML structures
  1033 for each theory).  INCOMPATIBILITY: new keywords 'file' and 'contains'
  1034 must be quoted when used as identifiers.
  1035 
  1036 * New 'value' command for reading, evaluating and printing terms using
  1037 the code generator.  INCOMPATIBILITY: command keyword 'value' must be
  1038 quoted when used as identifier.
  1039 
  1040 
  1041 *** Locales ***
  1042 
  1043 * New commands for the interpretation of locale expressions in
  1044 theories (1), locales (2) and proof contexts (3).  These generate
  1045 proof obligations from the expression specification.  After the
  1046 obligations have been discharged, theorems of the expression are added
  1047 to the theory, target locale or proof context.  The synopsis of the
  1048 commands is a follows:
  1049 
  1050   (1) interpretation expr inst
  1051   (2) interpretation target < expr
  1052   (3) interpret expr inst
  1053 
  1054 Interpretation in theories and proof contexts require a parameter
  1055 instantiation of terms from the current context.  This is applied to
  1056 specifications and theorems of the interpreted expression.
  1057 Interpretation in locales only permits parameter renaming through the
  1058 locale expression.  Interpretation is smart in that interpretations
  1059 that are active already do not occur in proof obligations, neither are
  1060 instantiated theorems stored in duplicate.  Use 'print_interps' to
  1061 inspect active interpretations of a particular locale.  For details,
  1062 see the Isar Reference manual.  Examples can be found in
  1063 HOL/Finite_Set.thy and HOL/Algebra/UnivPoly.thy.
  1064 
  1065 INCOMPATIBILITY: former 'instantiate' has been withdrawn, use
  1066 'interpret' instead.
  1067 
  1068 * New context element 'constrains' for adding type constraints to
  1069 parameters.
  1070 
  1071 * Context expressions: renaming of parameters with syntax
  1072 redeclaration.
  1073 
  1074 * Locale declaration: 'includes' disallowed.
  1075 
  1076 * Proper static binding of attribute syntax -- i.e. types / terms /
  1077 facts mentioned as arguments are always those of the locale definition
  1078 context, independently of the context of later invocations.  Moreover,
  1079 locale operations (renaming and type / term instantiation) are applied
  1080 to attribute arguments as expected.
  1081 
  1082 INCOMPATIBILITY of the ML interface: always pass Attrib.src instead of
  1083 actual attributes; rare situations may require Attrib.attribute to
  1084 embed those attributes into Attrib.src that lack concrete syntax.
  1085 Attribute implementations need to cooperate properly with the static
  1086 binding mechanism.  Basic parsers Args.XXX_typ/term/prop and
  1087 Attrib.XXX_thm etc. already do the right thing without further
  1088 intervention.  Only unusual applications -- such as "where" or "of"
  1089 (cf. src/Pure/Isar/attrib.ML), which process arguments depending both
  1090 on the context and the facts involved -- may have to assign parsed
  1091 values to argument tokens explicitly.
  1092 
  1093 * Changed parameter management in theorem generation for long goal
  1094 statements with 'includes'.  INCOMPATIBILITY: produces a different
  1095 theorem statement in rare situations.
  1096 
  1097 * Locale inspection command 'print_locale' omits notes elements.  Use
  1098 'print_locale!' to have them included in the output.
  1099 
  1100 
  1101 *** Provers ***
  1102 
  1103 * Provers/hypsubst.ML: improved version of the subst method, for
  1104 single-step rewriting: it now works in bound variable contexts. New is
  1105 'subst (asm)', for rewriting an assumption.  INCOMPATIBILITY: may
  1106 rewrite a different subterm than the original subst method, which is
  1107 still available as 'simplesubst'.
  1108 
  1109 * Provers/quasi.ML: new transitivity reasoners for transitivity only
  1110 and quasi orders.
  1111 
  1112 * Provers/trancl.ML: new transitivity reasoner for transitive and
  1113 reflexive-transitive closure of relations.
  1114 
  1115 * Provers/blast.ML: new reference depth_limit to make blast's depth
  1116 limit (previously hard-coded with a value of 20) user-definable.
  1117 
  1118 * Provers/simplifier.ML has been moved to Pure, where Simplifier.setup
  1119 is peformed already.  Object-logics merely need to finish their
  1120 initial simpset configuration as before.  INCOMPATIBILITY.
  1121 
  1122 
  1123 *** HOL ***
  1124 
  1125 * Symbolic syntax of Hilbert Choice Operator is now as follows:
  1126 
  1127   syntax (epsilon)
  1128     "_Eps" :: "[pttrn, bool] => 'a"    ("(3\<some>_./ _)" [0, 10] 10)
  1129 
  1130 The symbol \<some> is displayed as the alternative epsilon of LaTeX
  1131 and x-symbol; use option '-m epsilon' to get it actually printed.
  1132 Moreover, the mathematically important symbolic identifier \<epsilon>
  1133 becomes available as variable, constant etc.  INCOMPATIBILITY,
  1134 
  1135 * "x > y" abbreviates "y < x" and "x >= y" abbreviates "y <= x".
  1136 Similarly for all quantifiers: "ALL x > y" etc.  The x-symbol for >=
  1137 is \<ge>. New transitivity rules have been added to HOL/Orderings.thy to
  1138 support corresponding Isar calculations.
  1139 
  1140 * "{x:A. P}" abbreviates "{x. x:A & P}", and similarly for "\<in>"
  1141 instead of ":".
  1142 
  1143 * theory SetInterval: changed the syntax for open intervals:
  1144 
  1145   Old       New
  1146   {..n(}    {..<n}
  1147   {)n..}    {n<..}
  1148   {m..n(}   {m..<n}
  1149   {)m..n}   {m<..n}
  1150   {)m..n(}  {m<..<n}
  1151 
  1152 The old syntax is still supported but will disappear in the next
  1153 release.  For conversion use the following Emacs search and replace
  1154 patterns (these are not perfect but work quite well):
  1155 
  1156   {)\([^\.]*\)\.\.  ->  {\1<\.\.}
  1157   \.\.\([^(}]*\)(}  ->  \.\.<\1}
  1158 
  1159 * Theory Commutative_Ring (in Library): method comm_ring for proving
  1160 equalities in commutative rings; method 'algebra' provides a generic
  1161 interface.
  1162 
  1163 * Theory Finite_Set: changed the syntax for 'setsum', summation over
  1164 finite sets: "setsum (%x. e) A", which used to be "\<Sum>x:A. e", is
  1165 now either "SUM x:A. e" or "\<Sum>x \<in> A. e". The bound variable can
  1166 be a tuple pattern.
  1167 
  1168 Some new syntax forms are available:
  1169 
  1170   "\<Sum>x | P. e"      for     "setsum (%x. e) {x. P}"
  1171   "\<Sum>x = a..b. e"   for     "setsum (%x. e) {a..b}"
  1172   "\<Sum>x = a..<b. e"  for     "setsum (%x. e) {a..<b}"
  1173   "\<Sum>x < k. e"      for     "setsum (%x. e) {..<k}"
  1174 
  1175 The latter form "\<Sum>x < k. e" used to be based on a separate
  1176 function "Summation", which has been discontinued.
  1177 
  1178 * theory Finite_Set: in structured induction proofs, the insert case
  1179 is now 'case (insert x F)' instead of the old counterintuitive 'case
  1180 (insert F x)'.
  1181 
  1182 * The 'refute' command has been extended to support a much larger
  1183 fragment of HOL, including axiomatic type classes, constdefs and
  1184 typedefs, inductive datatypes and recursion.
  1185 
  1186 * New tactics 'sat' and 'satx' to prove propositional tautologies.
  1187 Requires zChaff with proof generation to be installed.  See
  1188 HOL/ex/SAT_Examples.thy for examples.
  1189 
  1190 * Datatype induction via method 'induct' now preserves the name of the
  1191 induction variable. For example, when proving P(xs::'a list) by
  1192 induction on xs, the induction step is now P(xs) ==> P(a#xs) rather
  1193 than P(list) ==> P(a#list) as previously.  Potential INCOMPATIBILITY
  1194 in unstructured proof scripts.
  1195 
  1196 * Reworked implementation of records.  Improved scalability for
  1197 records with many fields, avoiding performance problems for type
  1198 inference. Records are no longer composed of nested field types, but
  1199 of nested extension types. Therefore the record type only grows linear
  1200 in the number of extensions and not in the number of fields.  The
  1201 top-level (users) view on records is preserved.  Potential
  1202 INCOMPATIBILITY only in strange cases, where the theory depends on the
  1203 old record representation. The type generated for a record is called
  1204 <record_name>_ext_type.
  1205 
  1206 Flag record_quick_and_dirty_sensitive can be enabled to skip the
  1207 proofs triggered by a record definition or a simproc (if
  1208 quick_and_dirty is enabled).  Definitions of large records can take
  1209 quite long.
  1210 
  1211 New simproc record_upd_simproc for simplification of multiple record
  1212 updates enabled by default.  Moreover, trivial updates are also
  1213 removed: r(|x := x r|) = r.  INCOMPATIBILITY: old proofs break
  1214 occasionally, since simplification is more powerful by default.
  1215 
  1216 * typedef: proper support for polymorphic sets, which contain extra
  1217 type-variables in the term.
  1218 
  1219 * Simplifier: automatically reasons about transitivity chains
  1220 involving "trancl" (r^+) and "rtrancl" (r^*) by setting up tactics
  1221 provided by Provers/trancl.ML as additional solvers.  INCOMPATIBILITY:
  1222 old proofs break occasionally as simplification may now solve more
  1223 goals than previously.
  1224 
  1225 * Simplifier: converts x <= y into x = y if assumption y <= x is
  1226 present.  Works for all partial orders (class "order"), in particular
  1227 numbers and sets.  For linear orders (e.g. numbers) it treats ~ x < y
  1228 just like y <= x.
  1229 
  1230 * Simplifier: new simproc for "let x = a in f x".  If a is a free or
  1231 bound variable or a constant then the let is unfolded.  Otherwise
  1232 first a is simplified to b, and then f b is simplified to g. If
  1233 possible we abstract b from g arriving at "let x = b in h x",
  1234 otherwise we unfold the let and arrive at g.  The simproc can be
  1235 enabled/disabled by the reference use_let_simproc.  Potential
  1236 INCOMPATIBILITY since simplification is more powerful by default.
  1237 
  1238 * Classical reasoning: the meson method now accepts theorems as arguments.
  1239 
  1240 * Prover support: pre-release of the Isabelle-ATP linkup, which runs background
  1241 jobs to provide advice on the provability of subgoals.
  1242 
  1243 * Theory OrderedGroup and Ring_and_Field: various additions and
  1244 improvements to faciliate calculations involving equalities and
  1245 inequalities.
  1246 
  1247 The following theorems have been eliminated or modified
  1248 (INCOMPATIBILITY):
  1249 
  1250   abs_eq             now named abs_of_nonneg
  1251   abs_of_ge_0        now named abs_of_nonneg
  1252   abs_minus_eq       now named abs_of_nonpos
  1253   imp_abs_id         now named abs_of_nonneg
  1254   imp_abs_neg_id     now named abs_of_nonpos
  1255   mult_pos           now named mult_pos_pos
  1256   mult_pos_le        now named mult_nonneg_nonneg
  1257   mult_pos_neg_le    now named mult_nonneg_nonpos
  1258   mult_pos_neg2_le   now named mult_nonneg_nonpos2
  1259   mult_neg           now named mult_neg_neg
  1260   mult_neg_le        now named mult_nonpos_nonpos
  1261 
  1262 * Theory Parity: added rules for simplifying exponents.
  1263 
  1264 * Theory List:
  1265 
  1266 The following theorems have been eliminated or modified
  1267 (INCOMPATIBILITY):
  1268 
  1269   list_all_Nil       now named list_all.simps(1)
  1270   list_all_Cons      now named list_all.simps(2)
  1271   list_all_conv      now named list_all_iff
  1272   set_mem_eq         now named mem_iff
  1273 
  1274 * Theories SetsAndFunctions and BigO (see HOL/Library) support
  1275 asymptotic "big O" calculations.  See the notes in BigO.thy.
  1276 
  1277 
  1278 *** HOL-Complex ***
  1279 
  1280 * Theory RealDef: better support for embedding natural numbers and
  1281 integers in the reals.
  1282 
  1283 The following theorems have been eliminated or modified
  1284 (INCOMPATIBILITY):
  1285 
  1286   exp_ge_add_one_self  now requires no hypotheses
  1287   real_of_int_add      reversed direction of equality (use [symmetric])
  1288   real_of_int_minus    reversed direction of equality (use [symmetric])
  1289   real_of_int_diff     reversed direction of equality (use [symmetric])
  1290   real_of_int_mult     reversed direction of equality (use [symmetric])
  1291 
  1292 * Theory RComplete: expanded support for floor and ceiling functions.
  1293 
  1294 * Theory Ln is new, with properties of the natural logarithm
  1295 
  1296 * Hyperreal: There is a new type constructor "star" for making
  1297 nonstandard types.  The old type names are now type synonyms:
  1298 
  1299   hypreal = real star
  1300   hypnat = nat star
  1301   hcomplex = complex star
  1302 
  1303 * Hyperreal: Many groups of similarly-defined constants have been
  1304 replaced by polymorphic versions (INCOMPATIBILITY):
  1305 
  1306   star_of <-- hypreal_of_real, hypnat_of_nat, hcomplex_of_complex
  1307 
  1308   starset      <-- starsetNat, starsetC
  1309   *s*          <-- *sNat*, *sc*
  1310   starset_n    <-- starsetNat_n, starsetC_n
  1311   *sn*         <-- *sNatn*, *scn*
  1312   InternalSets <-- InternalNatSets, InternalCSets
  1313 
  1314   starfun      <-- starfun{Nat,Nat2,C,RC,CR}
  1315   *f*          <-- *fNat*, *fNat2*, *fc*, *fRc*, *fcR*
  1316   starfun_n    <-- starfun{Nat,Nat2,C,RC,CR}_n
  1317   *fn*         <-- *fNatn*, *fNat2n*, *fcn*, *fRcn*, *fcRn*
  1318   InternalFuns <-- InternalNatFuns, InternalNatFuns2, Internal{C,RC,CR}Funs
  1319 
  1320 * Hyperreal: Many type-specific theorems have been removed in favor of
  1321 theorems specific to various axiomatic type classes (INCOMPATIBILITY):
  1322 
  1323   add_commute <-- {hypreal,hypnat,hcomplex}_add_commute
  1324   add_assoc   <-- {hypreal,hypnat,hcomplex}_add_assocs
  1325   OrderedGroup.add_0 <-- {hypreal,hypnat,hcomplex}_add_zero_left
  1326   OrderedGroup.add_0_right <-- {hypreal,hcomplex}_add_zero_right
  1327   right_minus <-- hypreal_add_minus
  1328   left_minus <-- {hypreal,hcomplex}_add_minus_left
  1329   mult_commute <-- {hypreal,hypnat,hcomplex}_mult_commute
  1330   mult_assoc <-- {hypreal,hypnat,hcomplex}_mult_assoc
  1331   mult_1_left <-- {hypreal,hypnat}_mult_1, hcomplex_mult_one_left
  1332   mult_1_right <-- hcomplex_mult_one_right
  1333   mult_zero_left <-- hcomplex_mult_zero_left
  1334   left_distrib <-- {hypreal,hypnat,hcomplex}_add_mult_distrib
  1335   right_distrib <-- hypnat_add_mult_distrib2
  1336   zero_neq_one <-- {hypreal,hypnat,hcomplex}_zero_not_eq_one
  1337   right_inverse <-- hypreal_mult_inverse
  1338   left_inverse <-- hypreal_mult_inverse_left, hcomplex_mult_inv_left
  1339   order_refl <-- {hypreal,hypnat}_le_refl
  1340   order_trans <-- {hypreal,hypnat}_le_trans
  1341   order_antisym <-- {hypreal,hypnat}_le_anti_sym
  1342   order_less_le <-- {hypreal,hypnat}_less_le
  1343   linorder_linear <-- {hypreal,hypnat}_le_linear
  1344   add_left_mono <-- {hypreal,hypnat}_add_left_mono
  1345   mult_strict_left_mono <-- {hypreal,hypnat}_mult_less_mono2
  1346   add_nonneg_nonneg <-- hypreal_le_add_order
  1347 
  1348 * Hyperreal: Separate theorems having to do with type-specific
  1349 versions of constants have been merged into theorems that apply to the
  1350 new polymorphic constants (INCOMPATIBILITY):
  1351 
  1352   STAR_UNIV_set <-- {STAR_real,NatStar_real,STARC_complex}_set
  1353   STAR_empty_set <-- {STAR,NatStar,STARC}_empty_set
  1354   STAR_Un <-- {STAR,NatStar,STARC}_Un
  1355   STAR_Int <-- {STAR,NatStar,STARC}_Int
  1356   STAR_Compl <-- {STAR,NatStar,STARC}_Compl
  1357   STAR_subset <-- {STAR,NatStar,STARC}_subset
  1358   STAR_mem <-- {STAR,NatStar,STARC}_mem
  1359   STAR_mem_Compl <-- {STAR,STARC}_mem_Compl
  1360   STAR_diff <-- {STAR,STARC}_diff
  1361   STAR_star_of_image_subset <-- {STAR_hypreal_of_real, NatStar_hypreal_of_real,
  1362     STARC_hcomplex_of_complex}_image_subset
  1363   starset_n_Un <-- starset{Nat,C}_n_Un
  1364   starset_n_Int <-- starset{Nat,C}_n_Int
  1365   starset_n_Compl <-- starset{Nat,C}_n_Compl
  1366   starset_n_diff <-- starset{Nat,C}_n_diff
  1367   InternalSets_Un <-- Internal{Nat,C}Sets_Un
  1368   InternalSets_Int <-- Internal{Nat,C}Sets_Int
  1369   InternalSets_Compl <-- Internal{Nat,C}Sets_Compl
  1370   InternalSets_diff <-- Internal{Nat,C}Sets_diff
  1371   InternalSets_UNIV_diff <-- Internal{Nat,C}Sets_UNIV_diff
  1372   InternalSets_starset_n <-- Internal{Nat,C}Sets_starset{Nat,C}_n
  1373   starset_starset_n_eq <-- starset{Nat,C}_starset{Nat,C}_n_eq
  1374   starset_n_starset <-- starset{Nat,C}_n_starset{Nat,C}
  1375   starfun_n_starfun <-- starfun{Nat,Nat2,C,RC,CR}_n_starfun{Nat,Nat2,C,RC,CR}
  1376   starfun <-- starfun{Nat,Nat2,C,RC,CR}
  1377   starfun_mult <-- starfun{Nat,Nat2,C,RC,CR}_mult
  1378   starfun_add <-- starfun{Nat,Nat2,C,RC,CR}_add
  1379   starfun_minus <-- starfun{Nat,Nat2,C,RC,CR}_minus
  1380   starfun_diff <-- starfun{C,RC,CR}_diff
  1381   starfun_o <-- starfun{NatNat2,Nat2,_stafunNat,C,C_starfunRC,_starfunCR}_o
  1382   starfun_o2 <-- starfun{NatNat2,_stafunNat,C,C_starfunRC,_starfunCR}_o2
  1383   starfun_const_fun <-- starfun{Nat,Nat2,C,RC,CR}_const_fun
  1384   starfun_inverse <-- starfun{Nat,C,RC,CR}_inverse
  1385   starfun_eq <-- starfun{Nat,Nat2,C,RC,CR}_eq
  1386   starfun_eq_iff <-- starfun{C,RC,CR}_eq_iff
  1387   starfun_Id <-- starfunC_Id
  1388   starfun_approx <-- starfun{Nat,CR}_approx
  1389   starfun_capprox <-- starfun{C,RC}_capprox
  1390   starfun_abs <-- starfunNat_rabs
  1391   starfun_lambda_cancel <-- starfun{C,CR,RC}_lambda_cancel
  1392   starfun_lambda_cancel2 <-- starfun{C,CR,RC}_lambda_cancel2
  1393   starfun_mult_HFinite_approx <-- starfunCR_mult_HFinite_capprox
  1394   starfun_mult_CFinite_capprox <-- starfun{C,RC}_mult_CFinite_capprox
  1395   starfun_add_capprox <-- starfun{C,RC}_add_capprox
  1396   starfun_add_approx <-- starfunCR_add_approx
  1397   starfun_inverse_inverse <-- starfunC_inverse_inverse
  1398   starfun_divide <-- starfun{C,CR,RC}_divide
  1399   starfun_n <-- starfun{Nat,C}_n
  1400   starfun_n_mult <-- starfun{Nat,C}_n_mult
  1401   starfun_n_add <-- starfun{Nat,C}_n_add
  1402   starfun_n_add_minus <-- starfunNat_n_add_minus
  1403   starfun_n_const_fun <-- starfun{Nat,C}_n_const_fun
  1404   starfun_n_minus <-- starfun{Nat,C}_n_minus
  1405   starfun_n_eq <-- starfun{Nat,C}_n_eq
  1406 
  1407   star_n_add <-- {hypreal,hypnat,hcomplex}_add
  1408   star_n_minus <-- {hypreal,hcomplex}_minus
  1409   star_n_diff <-- {hypreal,hcomplex}_diff
  1410   star_n_mult <-- {hypreal,hcomplex}_mult
  1411   star_n_inverse <-- {hypreal,hcomplex}_inverse
  1412   star_n_le <-- {hypreal,hypnat}_le
  1413   star_n_less <-- {hypreal,hypnat}_less
  1414   star_n_zero_num <-- {hypreal,hypnat,hcomplex}_zero_num
  1415   star_n_one_num <-- {hypreal,hypnat,hcomplex}_one_num
  1416   star_n_abs <-- hypreal_hrabs
  1417   star_n_divide <-- hcomplex_divide
  1418 
  1419   star_of_add <-- {hypreal_of_real,hypnat_of_nat,hcomplex_of_complex}_add
  1420   star_of_minus <-- {hypreal_of_real,hcomplex_of_complex}_minus
  1421   star_of_diff <-- hypreal_of_real_diff
  1422   star_of_mult <-- {hypreal_of_real,hypnat_of_nat,hcomplex_of_complex}_mult
  1423   star_of_one <-- {hypreal_of_real,hcomplex_of_complex}_one
  1424   star_of_zero <-- {hypreal_of_real,hypnat_of_nat,hcomplex_of_complex}_zero
  1425   star_of_le <-- {hypreal_of_real,hypnat_of_nat}_le_iff
  1426   star_of_less <-- {hypreal_of_real,hypnat_of_nat}_less_iff
  1427   star_of_eq <-- {hypreal_of_real,hypnat_of_nat,hcomplex_of_complex}_eq_iff
  1428   star_of_inverse <-- {hypreal_of_real,hcomplex_of_complex}_inverse
  1429   star_of_divide <-- {hypreal_of_real,hcomplex_of_complex}_divide
  1430   star_of_of_nat <-- {hypreal_of_real,hcomplex_of_complex}_of_nat
  1431   star_of_of_int <-- {hypreal_of_real,hcomplex_of_complex}_of_int
  1432   star_of_number_of <-- {hypreal,hcomplex}_number_of
  1433   star_of_number_less <-- number_of_less_hypreal_of_real_iff
  1434   star_of_number_le <-- number_of_le_hypreal_of_real_iff
  1435   star_of_eq_number <-- hypreal_of_real_eq_number_of_iff
  1436   star_of_less_number <-- hypreal_of_real_less_number_of_iff
  1437   star_of_le_number <-- hypreal_of_real_le_number_of_iff
  1438   star_of_power <-- hypreal_of_real_power
  1439   star_of_eq_0 <-- hcomplex_of_complex_zero_iff
  1440 
  1441 * Hyperreal: new method "transfer" that implements the transfer
  1442 principle of nonstandard analysis. With a subgoal that mentions
  1443 nonstandard types like "'a star", the command "apply transfer"
  1444 replaces it with an equivalent one that mentions only standard types.
  1445 To be successful, all free variables must have standard types; non-
  1446 standard variables must have explicit universal quantifiers.
  1447 
  1448 * Hyperreal: A theory of Taylor series.
  1449 
  1450 
  1451 *** HOLCF ***
  1452 
  1453 * Discontinued special version of 'constdefs' (which used to support
  1454 continuous functions) in favor of the general Pure one with full
  1455 type-inference.
  1456 
  1457 * New simplification procedure for solving continuity conditions; it
  1458 is much faster on terms with many nested lambda abstractions (cubic
  1459 instead of exponential time).
  1460 
  1461 * New syntax for domain package: selector names are now optional.
  1462 Parentheses should be omitted unless argument is lazy, for example:
  1463 
  1464   domain 'a stream = cons "'a" (lazy "'a stream")
  1465 
  1466 * New command 'fixrec' for defining recursive functions with pattern
  1467 matching; defining multiple functions with mutual recursion is also
  1468 supported.  Patterns may include the constants cpair, spair, up, sinl,
  1469 sinr, or any data constructor defined by the domain package. The given
  1470 equations are proven as rewrite rules. See HOLCF/ex/Fixrec_ex.thy for
  1471 syntax and examples.
  1472 
  1473 * New commands 'cpodef' and 'pcpodef' for defining predicate subtypes
  1474 of cpo and pcpo types. Syntax is exactly like the 'typedef' command,
  1475 but the proof obligation additionally includes an admissibility
  1476 requirement. The packages generate instances of class cpo or pcpo,
  1477 with continuity and strictness theorems for Rep and Abs.
  1478 
  1479 * HOLCF: Many theorems have been renamed according to a more standard naming
  1480 scheme (INCOMPATIBILITY):
  1481 
  1482   foo_inject:  "foo$x = foo$y ==> x = y"
  1483   foo_eq:      "(foo$x = foo$y) = (x = y)"
  1484   foo_less:    "(foo$x << foo$y) = (x << y)"
  1485   foo_strict:  "foo$UU = UU"
  1486   foo_defined: "... ==> foo$x ~= UU"
  1487   foo_defined_iff: "(foo$x = UU) = (x = UU)"
  1488 
  1489 
  1490 *** ZF ***
  1491 
  1492 * ZF/ex: theories Group and Ring provide examples in abstract algebra,
  1493 including the First Isomorphism Theorem (on quotienting by the kernel
  1494 of a homomorphism).
  1495 
  1496 * ZF/Simplifier: install second copy of type solver that actually
  1497 makes use of TC rules declared to Isar proof contexts (or locales);
  1498 the old version is still required for ML proof scripts.
  1499 
  1500 
  1501 *** Cube ***
  1502 
  1503 * Converted to Isar theory format; use locales instead of axiomatic
  1504 theories.
  1505 
  1506 
  1507 *** ML ***
  1508 
  1509 * Pure/library.ML no longer defines its own option datatype, but uses
  1510 that of the SML basis, which has constructors NONE and SOME instead of
  1511 None and Some, as well as exception Option.Option instead of OPTION.
  1512 The functions the, if_none, is_some, is_none have been adapted
  1513 accordingly, while Option.map replaces apsome.
  1514 
  1515 * Pure/library.ML: the exception LIST has been given up in favour of
  1516 the standard exceptions Empty and Subscript, as well as
  1517 Library.UnequalLengths.  Function like Library.hd and Library.tl are
  1518 superceded by the standard hd and tl functions etc.
  1519 
  1520 A number of basic list functions are no longer exported to the ML
  1521 toplevel, as they are variants of predefined functions.  The following
  1522 suggests how one can translate existing code:
  1523 
  1524     rev_append xs ys = List.revAppend (xs, ys)
  1525     nth_elem (i, xs) = List.nth (xs, i)
  1526     last_elem xs = List.last xs
  1527     flat xss = List.concat xss
  1528     seq fs = List.app fs
  1529     partition P xs = List.partition P xs
  1530     mapfilter f xs = List.mapPartial f xs
  1531 
  1532 * Pure/library.ML: several combinators for linear functional
  1533 transformations, notably reverse application and composition:
  1534 
  1535   x |> f                f #> g
  1536   (x, y) |-> f          f #-> g
  1537 
  1538 * Pure/library.ML: introduced/changed precedence of infix operators:
  1539 
  1540   infix 1 |> |-> ||> ||>> |>> |>>> #> #->;
  1541   infix 2 ?;
  1542   infix 3 o oo ooo oooo;
  1543   infix 4 ~~ upto downto;
  1544 
  1545 Maybe INCOMPATIBILITY when any of those is used in conjunction with other
  1546 infix operators.
  1547 
  1548 * Pure/library.ML: natural list combinators fold, fold_rev, and
  1549 fold_map support linear functional transformations and nesting.  For
  1550 example:
  1551 
  1552   fold f [x1, ..., xN] y =
  1553     y |> f x1 |> ... |> f xN
  1554 
  1555   (fold o fold) f [xs1, ..., xsN] y =
  1556     y |> fold f xs1 |> ... |> fold f xsN
  1557 
  1558   fold f [x1, ..., xN] =
  1559     f x1 #> ... #> f xN
  1560 
  1561   (fold o fold) f [xs1, ..., xsN] =
  1562     fold f xs1 #> ... #> fold f xsN
  1563 
  1564 * Pure/library.ML: the following selectors on type 'a option are
  1565 available:
  1566 
  1567   the:               'a option -> 'a  (*partial*)
  1568   these:             'a option -> 'a  where 'a = 'b list
  1569   the_default: 'a -> 'a option -> 'a
  1570   the_list:          'a option -> 'a list
  1571 
  1572 * Pure/General: structure AList (cf. Pure/General/alist.ML) provides
  1573 basic operations for association lists, following natural argument
  1574 order; moreover the explicit equality predicate passed here avoids
  1575 potentially expensive polymorphic runtime equality checks.
  1576 The old functions may be expressed as follows:
  1577 
  1578   assoc = uncurry (AList.lookup (op =))
  1579   assocs = these oo AList.lookup (op =)
  1580   overwrite = uncurry (AList.update (op =)) o swap
  1581 
  1582 * Pure/General: structure AList (cf. Pure/General/alist.ML) provides
  1583 
  1584   val make: ('a -> 'b) -> 'a list -> ('a * 'b) list
  1585   val find: ('a * 'b -> bool) -> ('c * 'b) list -> 'a -> 'c list
  1586 
  1587 replacing make_keylist and keyfilter (occassionally used)
  1588 Naive rewrites:
  1589 
  1590   make_keylist = AList.make
  1591   keyfilter = AList.find (op =)
  1592 
  1593 * eq_fst and eq_snd now take explicit equality parameter, thus
  1594   avoiding eqtypes. Naive rewrites:
  1595 
  1596     eq_fst = eq_fst (op =)
  1597     eq_snd = eq_snd (op =)
  1598 
  1599 * Removed deprecated apl and apr (rarely used).
  1600   Naive rewrites:
  1601 
  1602     apl (n, op) =>>= curry op n
  1603     apr (op, m) =>>= fn n => op (n, m)
  1604 
  1605 * Pure/General: structure OrdList (cf. Pure/General/ord_list.ML)
  1606 provides a reasonably efficient light-weight implementation of sets as
  1607 lists.
  1608 
  1609 * Pure/General: generic tables (cf. Pure/General/table.ML) provide a
  1610 few new operations; existing lookup and update are now curried to
  1611 follow natural argument order (for use with fold etc.);
  1612 INCOMPATIBILITY, use (uncurry Symtab.lookup) etc. as last resort.
  1613 
  1614 * Pure/General: output via the Isabelle channels of
  1615 writeln/warning/error etc. is now passed through Output.output, with a
  1616 hook for arbitrary transformations depending on the print_mode
  1617 (cf. Output.add_mode -- the first active mode that provides a output
  1618 function wins).  Already formatted output may be embedded into further
  1619 text via Output.raw; the result of Pretty.string_of/str_of and derived
  1620 functions (string_of_term/cterm/thm etc.) is already marked raw to
  1621 accommodate easy composition of diagnostic messages etc.  Programmers
  1622 rarely need to care about Output.output or Output.raw at all, with
  1623 some notable exceptions: Output.output is required when bypassing the
  1624 standard channels (writeln etc.), or in token translations to produce
  1625 properly formatted results; Output.raw is required when capturing
  1626 already output material that will eventually be presented to the user
  1627 a second time.  For the default print mode, both Output.output and
  1628 Output.raw have no effect.
  1629 
  1630 * Pure/General: Output.time_accumulator NAME creates an operator ('a
  1631 -> 'b) -> 'a -> 'b to measure runtime and count invocations; the
  1632 cumulative results are displayed at the end of a batch session.
  1633 
  1634 * Pure/General: File.sysify_path and File.quote_sysify path have been
  1635 replaced by File.platform_path and File.shell_path (with appropriate
  1636 hooks).  This provides a clean interface for unusual systems where the
  1637 internal and external process view of file names are different.
  1638 
  1639 * Pure: more efficient orders for basic syntactic entities: added
  1640 fast_string_ord, fast_indexname_ord, fast_term_ord; changed sort_ord
  1641 and typ_ord to use fast_string_ord and fast_indexname_ord (term_ord is
  1642 NOT affected); structures Symtab, Vartab, Typtab, Termtab use the fast
  1643 orders now -- potential INCOMPATIBILITY for code that depends on a
  1644 particular order for Symtab.keys, Symtab.dest, etc. (consider using
  1645 Library.sort_strings on result).
  1646 
  1647 * Pure/term.ML: combinators fold_atyps, fold_aterms, fold_term_types,
  1648 fold_types traverse types/terms from left to right, observing natural
  1649 argument order.  Supercedes previous foldl_XXX versions, add_frees,
  1650 add_vars etc. have been adapted as well: INCOMPATIBILITY.
  1651 
  1652 * Pure: name spaces have been refined, with significant changes of the
  1653 internal interfaces -- INCOMPATIBILITY.  Renamed cond_extern(_table)
  1654 to extern(_table).  The plain name entry path is superceded by a
  1655 general 'naming' context, which also includes the 'policy' to produce
  1656 a fully qualified name and external accesses of a fully qualified
  1657 name; NameSpace.extend is superceded by context dependent
  1658 Sign.declare_name.  Several theory and proof context operations modify
  1659 the naming context.  Especially note Theory.restore_naming and
  1660 ProofContext.restore_naming to get back to a sane state; note that
  1661 Theory.add_path is no longer sufficient to recover from
  1662 Theory.absolute_path in particular.
  1663 
  1664 * Pure: new flags short_names (default false) and unique_names
  1665 (default true) for controlling output of qualified names.  If
  1666 short_names is set, names are printed unqualified.  If unique_names is
  1667 reset, the name prefix is reduced to the minimum required to achieve
  1668 the original result when interning again, even if there is an overlap
  1669 with earlier declarations.
  1670 
  1671 * Pure/TheoryDataFun: change of the argument structure; 'prep_ext' is
  1672 now 'extend', and 'merge' gets an additional Pretty.pp argument
  1673 (useful for printing error messages).  INCOMPATIBILITY.
  1674 
  1675 * Pure: major reorganization of the theory context.  Type Sign.sg and
  1676 Theory.theory are now identified, referring to the universal
  1677 Context.theory (see Pure/context.ML).  Actual signature and theory
  1678 content is managed as theory data.  The old code and interfaces were
  1679 spread over many files and structures; the new arrangement introduces
  1680 considerable INCOMPATIBILITY to gain more clarity:
  1681 
  1682   Context -- theory management operations (name, identity, inclusion,
  1683     parents, ancestors, merge, etc.), plus generic theory data;
  1684 
  1685   Sign -- logical signature and syntax operations (declaring consts,
  1686     types, etc.), plus certify/read for common entities;
  1687 
  1688   Theory -- logical theory operations (stating axioms, definitions,
  1689     oracles), plus a copy of logical signature operations (consts,
  1690     types, etc.); also a few basic management operations (Theory.copy,
  1691     Theory.merge, etc.)
  1692 
  1693 The most basic sign_of operations (Theory.sign_of, Thm.sign_of_thm
  1694 etc.) as well as the sign field in Thm.rep_thm etc. have been retained
  1695 for convenience -- they merely return the theory.
  1696 
  1697 * Pure: type Type.tsig is superceded by theory in most interfaces.
  1698 
  1699 * Pure: the Isar proof context type is already defined early in Pure
  1700 as Context.proof (note that ProofContext.context and Proof.context are
  1701 aliases, where the latter is the preferred name).  This enables other
  1702 Isabelle components to refer to that type even before Isar is present.
  1703 
  1704 * Pure/sign/theory: discontinued named name spaces (i.e. classK,
  1705 typeK, constK, axiomK, oracleK), but provide explicit operations for
  1706 any of these kinds.  For example, Sign.intern typeK is now
  1707 Sign.intern_type, Theory.hide_space Sign.typeK is now
  1708 Theory.hide_types.  Also note that former
  1709 Theory.hide_classes/types/consts are now
  1710 Theory.hide_classes_i/types_i/consts_i, while the non '_i' versions
  1711 internalize their arguments!  INCOMPATIBILITY.
  1712 
  1713 * Pure: get_thm interface (of PureThy and ProofContext) expects
  1714 datatype thmref (with constructors Name and NameSelection) instead of
  1715 plain string -- INCOMPATIBILITY;
  1716 
  1717 * Pure: cases produced by proof methods specify options, where NONE
  1718 means to remove case bindings -- INCOMPATIBILITY in
  1719 (RAW_)METHOD_CASES.
  1720 
  1721 * Pure: the following operations retrieve axioms or theorems from a
  1722 theory node or theory hierarchy, respectively:
  1723 
  1724   Theory.axioms_of: theory -> (string * term) list
  1725   Theory.all_axioms_of: theory -> (string * term) list
  1726   PureThy.thms_of: theory -> (string * thm) list
  1727   PureThy.all_thms_of: theory -> (string * thm) list
  1728 
  1729 * Pure: print_tac now outputs the goal through the trace channel.
  1730 
  1731 * Isar toplevel: improved diagnostics, mostly for Poly/ML only.
  1732 Reference Toplevel.debug (default false) controls detailed printing
  1733 and tracing of low-level exceptions; Toplevel.profiling (default 0)
  1734 controls execution profiling -- set to 1 for time and 2 for space
  1735 (both increase the runtime).
  1736 
  1737 * Isar session: The initial use of ROOT.ML is now always timed,
  1738 i.e. the log will show the actual process times, in contrast to the
  1739 elapsed wall-clock time that the outer shell wrapper produces.
  1740 
  1741 * Simplifier: improved handling of bound variables (nameless
  1742 representation, avoid allocating new strings).  Simprocs that invoke
  1743 the Simplifier recursively should use Simplifier.inherit_bounds to
  1744 avoid local name clashes.  Failure to do so produces warnings
  1745 "Simplifier: renamed bound variable ..."; set Simplifier.debug_bounds
  1746 for further details.
  1747 
  1748 * ML functions legacy_bindings and use_legacy_bindings produce ML fact
  1749 bindings for all theorems stored within a given theory; this may help
  1750 in porting non-Isar theories to Isar ones, while keeping ML proof
  1751 scripts for the time being.
  1752 
  1753 * ML operator HTML.with_charset specifies the charset begin used for
  1754 generated HTML files.  For example:
  1755 
  1756   HTML.with_charset "utf-8" use_thy "Hebrew";
  1757   HTML.with_charset "utf-8" use_thy "Chinese";
  1758 
  1759 
  1760 *** System ***
  1761 
  1762 * Allow symlinks to all proper Isabelle executables (Isabelle,
  1763 isabelle, isatool etc.).
  1764 
  1765 * ISABELLE_DOC_FORMAT setting specifies preferred document format (for
  1766 isatool doc, isatool mkdir, display_drafts etc.).
  1767 
  1768 * isatool usedir: option -f allows specification of the ML file to be
  1769 used by Isabelle; default is ROOT.ML.
  1770 
  1771 * New isatool version outputs the version identifier of the Isabelle
  1772 distribution being used.
  1773 
  1774 * HOL: new isatool dimacs2hol converts files in DIMACS CNF format
  1775 (containing Boolean satisfiability problems) into Isabelle/HOL
  1776 theories.
  1777 
  1778 
  1779 
  1780 New in Isabelle2004 (April 2004)
  1781 --------------------------------
  1782 
  1783 *** General ***
  1784 
  1785 * Provers/order.ML:  new efficient reasoner for partial and linear orders.
  1786   Replaces linorder.ML.
  1787 
  1788 * Pure: Greek letters (except small lambda, \<lambda>), as well as Gothic
  1789   (\<aa>...\<zz>\<AA>...\<ZZ>), calligraphic (\<A>...\<Z>), and Euler
  1790   (\<a>...\<z>), are now considered normal letters, and can therefore
  1791   be used anywhere where an ASCII letter (a...zA...Z) has until
  1792   now. COMPATIBILITY: This obviously changes the parsing of some
  1793   terms, especially where a symbol has been used as a binder, say
  1794   '\<Pi>x. ...', which is now a type error since \<Pi>x will be parsed
  1795   as an identifier.  Fix it by inserting a space around former
  1796   symbols.  Call 'isatool fixgreek' to try to fix parsing errors in
  1797   existing theory and ML files.
  1798 
  1799 * Pure: Macintosh and Windows line-breaks are now allowed in theory files.
  1800 
  1801 * Pure: single letter sub/superscripts (\<^isub> and \<^isup>) are now
  1802   allowed in identifiers. Similar to Greek letters \<^isub> is now considered
  1803   a normal (but invisible) letter. For multiple letter subscripts repeat
  1804   \<^isub> like this: x\<^isub>1\<^isub>2.
  1805 
  1806 * Pure: There are now sub-/superscripts that can span more than one
  1807   character. Text between \<^bsub> and \<^esub> is set in subscript in
  1808   ProofGeneral and LaTeX, text between \<^bsup> and \<^esup> in
  1809   superscript. The new control characters are not identifier parts.
  1810 
  1811 * Pure: Control-symbols of the form \<^raw:...> will literally print the
  1812   content of "..." to the latex file instead of \isacntrl... . The "..."
  1813   may consist of any printable characters excluding the end bracket >.
  1814 
  1815 * Pure: Using new Isar command "finalconsts" (or the ML functions
  1816   Theory.add_finals or Theory.add_finals_i) it is now possible to
  1817   declare constants "final", which prevents their being given a definition
  1818   later.  It is useful for constants whose behaviour is fixed axiomatically
  1819   rather than definitionally, such as the meta-logic connectives.
  1820 
  1821 * Pure: 'instance' now handles general arities with general sorts
  1822   (i.e. intersections of classes),
  1823 
  1824 * Presentation: generated HTML now uses a CSS style sheet to make layout
  1825   (somewhat) independent of content. It is copied from lib/html/isabelle.css.
  1826   It can be changed to alter the colors/layout of generated pages.
  1827 
  1828 
  1829 *** Isar ***
  1830 
  1831 * Tactic emulation methods rule_tac, erule_tac, drule_tac, frule_tac,
  1832   cut_tac, subgoal_tac and thin_tac:
  1833   - Now understand static (Isar) contexts.  As a consequence, users of Isar
  1834     locales are no longer forced to write Isar proof scripts.
  1835     For details see Isar Reference Manual, paragraph 4.3.2: Further tactic
  1836     emulations.
  1837   - INCOMPATIBILITY: names of variables to be instantiated may no
  1838     longer be enclosed in quotes.  Instead, precede variable name with `?'.
  1839     This is consistent with the instantiation attribute "where".
  1840 
  1841 * Attributes "where" and "of":
  1842   - Now take type variables of instantiated theorem into account when reading
  1843     the instantiation string.  This fixes a bug that caused instantiated
  1844     theorems to have too special types in some circumstances.
  1845   - "where" permits explicit instantiations of type variables.
  1846 
  1847 * Calculation commands "moreover" and "also" no longer interfere with
  1848   current facts ("this"), admitting arbitrary combinations with "then"
  1849   and derived forms.
  1850 
  1851 * Locales:
  1852   - Goal statements involving the context element "includes" no longer
  1853     generate theorems with internal delta predicates (those ending on
  1854     "_axioms") in the premise.
  1855     Resolve particular premise with <locale>.intro to obtain old form.
  1856   - Fixed bug in type inference ("unify_frozen") that prevented mix of target
  1857     specification and "includes" elements in goal statement.
  1858   - Rule sets <locale>.intro and <locale>.axioms no longer declared as
  1859     [intro?] and [elim?] (respectively) by default.
  1860   - Experimental command for instantiation of locales in proof contexts:
  1861         instantiate <label>[<attrs>]: <loc>
  1862     Instantiates locale <loc> and adds all its theorems to the current context
  1863     taking into account their attributes.  Label and attrs are optional
  1864     modifiers, like in theorem declarations.  If present, names of
  1865     instantiated theorems are qualified with <label>, and the attributes
  1866     <attrs> are applied after any attributes these theorems might have already.
  1867       If the locale has assumptions, a chained fact of the form
  1868     "<loc> t1 ... tn" is expected from which instantiations of the parameters
  1869     are derived.  The command does not support old-style locales declared
  1870     with "locale (open)".
  1871       A few (very simple) examples can be found in FOL/ex/LocaleInst.thy.
  1872 
  1873 * HOL: Tactic emulation methods induct_tac and case_tac understand static
  1874   (Isar) contexts.
  1875 
  1876 
  1877 *** HOL ***
  1878 
  1879 * Proof import: new image HOL4 contains the imported library from
  1880   the HOL4 system with about 2500 theorems. It is imported by
  1881   replaying proof terms produced by HOL4 in Isabelle. The HOL4 image
  1882   can be used like any other Isabelle image.  See
  1883   HOL/Import/HOL/README for more information.
  1884 
  1885 * Simplifier:
  1886   - Much improved handling of linear and partial orders.
  1887     Reasoners for linear and partial orders are set up for type classes
  1888     "linorder" and "order" respectively, and are added to the default simpset
  1889     as solvers.  This means that the simplifier can build transitivity chains
  1890     to solve goals from the assumptions.
  1891   - INCOMPATIBILITY: old proofs break occasionally.  Typically, applications
  1892     of blast or auto after simplification become unnecessary because the goal
  1893     is solved by simplification already.
  1894 
  1895 * Numerics: new theory Ring_and_Field contains over 250 basic numerical laws,
  1896     all proved in axiomatic type classes for semirings, rings and fields.
  1897 
  1898 * Numerics:
  1899   - Numeric types (nat, int, and in HOL-Complex rat, real, complex, etc.) are
  1900     now formalized using the Ring_and_Field theory mentioned above.
  1901   - INCOMPATIBILITY: simplification and arithmetic behaves somewhat differently
  1902     than before, because now they are set up once in a generic manner.
  1903   - INCOMPATIBILITY: many type-specific arithmetic laws have gone.
  1904     Look for the general versions in Ring_and_Field (and Power if they concern
  1905     exponentiation).
  1906 
  1907 * Type "rat" of the rational numbers is now available in HOL-Complex.
  1908 
  1909 * Records:
  1910   - Record types are now by default printed with their type abbreviation
  1911     instead of the list of all field types. This can be configured via
  1912     the reference "print_record_type_abbr".
  1913   - Simproc "record_upd_simproc" for simplification of multiple updates added
  1914     (not enabled by default).
  1915   - Simproc "record_ex_sel_eq_simproc" to simplify EX x. sel r = x resp.
  1916     EX x. x = sel r to True (not enabled by default).
  1917   - Tactic "record_split_simp_tac" to split and simplify records added.
  1918 
  1919 * 'specification' command added, allowing for definition by
  1920   specification.  There is also an 'ax_specification' command that
  1921   introduces the new constants axiomatically.
  1922 
  1923 * arith(_tac) is now able to generate counterexamples for reals as well.
  1924 
  1925 * HOL-Algebra: new locale "ring" for non-commutative rings.
  1926 
  1927 * HOL-ex: InductiveInvariant_examples illustrates advanced recursive function
  1928   definitions, thanks to Sava Krsti\'{c} and John Matthews.
  1929 
  1930 * HOL-Matrix: a first theory for matrices in HOL with an application of
  1931   matrix theory to linear programming.
  1932 
  1933 * Unions and Intersections:
  1934   The latex output syntax of UN and INT has been changed
  1935   from "\Union x \in A. B" to "\Union_{x \in A} B"
  1936   i.e. the index formulae has become a subscript.
  1937   Similarly for "\Union x. B", and for \Inter instead of \Union.
  1938 
  1939 * Unions and Intersections over Intervals:
  1940   There is new short syntax "UN i<=n. A" for "UN i:{0..n}. A". There is
  1941   also an x-symbol version with subscripts "\<Union>\<^bsub>i <= n\<^esub>. A"
  1942   like in normal math, and corresponding versions for < and for intersection.
  1943 
  1944 * HOL/List: Ordering "lexico" is renamed "lenlex" and the standard
  1945   lexicographic dictonary ordering has been added as "lexord".
  1946 
  1947 * ML: the legacy theory structures Int and List have been removed. They had
  1948   conflicted with ML Basis Library structures having the same names.
  1949 
  1950 * 'refute' command added to search for (finite) countermodels.  Only works
  1951   for a fragment of HOL.  The installation of an external SAT solver is
  1952   highly recommended.  See "HOL/Refute.thy" for details.
  1953 
  1954 * 'quickcheck' command: Allows to find counterexamples by evaluating
  1955   formulae under an assignment of free variables to random values.
  1956   In contrast to 'refute', it can deal with inductive datatypes,
  1957   but cannot handle quantifiers. See "HOL/ex/Quickcheck_Examples.thy"
  1958   for examples.
  1959 
  1960 
  1961 *** HOLCF ***
  1962 
  1963 * Streams now come with concatenation and are part of the HOLCF image
  1964 
  1965 
  1966 
  1967 New in Isabelle2003 (May 2003)
  1968 ------------------------------
  1969 
  1970 *** General ***
  1971 
  1972 * Provers/simplifier:
  1973 
  1974   - Completely reimplemented method simp (ML: Asm_full_simp_tac):
  1975     Assumptions are now subject to complete mutual simplification,
  1976     not just from left to right. The simplifier now preserves
  1977     the order of assumptions.
  1978 
  1979     Potential INCOMPATIBILITY:
  1980 
  1981     -- simp sometimes diverges where the old version did
  1982        not, e.g. invoking simp on the goal
  1983 
  1984         [| P (f x); y = x; f x = f y |] ==> Q
  1985 
  1986        now gives rise to the infinite reduction sequence
  1987 
  1988         P(f x) --(f x = f y)--> P(f y) --(y = x)--> P(f x) --(f x = f y)--> ...
  1989 
  1990        Using "simp (asm_lr)" (ML: Asm_lr_simp_tac) instead often solves this
  1991        kind of problem.
  1992 
  1993     -- Tactics combining classical reasoner and simplification (such as auto)
  1994        are also affected by this change, because many of them rely on
  1995        simp. They may sometimes diverge as well or yield a different numbers
  1996        of subgoals. Try to use e.g. force, fastsimp, or safe instead of auto
  1997        in case of problems. Sometimes subsequent calls to the classical
  1998        reasoner will fail because a preceeding call to the simplifier too
  1999        eagerly simplified the goal, e.g. deleted redundant premises.
  2000 
  2001   - The simplifier trace now shows the names of the applied rewrite rules
  2002 
  2003   - You can limit the number of recursive invocations of the simplifier
  2004     during conditional rewriting (where the simplifie tries to solve the
  2005     conditions before applying the rewrite rule):
  2006     ML "simp_depth_limit := n"
  2007     where n is an integer. Thus you can force termination where previously
  2008     the simplifier would diverge.
  2009 
  2010   - Accepts free variables as head terms in congruence rules.  Useful in Isar.
  2011 
  2012   - No longer aborts on failed congruence proof.  Instead, the
  2013     congruence is ignored.
  2014 
  2015 * Pure: New generic framework for extracting programs from constructive
  2016   proofs. See HOL/Extraction.thy for an example instantiation, as well
  2017   as HOL/Extraction for some case studies.
  2018 
  2019 * Pure: The main goal of the proof state is no longer shown by default, only
  2020 the subgoals. This behaviour is controlled by a new flag.
  2021    PG menu: Isabelle/Isar -> Settings -> Show Main Goal
  2022 (ML: Proof.show_main_goal).
  2023 
  2024 * Pure: You can find all matching introduction rules for subgoal 1, i.e. all
  2025 rules whose conclusion matches subgoal 1:
  2026       PG menu: Isabelle/Isar -> Show me -> matching rules
  2027 The rules are ordered by how closely they match the subgoal.
  2028 In particular, rules that solve a subgoal outright are displayed first
  2029 (or rather last, the way they are printed).
  2030 (ML: ProofGeneral.print_intros())
  2031 
  2032 * Pure: New flag trace_unify_fail causes unification to print
  2033 diagnostic information (PG: in trace buffer) when it fails. This is
  2034 useful for figuring out why single step proofs like rule, erule or
  2035 assumption failed.
  2036 
  2037 * Pure: Locale specifications now produce predicate definitions
  2038 according to the body of text (covering assumptions modulo local
  2039 definitions); predicate "loc_axioms" covers newly introduced text,
  2040 while "loc" is cumulative wrt. all included locale expressions; the
  2041 latter view is presented only on export into the global theory
  2042 context; potential INCOMPATIBILITY, use "(open)" option to fall back
  2043 on the old view without predicates;
  2044 
  2045 * Pure: predefined locales "var" and "struct" are useful for sharing
  2046 parameters (as in CASL, for example); just specify something like
  2047 ``var x + var y + struct M'' as import;
  2048 
  2049 * Pure: improved thms_containing: proper indexing of facts instead of
  2050 raw theorems; check validity of results wrt. current name space;
  2051 include local facts of proof configuration (also covers active
  2052 locales), cover fixed variables in index; may use "_" in term
  2053 specification; an optional limit for the number of printed facts may
  2054 be given (the default is 40);
  2055 
  2056 * Pure: disallow duplicate fact bindings within new-style theory files
  2057 (batch-mode only);
  2058 
  2059 * Provers: improved induct method: assumptions introduced by case
  2060 "foo" are split into "foo.hyps" (from the rule) and "foo.prems" (from
  2061 the goal statement); "foo" still refers to all facts collectively;
  2062 
  2063 * Provers: the function blast.overloaded has been removed: all constants
  2064 are regarded as potentially overloaded, which improves robustness in exchange
  2065 for slight decrease in efficiency;
  2066 
  2067 * Provers/linorder: New generic prover for transitivity reasoning over
  2068 linear orders.  Note: this prover is not efficient!
  2069 
  2070 * Isar: preview of problems to finish 'show' now produce an error
  2071 rather than just a warning (in interactive mode);
  2072 
  2073 
  2074 *** HOL ***
  2075 
  2076 * arith(_tac)
  2077 
  2078  - Produces a counter example if it cannot prove a goal.
  2079    Note that the counter example may be spurious if the goal is not a formula
  2080    of quantifier-free linear arithmetic.
  2081    In ProofGeneral the counter example appears in the trace buffer.
  2082 
  2083  - Knows about div k and mod k where k is a numeral of type nat or int.
  2084 
  2085  - Calls full Presburger arithmetic (by Amine Chaieb) if quantifier-free
  2086    linear arithmetic fails. This takes account of quantifiers and divisibility.
  2087    Presburger arithmetic can also be called explicitly via presburger(_tac).
  2088 
  2089 * simp's arithmetic capabilities have been enhanced a bit: it now
  2090 takes ~= in premises into account (by performing a case split);
  2091 
  2092 * simp reduces "m*(n div m) + n mod m" to n, even if the two summands
  2093 are distributed over a sum of terms;
  2094 
  2095 * New tactic "trans_tac" and method "trans" instantiate
  2096 Provers/linorder.ML for axclasses "order" and "linorder" (predicates
  2097 "<=", "<" and "=").
  2098 
  2099 * function INCOMPATIBILITIES: Pi-sets have been redefined and moved from main
  2100 HOL to Library/FuncSet; constant "Fun.op o" is now called "Fun.comp";
  2101 
  2102 * 'typedef' command has new option "open" to suppress the set
  2103 definition;
  2104 
  2105 * functions Min and Max on finite sets have been introduced (theory
  2106 Finite_Set);
  2107 
  2108 * attribute [symmetric] now works for relations as well; it turns
  2109 (x,y) : R^-1 into (y,x) : R, and vice versa;
  2110 
  2111 * induct over a !!-quantified statement (say !!x1..xn):
  2112   each "case" automatically performs "fix x1 .. xn" with exactly those names.
  2113 
  2114 * Map: `empty' is no longer a constant but a syntactic abbreviation for
  2115 %x. None. Warning: empty_def now refers to the previously hidden definition
  2116 of the empty set.
  2117 
  2118 * Algebra: formalization of classical algebra.  Intended as base for
  2119 any algebraic development in Isabelle.  Currently covers group theory
  2120 (up to Sylow's theorem) and ring theory (Universal Property of
  2121 Univariate Polynomials).  Contributions welcome;
  2122 
  2123 * GroupTheory: deleted, since its material has been moved to Algebra;
  2124 
  2125 * Complex: new directory of the complex numbers with numeric constants,
  2126 nonstandard complex numbers, and some complex analysis, standard and
  2127 nonstandard (Jacques Fleuriot);
  2128 
  2129 * HOL-Complex: new image for analysis, replacing HOL-Real and HOL-Hyperreal;
  2130 
  2131 * Hyperreal: introduced Gauge integration and hyperreal logarithms (Jacques
  2132 Fleuriot);
  2133 
  2134 * Real/HahnBanach: updated and adapted to locales;
  2135 
  2136 * NumberTheory: added Gauss's law of quadratic reciprocity (by Avigad,
  2137 Gray and Kramer);
  2138 
  2139 * UNITY: added the Meier-Sanders theory of progress sets;
  2140 
  2141 * MicroJava: bytecode verifier and lightweight bytecode verifier
  2142 as abstract algorithms, instantiated to the JVM;
  2143 
  2144 * Bali: Java source language formalization. Type system, operational
  2145 semantics, axiomatic semantics. Supported language features:
  2146 classes, interfaces, objects,virtual methods, static methods,
  2147 static/instance fields, arrays, access modifiers, definite
  2148 assignment, exceptions.
  2149 
  2150 
  2151 *** ZF ***
  2152 
  2153 * ZF/Constructible: consistency proof for AC (Gdel's constructible
  2154 universe, etc.);
  2155 
  2156 * Main ZF: virtually all theories converted to new-style format;
  2157 
  2158 
  2159 *** ML ***
  2160 
  2161 * Pure: Tactic.prove provides sane interface for internal proofs;
  2162 omits the infamous "standard" operation, so this is more appropriate
  2163 than prove_goalw_cterm in many situations (e.g. in simprocs);
  2164 
  2165 * Pure: improved error reporting of simprocs;
  2166 
  2167 * Provers: Simplifier.simproc(_i) provides sane interface for setting
  2168 up simprocs;
  2169 
  2170 
  2171 *** Document preparation ***
  2172 
  2173 * uses \par instead of \\ for line breaks in theory text. This may
  2174 shift some page breaks in large documents. To get the old behaviour
  2175 use \renewcommand{\isanewline}{\mbox{}\\\mbox{}} in root.tex.
  2176 
  2177 * minimized dependencies of isabelle.sty and isabellesym.sty on
  2178 other packages
  2179 
  2180 * \<euro> now needs package babel/greek instead of marvosym (which
  2181 broke \Rightarrow)
  2182 
  2183 * normal size for \<zero>...\<nine> (uses \mathbf instead of
  2184 textcomp package)
  2185 
  2186 
  2187 
  2188 New in Isabelle2002 (March 2002)
  2189 --------------------------------
  2190 
  2191 *** Document preparation ***
  2192 
  2193 * greatly simplified document preparation setup, including more
  2194 graceful interpretation of isatool usedir -i/-d/-D options, and more
  2195 instructive isatool mkdir; users should basically be able to get
  2196 started with "isatool mkdir HOL Test && isatool make"; alternatively,
  2197 users may run a separate document processing stage manually like this:
  2198 "isatool usedir -D output HOL Test && isatool document Test/output";
  2199 
  2200 * theory dependency graph may now be incorporated into documents;
  2201 isatool usedir -g true will produce session_graph.eps/.pdf for use
  2202 with \includegraphics of LaTeX;
  2203 
  2204 * proper spacing of consecutive markup elements, especially text
  2205 blocks after section headings;
  2206 
  2207 * support bold style (for single symbols only), input syntax is like
  2208 this: "\<^bold>\<alpha>" or "\<^bold>A";
  2209 
  2210 * \<bullet> is now output as bold \cdot by default, which looks much
  2211 better in printed text;
  2212 
  2213 * added default LaTeX bindings for \<tturnstile> and \<TTurnstile>;
  2214 note that these symbols are currently unavailable in Proof General /
  2215 X-Symbol; new symbols \<zero>, \<one>, ..., \<nine>, and \<euro>;
  2216 
  2217 * isatool latex no longer depends on changed TEXINPUTS, instead
  2218 isatool document copies the Isabelle style files to the target
  2219 location;
  2220 
  2221 
  2222 *** Isar ***
  2223 
  2224 * Pure/Provers: improved proof by cases and induction;
  2225   - 'case' command admits impromptu naming of parameters (such as
  2226     "case (Suc n)");
  2227   - 'induct' method divinates rule instantiation from the inductive
  2228     claim; no longer requires excessive ?P bindings for proper
  2229     instantiation of cases;
  2230   - 'induct' method properly enumerates all possibilities of set/type
  2231     rules; as a consequence facts may be also passed through *type*
  2232     rules without further ado;
  2233   - 'induct' method now derives symbolic cases from the *rulified*
  2234     rule (before it used to rulify cases stemming from the internal
  2235     atomized version); this means that the context of a non-atomic
  2236     statement becomes is included in the hypothesis, avoiding the
  2237     slightly cumbersome show "PROP ?case" form;
  2238   - 'induct' may now use elim-style induction rules without chaining
  2239     facts, using ``missing'' premises from the goal state; this allows
  2240     rules stemming from inductive sets to be applied in unstructured
  2241     scripts, while still benefitting from proper handling of non-atomic
  2242     statements; NB: major inductive premises need to be put first, all
  2243     the rest of the goal is passed through the induction;
  2244   - 'induct' proper support for mutual induction involving non-atomic
  2245     rule statements (uses the new concept of simultaneous goals, see
  2246     below);
  2247   - append all possible rule selections, but only use the first
  2248     success (no backtracking);
  2249   - removed obsolete "(simplified)" and "(stripped)" options of methods;
  2250   - undeclared rule case names default to numbers 1, 2, 3, ...;
  2251   - added 'print_induct_rules' (covered by help item in recent Proof
  2252     General versions);
  2253   - moved induct/cases attributes to Pure, methods to Provers;
  2254   - generic method setup instantiated for FOL and HOL;
  2255 
  2256 * Pure: support multiple simultaneous goal statements, for example
  2257 "have a: A and b: B" (same for 'theorem' etc.); being a pure
  2258 meta-level mechanism, this acts as if several individual goals had
  2259 been stated separately; in particular common proof methods need to be
  2260 repeated in order to cover all claims; note that a single elimination
  2261 step is *not* sufficient to establish the two conjunctions, so this
  2262 fails:
  2263 
  2264   assume "A & B" then have A and B ..   (*".." fails*)
  2265 
  2266 better use "obtain" in situations as above; alternative refer to
  2267 multi-step methods like 'auto', 'simp_all', 'blast+' etc.;
  2268 
  2269 * Pure: proper integration with ``locales''; unlike the original
  2270 version by Florian Kammller, Isar locales package high-level proof
  2271 contexts rather than raw logical ones (e.g. we admit to include
  2272 attributes everywhere); operations on locales include merge and
  2273 rename; support for implicit arguments (``structures''); simultaneous
  2274 type-inference over imports and text; see also HOL/ex/Locales.thy for
  2275 some examples;
  2276 
  2277 * Pure: the following commands have been ``localized'', supporting a
  2278 target locale specification "(in name)": 'lemma', 'theorem',
  2279 'corollary', 'lemmas', 'theorems', 'declare'; the results will be
  2280 stored both within the locale and at the theory level (exported and
  2281 qualified by the locale name);
  2282 
  2283 * Pure: theory goals may now be specified in ``long'' form, with
  2284 ad-hoc contexts consisting of arbitrary locale elements. for example
  2285 ``lemma foo: fixes x assumes "A x" shows "B x"'' (local syntax and
  2286 definitions may be given, too); the result is a meta-level rule with
  2287 the context elements being discharged in the obvious way;
  2288 
  2289 * Pure: new proof command 'using' allows to augment currently used
  2290 facts after a goal statement ('using' is syntactically analogous to
  2291 'apply', but acts on the goal's facts only); this allows chained facts
  2292 to be separated into parts given before and after a claim, as in
  2293 ``from a and b have C using d and e <proof>'';
  2294 
  2295 * Pure: renamed "antecedent" case to "rule_context";
  2296 
  2297 * Pure: new 'judgment' command records explicit information about the
  2298 object-logic embedding (used by several tools internally); no longer
  2299 use hard-wired "Trueprop";
  2300 
  2301 * Pure: added 'corollary' command;
  2302 
  2303 * Pure: fixed 'token_translation' command;
  2304 
  2305 * Pure: removed obsolete 'exported' attribute;
  2306 
  2307 * Pure: dummy pattern "_" in is/let is now automatically lifted over
  2308 bound variables: "ALL x. P x --> Q x" (is "ALL x. _ --> ?C x")
  2309 supersedes more cumbersome ... (is "ALL x. _ x --> ?C x");
  2310 
  2311 * Pure: method 'atomize' presents local goal premises as object-level
  2312 statements (atomic meta-level propositions); setup controlled via
  2313 rewrite rules declarations of 'atomize' attribute; example
  2314 application: 'induct' method with proper rule statements in improper
  2315 proof *scripts*;
  2316 
  2317 * Pure: emulation of instantiation tactics (rule_tac, cut_tac, etc.)
  2318 now consider the syntactic context of assumptions, giving a better
  2319 chance to get type-inference of the arguments right (this is
  2320 especially important for locales);
  2321 
  2322 * Pure: "sorry" no longer requires quick_and_dirty in interactive
  2323 mode;
  2324 
  2325 * Pure/obtain: the formal conclusion "thesis", being marked as
  2326 ``internal'', may no longer be reference directly in the text;
  2327 potential INCOMPATIBILITY, may need to use "?thesis" in rare
  2328 situations;
  2329 
  2330 * Pure: generic 'sym' attribute which declares a rule both as pure
  2331 'elim?' and for the 'symmetric' operation;
  2332 
  2333 * Pure: marginal comments ``--'' may now occur just anywhere in the
  2334 text; the fixed correlation with particular command syntax has been
  2335 discontinued;
  2336 
  2337 * Pure: new method 'rules' is particularly well-suited for proof
  2338 search in intuitionistic logic; a bit slower than 'blast' or 'fast',
  2339 but often produces more compact proof terms with less detours;
  2340 
  2341 * Pure/Provers/classical: simplified integration with pure rule
  2342 attributes and methods; the classical "intro?/elim?/dest?"
  2343 declarations coincide with the pure ones; the "rule" method no longer
  2344 includes classically swapped intros; "intro" and "elim" methods no
  2345 longer pick rules from the context; also got rid of ML declarations
  2346 AddXIs/AddXEs/AddXDs; all of this has some potential for
  2347 INCOMPATIBILITY;
  2348 
  2349 * Provers/classical: attribute 'swapped' produces classical inversions
  2350 of introduction rules;
  2351 
  2352 * Provers/simplifier: 'simplified' attribute may refer to explicit
  2353 rules instead of full simplifier context; 'iff' attribute handles
  2354 conditional rules;
  2355 
  2356 * HOL: 'typedef' now allows alternative names for Rep/Abs morphisms;
  2357 
  2358 * HOL: 'recdef' now fails on unfinished automated proofs, use
  2359 "(permissive)" option to recover old behavior;
  2360 
  2361 * HOL: 'inductive' no longer features separate (collective) attributes
  2362 for 'intros' (was found too confusing);
  2363 
  2364 * HOL: properly declared induction rules less_induct and
  2365 wf_induct_rule;
  2366 
  2367 
  2368 *** HOL ***
  2369 
  2370 * HOL: moved over to sane numeral syntax; the new policy is as
  2371 follows:
  2372 
  2373   - 0 and 1 are polymorphic constants, which are defined on any
  2374   numeric type (nat, int, real etc.);
  2375 
  2376   - 2, 3, 4, ... and -1, -2, -3, ... are polymorphic numerals, based
  2377   binary representation internally;
  2378 
  2379   - type nat has special constructor Suc, and generally prefers Suc 0
  2380   over 1::nat and Suc (Suc 0) over 2::nat;
  2381 
  2382 This change may cause significant problems of INCOMPATIBILITY; here
  2383 are some hints on converting existing sources:
  2384 
  2385   - due to the new "num" token, "-0" and "-1" etc. are now atomic
  2386   entities, so expressions involving "-" (unary or binary minus) need
  2387   to be spaced properly;
  2388 
  2389   - existing occurrences of "1" may need to be constraint "1::nat" or
  2390   even replaced by Suc 0; similar for old "2";
  2391 
  2392   - replace "#nnn" by "nnn", and "#-nnn" by "-nnn";
  2393 
  2394   - remove all special provisions on numerals in proofs;
  2395 
  2396 * HOL: simp rules nat_number expand numerals on nat to Suc/0
  2397 representation (depends on bin_arith_simps in the default context);
  2398 
  2399 * HOL: symbolic syntax for x^2 (numeral 2);
  2400 
  2401 * HOL: the class of all HOL types is now called "type" rather than
  2402 "term"; INCOMPATIBILITY, need to adapt references to this type class
  2403 in axclass/classes, instance/arities, and (usually rare) occurrences
  2404 in typings (of consts etc.); internally the class is called
  2405 "HOL.type", ML programs should refer to HOLogic.typeS;
  2406 
  2407 * HOL/record package improvements:
  2408   - new derived operations "fields" to build a partial record section,
  2409     "extend" to promote a fixed record to a record scheme, and
  2410     "truncate" for the reverse; cf. theorems "xxx.defs", which are *not*
  2411     declared as simp by default;
  2412   - shared operations ("more", "fields", etc.) now need to be always
  2413     qualified) --- potential INCOMPATIBILITY;
  2414   - removed "make_scheme" operations (use "make" with "extend") --
  2415     INCOMPATIBILITY;
  2416   - removed "more" class (simply use "term") -- INCOMPATIBILITY;
  2417   - provides cases/induct rules for use with corresponding Isar
  2418     methods (for concrete records, record schemes, concrete more
  2419     parts, and schematic more parts -- in that order);
  2420   - internal definitions directly based on a light-weight abstract
  2421     theory of product types over typedef rather than datatype;
  2422 
  2423 * HOL: generic code generator for generating executable ML code from
  2424 specifications; specific support for HOL constructs such as inductive
  2425 datatypes and sets, as well as recursive functions; can be invoked
  2426 via 'generate_code' theory section;
  2427 
  2428 * HOL: canonical cases/induct rules for n-tuples (n = 3..7);
  2429 
  2430 * HOL: consolidated and renamed several theories.  In particular:
  2431         Ord.thy has been absorbed into HOL.thy
  2432         String.thy has been absorbed into List.thy
  2433 
  2434 * HOL: concrete setsum syntax "\<Sum>i:A. b" == "setsum (%i. b) A"
  2435 (beware of argument permutation!);
  2436 
  2437 * HOL: linorder_less_split superseded by linorder_cases;
  2438 
  2439 * HOL/List: "nodups" renamed to "distinct";
  2440 
  2441 * HOL: added "The" definite description operator; move Hilbert's "Eps"
  2442 to peripheral theory "Hilbert_Choice"; some INCOMPATIBILITIES:
  2443   - Ex_def has changed, now need to use some_eq_ex
  2444 
  2445 * HOL: made split_all_tac safe; EXISTING PROOFS MAY FAIL OR LOOP, so
  2446 in this (rare) case use:
  2447 
  2448   delSWrapper "split_all_tac"
  2449   addSbefore ("unsafe_split_all_tac", unsafe_split_all_tac)
  2450 
  2451 * HOL: added safe wrapper "split_conv_tac" to claset; EXISTING PROOFS
  2452 MAY FAIL;
  2453 
  2454 * HOL: introduced f^n = f o ... o f; warning: due to the limits of
  2455 Isabelle's type classes, ^ on functions and relations has too general
  2456 a domain, namely ('a * 'b) set and 'a => 'b; this means that it may be
  2457 necessary to attach explicit type constraints;
  2458 
  2459 * HOL/Relation: the prefix name of the infix "O" has been changed from
  2460 "comp" to "rel_comp"; INCOMPATIBILITY: a few theorems have been
  2461 renamed accordingly (eg "compI" -> "rel_compI").
  2462 
  2463 * HOL: syntax translations now work properly with numerals and records
  2464 expressions;
  2465 
  2466 * HOL: bounded abstraction now uses syntax "%" / "\<lambda>" instead
  2467 of "lam" -- INCOMPATIBILITY;
  2468 
  2469 * HOL: got rid of some global declarations (potential INCOMPATIBILITY
  2470 for ML tools): const "()" renamed "Product_Type.Unity", type "unit"
  2471 renamed "Product_Type.unit";
  2472 
  2473 * HOL: renamed rtrancl_into_rtrancl2 to converse_rtrancl_into_rtrancl
  2474 
  2475 * HOL: removed obsolete theorem "optionE" (use "option.exhaust", or
  2476 the "cases" method);
  2477 
  2478 * HOL/GroupTheory: group theory examples including Sylow's theorem (by
  2479 Florian Kammller);
  2480 
  2481 * HOL/IMP: updated and converted to new-style theory format; several
  2482 parts turned into readable document, with proper Isar proof texts and
  2483 some explanations (by Gerwin Klein);
  2484 
  2485 * HOL-Real: added Complex_Numbers (by Gertrud Bauer);
  2486 
  2487 * HOL-Hyperreal is now a logic image;
  2488 
  2489 
  2490 *** HOLCF ***
  2491 
  2492 * Isar: consts/constdefs supports mixfix syntax for continuous
  2493 operations;
  2494 
  2495 * Isar: domain package adapted to new-style theory format, e.g. see
  2496 HOLCF/ex/Dnat.thy;
  2497 
  2498 * theory Lift: proper use of rep_datatype lift instead of ML hacks --
  2499 potential INCOMPATIBILITY; now use plain induct_tac instead of former
  2500 lift.induct_tac, always use UU instead of Undef;
  2501 
  2502 * HOLCF/IMP: updated and converted to new-style theory;
  2503 
  2504 
  2505 *** ZF ***
  2506 
  2507 * Isar: proper integration of logic-specific tools and packages,
  2508 including theory commands '(co)inductive', '(co)datatype',
  2509 'rep_datatype', 'inductive_cases', as well as methods 'ind_cases',
  2510 'induct_tac', 'case_tac', and 'typecheck' (with attribute 'TC');
  2511 
  2512 * theory Main no longer includes AC; for the Axiom of Choice, base
  2513 your theory on Main_ZFC;
  2514 
  2515 * the integer library now covers quotients and remainders, with many
  2516 laws relating division to addition, multiplication, etc.;
  2517 
  2518 * ZF/UNITY: Chandy and Misra's UNITY is now available in ZF, giving a
  2519 typeless version of the formalism;
  2520 
  2521 * ZF/AC, Coind, IMP, Resid: updated and converted to new-style theory
  2522 format;
  2523 
  2524 * ZF/Induct: new directory for examples of inductive definitions,
  2525 including theory Multiset for multiset orderings; converted to
  2526 new-style theory format;
  2527 
  2528 * ZF: many new theorems about lists, ordinals, etc.;
  2529 
  2530 
  2531 *** General ***
  2532 
  2533 * Pure/kernel: meta-level proof terms (by Stefan Berghofer); reference
  2534 variable proof controls level of detail: 0 = no proofs (only oracle
  2535 dependencies), 1 = lemma dependencies, 2 = compact proof terms; see
  2536 also ref manual for further ML interfaces;
  2537 
  2538 * Pure/axclass: removed obsolete ML interface
  2539 goal_subclass/goal_arity;
  2540 
  2541 * Pure/syntax: new token syntax "num" for plain numerals (without "#"
  2542 of "xnum"); potential INCOMPATIBILITY, since -0, -1 etc. are now
  2543 separate tokens, so expressions involving minus need to be spaced
  2544 properly;
  2545 
  2546 * Pure/syntax: support non-oriented infixes, using keyword "infix"
  2547 rather than "infixl" or "infixr";
  2548 
  2549 * Pure/syntax: concrete syntax for dummy type variables admits genuine
  2550 sort constraint specifications in type inference; e.g. "x::_::foo"
  2551 ensures that the type of "x" is of sort "foo" (but not necessarily a
  2552 type variable);
  2553 
  2554 * Pure/syntax: print modes "type_brackets" and "no_type_brackets"
  2555 control output of nested => (types); the default behavior is
  2556 "type_brackets";
  2557 
  2558 * Pure/syntax: builtin parse translation for "_constify" turns valued
  2559 tokens into AST constants;
  2560 
  2561 * Pure/syntax: prefer later declarations of translations and print
  2562 translation functions; potential INCOMPATIBILITY: need to reverse
  2563 multiple declarations for same syntax element constant;
  2564 
  2565 * Pure/show_hyps reset by default (in accordance to existing Isar
  2566 practice);
  2567 
  2568 * Provers/classical: renamed addaltern to addafter, addSaltern to
  2569 addSafter;
  2570 
  2571 * Provers/clasimp: ``iff'' declarations now handle conditional rules
  2572 as well;
  2573 
  2574 * system: tested support for MacOS X; should be able to get Isabelle +
  2575 Proof General to work in a plain Terminal after installing Poly/ML
  2576 (e.g. from the Isabelle distribution area) and GNU bash alone
  2577 (e.g. from http://www.apple.com); full X11, XEmacs and X-Symbol
  2578 support requires further installations, e.g. from
  2579 http://fink.sourceforge.net/);
  2580 
  2581 * system: support Poly/ML 4.1.1 (able to manage larger heaps);
  2582 
  2583 * system: reduced base memory usage by Poly/ML (approx. 20 MB instead
  2584 of 40 MB), cf. ML_OPTIONS;
  2585 
  2586 * system: Proof General keywords specification is now part of the
  2587 Isabelle distribution (see etc/isar-keywords.el);
  2588 
  2589 * system: support for persistent Proof General sessions (refrain from
  2590 outdating all loaded theories on startup); user may create writable
  2591 logic images like this: ``isabelle -q HOL Test'';
  2592 
  2593 * system: smart selection of Isabelle process versus Isabelle
  2594 interface, accommodates case-insensitive file systems (e.g. HFS+); may
  2595 run both "isabelle" and "Isabelle" even if file names are badly
  2596 damaged (executable inspects the case of the first letter of its own
  2597 name); added separate "isabelle-process" and "isabelle-interface";
  2598 
  2599 * system: refrain from any attempt at filtering input streams; no
  2600 longer support ``8bit'' encoding of old isabelle font, instead proper
  2601 iso-latin characters may now be used; the related isatools
  2602 "symbolinput" and "nonascii" have disappeared as well;
  2603 
  2604 * system: removed old "xterm" interface (the print modes "xterm" and
  2605 "xterm_color" are still available for direct use in a suitable
  2606 terminal);
  2607 
  2608 
  2609 
  2610 New in Isabelle99-2 (February 2001)
  2611 -----------------------------------
  2612 
  2613 *** Overview of INCOMPATIBILITIES ***
  2614 
  2615 * HOL: please note that theories in the Library and elsewhere often use the
  2616 new-style (Isar) format; to refer to their theorems in an ML script you must
  2617 bind them to ML identifers by e.g.      val thm_name = thm "thm_name";
  2618 
  2619 * HOL: inductive package no longer splits induction rule aggressively,
  2620 but only as far as specified by the introductions given; the old
  2621 format may be recovered via ML function complete_split_rule or attribute
  2622 'split_rule (complete)';
  2623 
  2624 * HOL: induct renamed to lfp_induct, lfp_Tarski to lfp_unfold,
  2625 gfp_Tarski to gfp_unfold;
  2626 
  2627 * HOL: contrapos, contrapos2 renamed to contrapos_nn, contrapos_pp;
  2628 
  2629 * HOL: infix "dvd" now has priority 50 rather than 70 (because it is a
  2630 relation); infix "^^" has been renamed "``"; infix "``" has been
  2631 renamed "`"; "univalent" has been renamed "single_valued";
  2632 
  2633 * HOL/Real: "rinv" and "hrinv" replaced by overloaded "inverse"
  2634 operation;
  2635 
  2636 * HOLCF: infix "`" has been renamed "$"; the symbol syntax is \<cdot>;
  2637 
  2638 * Isar: 'obtain' no longer declares "that" fact as simp/intro;
  2639 
  2640 * Isar/HOL: method 'induct' now handles non-atomic goals; as a
  2641 consequence, it is no longer monotonic wrt. the local goal context
  2642 (which is now passed through the inductive cases);
  2643 
  2644 * Document preparation: renamed standard symbols \<ll> to \<lless> and
  2645 \<gg> to \<ggreater>;
  2646 
  2647 
  2648 *** Document preparation ***
  2649 
  2650 * \isabellestyle{NAME} selects version of Isabelle output (currently
  2651 available: are "it" for near math-mode best-style output, "sl" for
  2652 slanted text style, and "tt" for plain type-writer; if no
  2653 \isabellestyle command is given, output is according to slanted
  2654 type-writer);
  2655 
  2656 * support sub/super scripts (for single symbols only), input syntax is
  2657 like this: "A\<^sup>*" or "A\<^sup>\<star>";
  2658 
  2659 * some more standard symbols; see Appendix A of the system manual for
  2660 the complete list of symbols defined in isabellesym.sty;
  2661 
  2662 * improved isabelle style files; more abstract symbol implementation
  2663 (should now use \isamath{...} and \isatext{...} in custom symbol
  2664 definitions);
  2665 
  2666 * antiquotation @{goals} and @{subgoals} for output of *dynamic* goals
  2667 state; Note that presentation of goal states does not conform to
  2668 actual human-readable proof documents.  Please do not include goal
  2669 states into document output unless you really know what you are doing!
  2670 
  2671 * proper indentation of antiquoted output with proportional LaTeX
  2672 fonts;
  2673 
  2674 * no_document ML operator temporarily disables LaTeX document
  2675 generation;
  2676 
  2677 * isatool unsymbolize tunes sources for plain ASCII communication;
  2678 
  2679 
  2680 *** Isar ***
  2681 
  2682 * Pure: Isar now suffers initial goal statements to contain unbound
  2683 schematic variables (this does not conform to actual readable proof
  2684 documents, due to unpredictable outcome and non-compositional proof
  2685 checking); users who know what they are doing may use schematic goals
  2686 for Prolog-style synthesis of proven results;
  2687 
  2688 * Pure: assumption method (an implicit finishing) now handles actual
  2689 rules as well;
  2690 
  2691 * Pure: improved 'obtain' --- moved to Pure, insert "that" into
  2692 initial goal, declare "that" only as Pure intro (only for single
  2693 steps); the "that" rule assumption may now be involved in implicit
  2694 finishing, thus ".." becomes a feasible for trivial obtains;
  2695 
  2696 * Pure: default proof step now includes 'intro_classes'; thus trivial
  2697 instance proofs may be performed by "..";
  2698 
  2699 * Pure: ?thesis / ?this / "..." now work for pure meta-level
  2700 statements as well;
  2701 
  2702 * Pure: more robust selection of calculational rules;
  2703 
  2704 * Pure: the builtin notion of 'finished' goal now includes the ==-refl
  2705 rule (as well as the assumption rule);
  2706 
  2707 * Pure: 'thm_deps' command visualizes dependencies of theorems and
  2708 lemmas, using the graph browser tool;
  2709 
  2710 * Pure: predict failure of "show" in interactive mode;
  2711 
  2712 * Pure: 'thms_containing' now takes actual terms as arguments;
  2713 
  2714 * HOL: improved method 'induct' --- now handles non-atomic goals
  2715 (potential INCOMPATIBILITY); tuned error handling;
  2716 
  2717 * HOL: cases and induct rules now provide explicit hints about the
  2718 number of facts to be consumed (0 for "type" and 1 for "set" rules);
  2719 any remaining facts are inserted into the goal verbatim;
  2720 
  2721 * HOL: local contexts (aka cases) may now contain term bindings as
  2722 well; the 'cases' and 'induct' methods new provide a ?case binding for
  2723 the result to be shown in each case;
  2724 
  2725 * HOL: added 'recdef_tc' command;
  2726 
  2727 * isatool convert assists in eliminating legacy ML scripts;
  2728 
  2729 
  2730 *** HOL ***
  2731 
  2732 * HOL/Library: a collection of generic theories to be used together
  2733 with main HOL; the theory loader path already includes this directory
  2734 by default; the following existing theories have been moved here:
  2735 HOL/Induct/Multiset, HOL/Induct/Acc (as Accessible_Part), HOL/While
  2736 (as While_Combinator), HOL/Lex/Prefix (as List_Prefix);
  2737 
  2738 * HOL/Unix: "Some aspects of Unix file-system security", a typical
  2739 modelling and verification task performed in Isabelle/HOL +
  2740 Isabelle/Isar + Isabelle document preparation (by Markus Wenzel).
  2741 
  2742 * HOL/Algebra: special summation operator SUM no longer exists, it has
  2743 been replaced by setsum; infix 'assoc' now has priority 50 (like
  2744 'dvd'); axiom 'one_not_zero' has been moved from axclass 'ring' to
  2745 'domain', this makes the theory consistent with mathematical
  2746 literature;
  2747 
  2748 * HOL basics: added overloaded operations "inverse" and "divide"
  2749 (infix "/"), syntax for generic "abs" operation, generic summation
  2750 operator \<Sum>;
  2751 
  2752 * HOL/typedef: simplified package, provide more useful rules (see also
  2753 HOL/subset.thy);
  2754 
  2755 * HOL/datatype: induction rule for arbitrarily branching datatypes is
  2756 now expressed as a proper nested rule (old-style tactic scripts may
  2757 require atomize_strip_tac to cope with non-atomic premises);
  2758 
  2759 * HOL: renamed theory "Prod" to "Product_Type", renamed "split" rule
  2760 to "split_conv" (old name still available for compatibility);
  2761 
  2762 * HOL: improved concrete syntax for strings (e.g. allows translation
  2763 rules with string literals);
  2764 
  2765 * HOL-Real-Hyperreal: this extends HOL-Real with the hyperreals
  2766  and Fleuriot's mechanization of analysis, including the transcendental
  2767  functions for the reals;
  2768 
  2769 * HOL/Real, HOL/Hyperreal: improved arithmetic simplification;
  2770 
  2771 
  2772 *** CTT ***
  2773 
  2774 * CTT: x-symbol support for Pi, Sigma, -->, : (membership); note that
  2775 "lam" is displayed as TWO lambda-symbols
  2776 
  2777 * CTT: theory Main now available, containing everything (that is, Bool
  2778 and Arith);
  2779 
  2780 
  2781 *** General ***
  2782 
  2783 * Pure: the Simplifier has been implemented properly as a derived rule
  2784 outside of the actual kernel (at last!); the overall performance
  2785 penalty in practical applications is about 50%, while reliability of
  2786 the Isabelle inference kernel has been greatly improved;
  2787 
  2788 * print modes "brackets" and "no_brackets" control output of nested =>
  2789 (types) and ==> (props); the default behaviour is "brackets";
  2790 
  2791 * Provers: fast_tac (and friends) now handle actual object-logic rules
  2792 as assumptions as well;
  2793 
  2794 * system: support Poly/ML 4.0;
  2795 
  2796 * system: isatool install handles KDE version 1 or 2;
  2797 
  2798 
  2799 
  2800 New in Isabelle99-1 (October 2000)
  2801 ----------------------------------
  2802 
  2803 *** Overview of INCOMPATIBILITIES ***
  2804 
  2805 * HOL: simplification of natural numbers is much changed; to partly
  2806 recover the old behaviour (e.g. to prevent n+n rewriting to #2*n)
  2807 issue the following ML commands:
  2808 
  2809   Delsimprocs Nat_Numeral_Simprocs.cancel_numerals;
  2810   Delsimprocs [Nat_Numeral_Simprocs.combine_numerals];
  2811 
  2812 * HOL: simplification no longer dives into case-expressions; this is
  2813 controlled by "t.weak_case_cong" for each datatype t;
  2814 
  2815 * HOL: nat_less_induct renamed to less_induct;
  2816 
  2817 * HOL: systematic renaming of the SOME (Eps) rules, may use isatool
  2818 fixsome to patch .thy and .ML sources automatically;
  2819 
  2820   select_equality  -> some_equality
  2821   select_eq_Ex     -> some_eq_ex
  2822   selectI2EX       -> someI2_ex
  2823   selectI2         -> someI2
  2824   selectI          -> someI
  2825   select1_equality -> some1_equality
  2826   Eps_sym_eq       -> some_sym_eq_trivial
  2827   Eps_eq           -> some_eq_trivial
  2828 
  2829 * HOL: exhaust_tac on datatypes superceded by new generic case_tac;
  2830 
  2831 * HOL: removed obsolete theorem binding expand_if (refer to split_if
  2832 instead);
  2833 
  2834 * HOL: the recursion equations generated by 'recdef' are now called
  2835 f.simps instead of f.rules;
  2836 
  2837 * HOL: qed_spec_mp now also handles bounded ALL as well;
  2838 
  2839 * HOL: 0 is now overloaded, so the type constraint ":: nat" may
  2840 sometimes be needed;
  2841 
  2842 * HOL: the constant for "f``x" is now "image" rather than "op ``";
  2843 
  2844 * HOL: the constant for "f-``x" is now "vimage" rather than "op -``";
  2845 
  2846 * HOL: the disjoint sum is now "<+>" instead of "Plus"; the cartesian
  2847 product is now "<*>" instead of "Times"; the lexicographic product is
  2848 now "<*lex*>" instead of "**";
  2849 
  2850 * HOL: theory Sexp is now in HOL/Induct examples (it used to be part
  2851 of main HOL, but was unused); better use HOL's datatype package;
  2852 
  2853 * HOL: removed "symbols" syntax for constant "override" of theory Map;
  2854 the old syntax may be recovered as follows:
  2855 
  2856   syntax (symbols)
  2857     override  :: "('a ~=> 'b) => ('a ~=> 'b) => ('a ~=> 'b)"
  2858       (infixl "\\<oplus>" 100)
  2859 
  2860 * HOL/Real: "rabs" replaced by overloaded "abs" function;
  2861 
  2862 * HOL/ML: even fewer consts are declared as global (see theories Ord,
  2863 Lfp, Gfp, WF); this only affects ML packages that refer to const names
  2864 internally;
  2865 
  2866 * HOL and ZF: syntax for quotienting wrt an equivalence relation
  2867 changed from A/r to A//r;
  2868 
  2869 * ZF: new treatment of arithmetic (nat & int) may break some old
  2870 proofs;
  2871 
  2872 * Isar: renamed some attributes (RS -> THEN, simplify -> simplified,
  2873 rulify -> rule_format, elimify -> elim_format, ...);
  2874 
  2875 * Isar/Provers: intro/elim/dest attributes changed; renamed
  2876 intro/intro!/intro!! flags to intro!/intro/intro? (in most cases, one
  2877 should have to change intro!! to intro? only); replaced "delrule" by
  2878 "rule del";
  2879 
  2880 * Isar/HOL: renamed "intrs" to "intros" in inductive definitions;
  2881 
  2882 * Provers: strengthened force_tac by using new first_best_tac;
  2883 
  2884 * LaTeX document preparation: several changes of isabelle.sty (see
  2885 lib/texinputs);
  2886 
  2887 
  2888 *** Document preparation ***
  2889 
  2890 * formal comments (text blocks etc.) in new-style theories may now
  2891 contain antiquotations of thm/prop/term/typ/text to be presented
  2892 according to latex print mode; concrete syntax is like this:
  2893 @{term[show_types] "f(x) = a + x"};
  2894 
  2895 * isatool mkdir provides easy setup of Isabelle session directories,
  2896 including proper document sources;
  2897 
  2898 * generated LaTeX sources are now deleted after successful run
  2899 (isatool document -c); may retain a copy somewhere else via -D option
  2900 of isatool usedir;
  2901 
  2902 * isatool usedir -D now lets isatool latex -o sty update the Isabelle
  2903 style files, achieving self-contained LaTeX sources and simplifying
  2904 LaTeX debugging;
  2905 
  2906 * old-style theories now produce (crude) LaTeX output as well;
  2907 
  2908 * browser info session directories are now self-contained (may be put
  2909 on WWW server seperately); improved graphs of nested sessions; removed
  2910 graph for 'all sessions';
  2911 
  2912 * several improvements in isabelle style files; \isabellestyle{it}
  2913 produces fake math mode output; \isamarkupheader is now \section by
  2914 default; see lib/texinputs/isabelle.sty etc.;
  2915 
  2916 
  2917 *** Isar ***
  2918 
  2919 * Isar/Pure: local results and corresponding term bindings are now
  2920 subject to Hindley-Milner polymorphism (similar to ML); this
  2921 accommodates incremental type-inference very nicely;
  2922 
  2923 * Isar/Pure: new derived language element 'obtain' supports
  2924 generalized existence reasoning;
  2925 
  2926 * Isar/Pure: new calculational elements 'moreover' and 'ultimately'
  2927 support accumulation of results, without applying any rules yet;
  2928 useful to collect intermediate results without explicit name
  2929 references, and for use with transitivity rules with more than 2
  2930 premises;
  2931 
  2932 * Isar/Pure: scalable support for case-analysis type proofs: new
  2933 'case' language element refers to local contexts symbolically, as
  2934 produced by certain proof methods; internally, case names are attached
  2935 to theorems as "tags";
  2936 
  2937 * Isar/Pure: theory command 'hide' removes declarations from
  2938 class/type/const name spaces;
  2939 
  2940 * Isar/Pure: theory command 'defs' supports option "(overloaded)" to
  2941 indicate potential overloading;
  2942 
  2943 * Isar/Pure: changed syntax of local blocks from {{ }} to { };
  2944 
  2945 * Isar/Pure: syntax of sorts made 'inner', i.e. have to write
  2946 "{a,b,c}" instead of {a,b,c};
  2947 
  2948 * Isar/Pure now provides its own version of intro/elim/dest
  2949 attributes; useful for building new logics, but beware of confusion
  2950 with the version in Provers/classical;
  2951 
  2952 * Isar/Pure: the local context of (non-atomic) goals is provided via
  2953 case name 'antecedent';
  2954 
  2955 * Isar/Pure: removed obsolete 'transfer' attribute (transfer of thms
  2956 to the current context is now done automatically);
  2957 
  2958 * Isar/Pure: theory command 'method_setup' provides a simple interface
  2959 for definining proof methods in ML;
  2960 
  2961 * Isar/Provers: intro/elim/dest attributes changed; renamed
  2962 intro/intro!/intro!! flags to intro!/intro/intro? (INCOMPATIBILITY, in
  2963 most cases, one should have to change intro!! to intro? only);
  2964 replaced "delrule" by "rule del";
  2965 
  2966 * Isar/Provers: new 'hypsubst' method, plain 'subst' method and
  2967 'symmetric' attribute (the latter supercedes [RS sym]);
  2968 
  2969 * Isar/Provers: splitter support (via 'split' attribute and 'simp'
  2970 method modifier); 'simp' method: 'only:' modifier removes loopers as
  2971 well (including splits);
  2972 
  2973 * Isar/Provers: Simplifier and Classical methods now support all kind
  2974 of modifiers used in the past, including 'cong', 'iff', etc.
  2975 
  2976 * Isar/Provers: added 'fastsimp' and 'clarsimp' methods (combination
  2977 of Simplifier and Classical reasoner);
  2978 
  2979 * Isar/HOL: new proof method 'cases' and improved version of 'induct'
  2980 now support named cases; major packages (inductive, datatype, primrec,
  2981 recdef) support case names and properly name parameters;
  2982 
  2983 * Isar/HOL: new transitivity rules for substitution in inequalities --
  2984 monotonicity conditions are extracted to be proven at end of
  2985 calculations;
  2986 
  2987 * Isar/HOL: removed 'case_split' thm binding, should use 'cases' proof
  2988 method anyway;
  2989 
  2990 * Isar/HOL: removed old expand_if = split_if; theorems if_splits =
  2991 split_if split_if_asm; datatype package provides theorems foo.splits =
  2992 foo.split foo.split_asm for each datatype;
  2993 
  2994 * Isar/HOL: tuned inductive package, rename "intrs" to "intros"
  2995 (potential INCOMPATIBILITY), emulation of mk_cases feature for proof
  2996 scripts: new 'inductive_cases' command and 'ind_cases' method; (Note:
  2997 use "(cases (simplified))" method in proper proof texts);
  2998 
  2999 * Isar/HOL: added global 'arith_split' attribute for 'arith' method;
  3000 
  3001 * Isar: names of theorems etc. may be natural numbers as well;
  3002 
  3003 * Isar: 'pr' command: optional arguments for goals_limit and
  3004 ProofContext.prems_limit; no longer prints theory contexts, but only
  3005 proof states;
  3006 
  3007 * Isar: diagnostic commands 'pr', 'thm', 'prop', 'term', 'typ' admit
  3008 additional print modes to be specified; e.g. "pr(latex)" will print
  3009 proof state according to the Isabelle LaTeX style;
  3010 
  3011 * Isar: improved support for emulating tactic scripts, including proof
  3012 methods 'rule_tac' etc., 'cut_tac', 'thin_tac', 'subgoal_tac',
  3013 'rename_tac', 'rotate_tac', 'tactic', and 'case_tac' / 'induct_tac'
  3014 (for HOL datatypes);
  3015 
  3016 * Isar: simplified (more robust) goal selection of proof methods: 1st
  3017 goal, all goals, or explicit goal specifier (tactic emulation); thus
  3018 'proof method scripts' have to be in depth-first order;
  3019 
  3020 * Isar: tuned 'let' syntax: replaced 'as' keyword by 'and';
  3021 
  3022 * Isar: removed 'help' command, which hasn't been too helpful anyway;
  3023 should instead use individual commands for printing items
  3024 (print_commands, print_methods etc.);
  3025 
  3026 * Isar: added 'nothing' --- the empty list of theorems;
  3027 
  3028 
  3029 *** HOL ***
  3030 
  3031 * HOL/MicroJava: formalization of a fragment of Java, together with a
  3032 corresponding virtual machine and a specification of its bytecode
  3033 verifier and a lightweight bytecode verifier, including proofs of
  3034 type-safety; by Gerwin Klein, Tobias Nipkow, David von Oheimb, and
  3035 Cornelia Pusch (see also the homepage of project Bali at
  3036 http://isabelle.in.tum.de/Bali/);
  3037 
  3038 * HOL/Algebra: new theory of rings and univariate polynomials, by
  3039 Clemens Ballarin;
  3040 
  3041 * HOL/NumberTheory: fundamental Theorem of Arithmetic, Chinese
  3042 Remainder Theorem, Fermat/Euler Theorem, Wilson's Theorem, by Thomas M
  3043 Rasmussen;
  3044 
  3045 * HOL/Lattice: fundamental concepts of lattice theory and order
  3046 structures, including duals, properties of bounds versus algebraic
  3047 laws, lattice operations versus set-theoretic ones, the Knaster-Tarski
  3048 Theorem for complete lattices etc.; may also serve as a demonstration
  3049 for abstract algebraic reasoning using axiomatic type classes, and
  3050 mathematics-style proof in Isabelle/Isar; by Markus Wenzel;
  3051 
  3052 * HOL/Prolog: a (bare-bones) implementation of Lambda-Prolog, by David
  3053 von Oheimb;
  3054 
  3055 * HOL/IMPP: extension of IMP with local variables and mutually
  3056 recursive procedures, by David von Oheimb;
  3057 
  3058 * HOL/Lambda: converted into new-style theory and document;
  3059 
  3060 * HOL/ex/Multiquote: example of multiple nested quotations and
  3061 anti-quotations -- basically a generalized version of de-Bruijn
  3062 representation; very useful in avoiding lifting of operations;
  3063 
  3064 * HOL/record: added general record equality rule to simpset; fixed
  3065 select-update simplification procedure to handle extended records as
  3066 well; admit "r" as field name;
  3067 
  3068 * HOL: 0 is now overloaded over the new sort "zero", allowing its use with
  3069 other numeric types and also as the identity of groups, rings, etc.;
  3070 
  3071 * HOL: new axclass plus_ac0 for addition with the AC-laws and 0 as identity.
  3072 Types nat and int belong to this axclass;
  3073 
  3074 * HOL: greatly improved simplification involving numerals of type nat, int, real:
  3075    (i + #8 + j) = Suc k simplifies to  #7 + (i + j) = k
  3076    i*j + k + j*#3*i     simplifies to  #4*(i*j) + k
  3077   two terms #m*u and #n*u are replaced by #(m+n)*u
  3078     (where #m, #n and u can implicitly be 1; this is simproc combine_numerals)
  3079   and the term/formula #m*u+x ~~ #n*u+y simplifies simplifies to #(m-n)+x ~~ y
  3080     or x ~~ #(n-m)+y, where ~~ is one of = < <= or - (simproc cancel_numerals);
  3081 
  3082 * HOL: meson_tac is available (previously in ex/meson.ML); it is a
  3083 powerful prover for predicate logic but knows nothing of clasets; see
  3084 ex/mesontest.ML and ex/mesontest2.ML for example applications;
  3085 
  3086 * HOL: new version of "case_tac" subsumes both boolean case split and
  3087 "exhaust_tac" on datatypes; INCOMPATIBILITY: exhaust_tac no longer
  3088 exists, may define val exhaust_tac = case_tac for ad-hoc portability;
  3089 
  3090 * HOL: simplification no longer dives into case-expressions: only the
  3091 selector expression is simplified, but not the remaining arms; to
  3092 enable full simplification of case-expressions for datatype t, you may
  3093 remove t.weak_case_cong from the simpset, either globally (Delcongs
  3094 [thm"t.weak_case_cong"];) or locally (delcongs [...]).
  3095 
  3096 * HOL/recdef: the recursion equations generated by 'recdef' for
  3097 function 'f' are now called f.simps instead of f.rules; if all
  3098 termination conditions are proved automatically, these simplification
  3099 rules are added to the simpset, as in primrec; rules may be named
  3100 individually as well, resulting in a separate list of theorems for
  3101 each equation;
  3102 
  3103 * HOL/While is a new theory that provides a while-combinator. It
  3104 permits the definition of tail-recursive functions without the
  3105 provision of a termination measure. The latter is necessary once the
  3106 invariant proof rule for while is applied.
  3107 
  3108 * HOL: new (overloaded) notation for the set of elements below/above
  3109 some element: {..u}, {..u(}, {l..}, {)l..}. See theory SetInterval.
  3110 
  3111 * HOL: theorems impI, allI, ballI bound as "strip";
  3112 
  3113 * HOL: new tactic induct_thm_tac: thm -> string -> int -> tactic
  3114 induct_tac th "x1 ... xn" expects th to have a conclusion of the form
  3115 P v1 ... vn and abbreviates res_inst_tac [("v1","x1"),...,("vn","xn")] th;
  3116 
  3117 * HOL/Real: "rabs" replaced by overloaded "abs" function;
  3118 
  3119 * HOL: theory Sexp now in HOL/Induct examples (it used to be part of
  3120 main HOL, but was unused);
  3121 
  3122 * HOL: fewer consts declared as global (e.g. have to refer to
  3123 "Lfp.lfp" instead of "lfp" internally; affects ML packages only);
  3124 
  3125 * HOL: tuned AST representation of nested pairs, avoiding bogus output
  3126 in case of overlap with user translations (e.g. judgements over
  3127 tuples); (note that the underlying logical represenation is still
  3128 bogus);
  3129 
  3130 
  3131 *** ZF ***
  3132 
  3133 * ZF: simplification automatically cancels common terms in arithmetic
  3134 expressions over nat and int;
  3135 
  3136 * ZF: new treatment of nat to minimize type-checking: all operators
  3137 coerce their operands to a natural number using the function natify,
  3138 making the algebraic laws unconditional;
  3139 
  3140 * ZF: as above, for int: operators coerce their operands to an integer
  3141 using the function intify;
  3142 
  3143 * ZF: the integer library now contains many of the usual laws for the
  3144 orderings, including $<=, and monotonicity laws for $+ and $*;
  3145 
  3146 * ZF: new example ZF/ex/NatSum to demonstrate integer arithmetic
  3147 simplification;
  3148 
  3149 * FOL and ZF: AddIffs now available, giving theorems of the form P<->Q
  3150 to the simplifier and classical reasoner simultaneously;
  3151 
  3152 
  3153 *** General ***
  3154 
  3155 * Provers: blast_tac now handles actual object-logic rules as
  3156 assumptions; note that auto_tac uses blast_tac internally as well;
  3157 
  3158 * Provers: new functions rulify/rulify_no_asm: thm -> thm for turning
  3159 outer -->/All/Ball into ==>/!!; qed_spec_mp now uses rulify_no_asm;
  3160 
  3161 * Provers: delrules now handles destruct rules as well (no longer need
  3162 explicit make_elim);
  3163 
  3164 * Provers: Blast_tac now warns of and ignores "weak elimination rules" e.g.
  3165   [| inj ?f;          ?f ?x = ?f ?y; ?x = ?y ==> ?W |] ==> ?W
  3166 use instead the strong form,
  3167   [| inj ?f; ~ ?W ==> ?f ?x = ?f ?y; ?x = ?y ==> ?W |] ==> ?W
  3168 in HOL, FOL and ZF the function cla_make_elim will create such rules
  3169 from destruct-rules;
  3170 
  3171 * Provers: Simplifier.easy_setup provides a fast path to basic
  3172 Simplifier setup for new object-logics;
  3173 
  3174 * Pure: AST translation rules no longer require constant head on LHS;
  3175 
  3176 * Pure: improved name spaces: ambiguous output is qualified; support
  3177 for hiding of names;
  3178 
  3179 * system: smart setup of canonical ML_HOME, ISABELLE_INTERFACE, and
  3180 XSYMBOL_HOME; no longer need to do manual configuration in most
  3181 situations;
  3182 
  3183 * system: compression of ML heaps images may now be controlled via -c
  3184 option of isabelle and isatool usedir (currently only observed by
  3185 Poly/ML);
  3186 
  3187 * system: isatool installfonts may handle X-Symbol fonts as well (very
  3188 useful for remote X11);
  3189 
  3190 * system: provide TAGS file for Isabelle sources;
  3191 
  3192 * ML: infix 'OF' is a version of 'MRS' with more appropriate argument
  3193 order;
  3194 
  3195 * ML: renamed flags Syntax.trace_norm_ast to Syntax.trace_ast; global
  3196 timing flag supersedes proof_timing and Toplevel.trace;
  3197 
  3198 * ML: new combinators |>> and |>>> for incremental transformations
  3199 with secondary results (e.g. certain theory extensions):
  3200 
  3201 * ML: PureThy.add_defs gets additional argument to indicate potential
  3202 overloading (usually false);
  3203 
  3204 * ML: PureThy.add_thms/add_axioms/add_defs now return theorems as
  3205 results;
  3206 
  3207 
  3208 
  3209 New in Isabelle99 (October 1999)
  3210 --------------------------------
  3211 
  3212 *** Overview of INCOMPATIBILITIES (see below for more details) ***
  3213 
  3214 * HOL: The THEN and ELSE parts of conditional expressions (if P then x else y)
  3215 are no longer simplified.  (This allows the simplifier to unfold recursive
  3216 functional programs.)  To restore the old behaviour, declare
  3217 
  3218     Delcongs [if_weak_cong];
  3219 
  3220 * HOL: Removed the obsolete syntax "Compl A"; use -A for set
  3221 complement;
  3222 
  3223 * HOL: the predicate "inj" is now defined by translation to "inj_on";
  3224 
  3225 * HOL/datatype: mutual_induct_tac no longer exists --
  3226   use induct_tac "x_1 ... x_n" instead of mutual_induct_tac ["x_1", ..., "x_n"]
  3227 
  3228 * HOL/typedef: fixed type inference for representing set; type
  3229 arguments now have to occur explicitly on the rhs as type constraints;
  3230 
  3231 * ZF: The con_defs part of an inductive definition may no longer refer
  3232 to constants declared in the same theory;
  3233 
  3234 * HOL, ZF: the function mk_cases, generated by the inductive
  3235 definition package, has lost an argument.  To simplify its result, it
  3236 uses the default simpset instead of a supplied list of theorems.
  3237 
  3238 * HOL/List: the constructors of type list are now Nil and Cons;
  3239 
  3240 * Simplifier: the type of the infix ML functions
  3241         setSSolver addSSolver setSolver addSolver
  3242 is now  simpset * solver -> simpset  where `solver' is a new abstract type
  3243 for packaging solvers. A solver is created via
  3244         mk_solver: string -> (thm list -> int -> tactic) -> solver
  3245 where the string argument is only a comment.
  3246 
  3247 
  3248 *** Proof tools ***
  3249 
  3250 * Provers/Arith/fast_lin_arith.ML contains a functor for creating a
  3251 decision procedure for linear arithmetic. Currently it is used for
  3252 types `nat', `int', and `real' in HOL (see below); it can, should and
  3253 will be instantiated for other types and logics as well.
  3254 
  3255 * The simplifier now accepts rewrite rules with flexible heads, eg
  3256      hom ?f ==> ?f(?x+?y) = ?f ?x + ?f ?y
  3257   They are applied like any rule with a non-pattern lhs, i.e. by first-order
  3258   matching.
  3259 
  3260 
  3261 *** General ***
  3262 
  3263 * New Isabelle/Isar subsystem provides an alternative to traditional
  3264 tactical theorem proving; together with the ProofGeneral/isar user
  3265 interface it offers an interactive environment for developing human
  3266 readable proof documents (Isar == Intelligible semi-automated
  3267 reasoning); for further information see isatool doc isar-ref,
  3268 src/HOL/Isar_examples and http://isabelle.in.tum.de/Isar/
  3269 
  3270 * improved and simplified presentation of theories: better HTML markup
  3271 (including colors), graph views in several sizes; isatool usedir now
  3272 provides a proper interface for user theories (via -P option); actual
  3273 document preparation based on (PDF)LaTeX is available as well (for
  3274 new-style theories only); see isatool doc system for more information;
  3275 
  3276 * native support for Proof General, both for classic Isabelle and
  3277 Isabelle/Isar;
  3278 
  3279 * ML function thm_deps visualizes dependencies of theorems and lemmas,
  3280 using the graph browser tool;
  3281 
  3282 * Isabelle manuals now also available as PDF;
  3283 
  3284 * theory loader rewritten from scratch (may not be fully
  3285 bug-compatible); old loadpath variable has been replaced by show_path,
  3286 add_path, del_path, reset_path functions; new operations such as
  3287 update_thy, touch_thy, remove_thy, use/update_thy_only (see also
  3288 isatool doc ref);
  3289 
  3290 * improved isatool install: option -k creates KDE application icon,
  3291 option -p DIR installs standalone binaries;
  3292 
  3293 * added ML_PLATFORM setting (useful for cross-platform installations);
  3294 more robust handling of platform specific ML images for SML/NJ;
  3295 
  3296 * the settings environment is now statically scoped, i.e. it is never
  3297 created again in sub-processes invoked from isabelle, isatool, or
  3298 Isabelle;
  3299 
  3300 * path element specification '~~' refers to '$ISABELLE_HOME';
  3301 
  3302 * in locales, the "assumes" and "defines" parts may be omitted if
  3303 empty;
  3304 
  3305 * new print_mode "xsymbols" for extended symbol support (e.g. genuine
  3306 long arrows);
  3307 
  3308 * new print_mode "HTML";
  3309 
  3310 * new flag show_tags controls display of tags of theorems (which are
  3311 basically just comments that may be attached by some tools);
  3312 
  3313 * Isamode 2.6 requires patch to accomodate change of Isabelle font
  3314 mode and goal output format:
  3315 
  3316 diff -r Isamode-2.6/elisp/isa-load.el Isamode/elisp/isa-load.el
  3317 244c244
  3318 <       (list (isa-getenv "ISABELLE") "-msymbols" logic-name)
  3319 ---
  3320 >       (list (isa-getenv "ISABELLE") "-misabelle_font" "-msymbols" logic-name)
  3321 diff -r Isabelle-2.6/elisp/isa-proofstate.el Isamode/elisp/isa-proofstate.el
  3322 181c181
  3323 < (defconst proofstate-proofstart-regexp "^Level [0-9]+$"
  3324 ---
  3325 > (defconst proofstate-proofstart-regexp "^Level [0-9]+"
  3326 
  3327 * function bind_thms stores lists of theorems (cf. bind_thm);
  3328 
  3329 * new shorthand tactics ftac, eatac, datac, fatac;
  3330 
  3331 * qed (and friends) now accept "" as result name; in that case the
  3332 theorem is not stored, but proper checks and presentation of the
  3333 result still apply;
  3334 
  3335 * theorem database now also indexes constants "Trueprop", "all",
  3336 "==>", "=="; thus thms_containing, findI etc. may retrieve more rules;
  3337 
  3338 
  3339 *** HOL ***
  3340 
  3341 ** HOL arithmetic **
  3342 
  3343 * There are now decision procedures for linear arithmetic over nat and
  3344 int:
  3345 
  3346 1. arith_tac copes with arbitrary formulae involving `=', `<', `<=',
  3347 `+', `-', `Suc', `min', `max' and numerical constants; other subterms
  3348 are treated as atomic; subformulae not involving type `nat' or `int'
  3349 are ignored; quantified subformulae are ignored unless they are
  3350 positive universal or negative existential. The tactic has to be
  3351 invoked by hand and can be a little bit slow. In particular, the
  3352 running time is exponential in the number of occurrences of `min' and
  3353 `max', and `-' on `nat'.
  3354 
  3355 2. fast_arith_tac is a cut-down version of arith_tac: it only takes
  3356 (negated) (in)equalities among the premises and the conclusion into
  3357 account (i.e. no compound formulae) and does not know about `min' and
  3358 `max', and `-' on `nat'. It is fast and is used automatically by the
  3359 simplifier.
  3360 
  3361 NB: At the moment, these decision procedures do not cope with mixed
  3362 nat/int formulae where the two parts interact, such as `m < n ==>
  3363 int(m) < int(n)'.
  3364 
  3365 * HOL/Numeral provides a generic theory of numerals (encoded
  3366 efficiently as bit strings); setup for types nat/int/real is in place;
  3367 INCOMPATIBILITY: since numeral syntax is now polymorphic, rather than
  3368 int, existing theories and proof scripts may require a few additional
  3369 type constraints;
  3370 
  3371 * integer division and remainder can now be performed on constant
  3372 arguments;
  3373 
  3374 * many properties of integer multiplication, division and remainder
  3375 are now available;
  3376 
  3377 * An interface to the Stanford Validity Checker (SVC) is available through the
  3378 tactic svc_tac.  Propositional tautologies and theorems of linear arithmetic
  3379 are proved automatically.  SVC must be installed separately, and its results
  3380 must be TAKEN ON TRUST (Isabelle does not check the proofs, but tags any
  3381 invocation of the underlying oracle).  For SVC see
  3382   http://verify.stanford.edu/SVC
  3383 
  3384 * IsaMakefile: the HOL-Real target now builds an actual image;
  3385 
  3386 
  3387 ** HOL misc **
  3388 
  3389 * HOL/Real/HahnBanach: the Hahn-Banach theorem for real vector spaces
  3390 (in Isabelle/Isar) -- by Gertrud Bauer;
  3391 
  3392 * HOL/BCV: generic model of bytecode verification, i.e. data-flow
  3393 analysis for assembly languages with subtypes;
  3394 
  3395 * HOL/TLA (Lamport's Temporal Logic of Actions): major reorganization
  3396 -- avoids syntactic ambiguities and treats state, transition, and
  3397 temporal levels more uniformly; introduces INCOMPATIBILITIES due to
  3398 changed syntax and (many) tactics;
  3399 
  3400 * HOL/inductive: Now also handles more general introduction rules such
  3401   as "ALL y. (y, x) : r --> y : acc r ==> x : acc r"; monotonicity
  3402   theorems are now maintained within the theory (maintained via the
  3403   "mono" attribute);
  3404 
  3405 * HOL/datatype: Now also handles arbitrarily branching datatypes
  3406   (using function types) such as
  3407 
  3408   datatype 'a tree = Atom 'a | Branch "nat => 'a tree"
  3409 
  3410 * HOL/record: record_simproc (part of the default simpset) takes care
  3411 of selectors applied to updated records; record_split_tac is no longer
  3412 part of the default claset; update_defs may now be removed from the
  3413 simpset in many cases; COMPATIBILITY: old behavior achieved by
  3414 
  3415   claset_ref () := claset() addSWrapper record_split_wrapper;
  3416   Delsimprocs [record_simproc]
  3417 
  3418 * HOL/typedef: fixed type inference for representing set; type
  3419 arguments now have to occur explicitly on the rhs as type constraints;
  3420 
  3421 * HOL/recdef (TFL): 'congs' syntax now expects comma separated list of theorem
  3422 names rather than an ML expression;
  3423 
  3424 * HOL/defer_recdef (TFL): like recdef but the well-founded relation can be
  3425 supplied later.  Program schemes can be defined, such as
  3426     "While B C s = (if B s then While B C (C s) else s)"
  3427 where the well-founded relation can be chosen after B and C have been given.
  3428 
  3429 * HOL/List: the constructors of type list are now Nil and Cons;
  3430 INCOMPATIBILITY: while [] and infix # syntax is still there, of
  3431 course, ML tools referring to List.list.op # etc. have to be adapted;
  3432 
  3433 * HOL_quantifiers flag superseded by "HOL" print mode, which is
  3434 disabled by default; run isabelle with option -m HOL to get back to
  3435 the original Gordon/HOL-style output;
  3436 
  3437 * HOL/Ord.thy: new bounded quantifier syntax (input only): ALL x<y. P,
  3438 ALL x<=y. P, EX x<y. P, EX x<=y. P;
  3439 
  3440 * HOL basic syntax simplified (more orthogonal): all variants of
  3441 All/Ex now support plain / symbolic / HOL notation; plain syntax for
  3442 Eps operator is provided as well: "SOME x. P[x]";
  3443 
  3444 * HOL/Sum.thy: sum_case has been moved to HOL/Datatype;
  3445 
  3446 * HOL/Univ.thy: infix syntax <*>, <+>, <**>, <+> eliminated and made
  3447 thus available for user theories;
  3448 
  3449 * HOLCF/IOA/Sequents: renamed 'Cons' to 'Consq' to avoid clash with
  3450 HOL/List; hardly an INCOMPATIBILITY since '>>' syntax is used all the
  3451 time;
  3452 
  3453 * HOL: new tactic smp_tac: int -> int -> tactic, which applies spec
  3454 several times and then mp;
  3455 
  3456 
  3457 *** LK ***
  3458 
  3459 * the notation <<...>> is now available as a notation for sequences of
  3460 formulas;
  3461 
  3462 * the simplifier is now installed
  3463 
  3464 * the axiom system has been generalized (thanks to Soren Heilmann)
  3465 
  3466 * the classical reasoner now has a default rule database
  3467 
  3468 
  3469 *** ZF ***
  3470 
  3471 * new primrec section allows primitive recursive functions to be given
  3472 directly (as in HOL) over datatypes and the natural numbers;
  3473 
  3474 * new tactics induct_tac and exhaust_tac for induction (or case
  3475 analysis) over datatypes and the natural numbers;
  3476 
  3477 * the datatype declaration of type T now defines the recursor T_rec;
  3478 
  3479 * simplification automatically does freeness reasoning for datatype
  3480 constructors;
  3481 
  3482 * automatic type-inference, with AddTCs command to insert new
  3483 type-checking rules;
  3484 
  3485 * datatype introduction rules are now added as Safe Introduction rules
  3486 to the claset;
  3487 
  3488 * the syntax "if P then x else y" is now available in addition to
  3489 if(P,x,y);
  3490 
  3491 
  3492 *** Internal programming interfaces ***
  3493 
  3494 * tuned simplifier trace output; new flag debug_simp;
  3495 
  3496 * structures Vartab / Termtab (instances of TableFun) offer efficient
  3497 tables indexed by indexname_ord / term_ord (compatible with aconv);
  3498 
  3499 * AxClass.axclass_tac lost the theory argument;
  3500 
  3501 * tuned current_goals_markers semantics: begin / end goal avoids
  3502 printing empty lines;
  3503 
  3504 * removed prs and prs_fn hook, which was broken because it did not
  3505 include \n in its semantics, forcing writeln to add one
  3506 uncoditionally; replaced prs_fn by writeln_fn; consider std_output:
  3507 string -> unit if you really want to output text without newline;
  3508 
  3509 * Symbol.output subject to print mode; INCOMPATIBILITY: defaults to
  3510 plain output, interface builders may have to enable 'isabelle_font'
  3511 mode to get Isabelle font glyphs as before;
  3512 
  3513 * refined token_translation interface; INCOMPATIBILITY: output length
  3514 now of type real instead of int;
  3515 
  3516 * theory loader actions may be traced via new ThyInfo.add_hook
  3517 interface (see src/Pure/Thy/thy_info.ML); example application: keep
  3518 your own database of information attached to *whole* theories -- as
  3519 opposed to intra-theory data slots offered via TheoryDataFun;
  3520 
  3521 * proper handling of dangling sort hypotheses (at last!);
  3522 Thm.strip_shyps and Drule.strip_shyps_warning take care of removing
  3523 extra sort hypotheses that can be witnessed from the type signature;
  3524 the force_strip_shyps flag is gone, any remaining shyps are simply
  3525 left in the theorem (with a warning issued by strip_shyps_warning);
  3526 
  3527 
  3528 
  3529 New in Isabelle98-1 (October 1998)
  3530 ----------------------------------
  3531 
  3532 *** Overview of INCOMPATIBILITIES (see below for more details) ***
  3533 
  3534 * several changes of automated proof tools;
  3535 
  3536 * HOL: major changes to the inductive and datatype packages, including
  3537 some minor incompatibilities of theory syntax;
  3538 
  3539 * HOL: renamed r^-1 to 'converse' from 'inverse'; 'inj_onto' is now
  3540 called `inj_on';
  3541 
  3542 * HOL: removed duplicate thms in Arith:
  3543   less_imp_add_less  should be replaced by  trans_less_add1
  3544   le_imp_add_le      should be replaced by  trans_le_add1
  3545 
  3546 * HOL: unary minus is now overloaded (new type constraints may be
  3547 required);
  3548 
  3549 * HOL and ZF: unary minus for integers is now #- instead of #~.  In
  3550 ZF, expressions such as n#-1 must be changed to n#- 1, since #-1 is
  3551 now taken as an integer constant.
  3552 
  3553 * Pure: ML function 'theory_of' renamed to 'theory';
  3554 
  3555 
  3556 *** Proof tools ***
  3557 
  3558 * Simplifier:
  3559   1. Asm_full_simp_tac is now more aggressive.
  3560      1. It will sometimes reorient premises if that increases their power to
  3561         simplify.
  3562      2. It does no longer proceed strictly from left to right but may also
  3563         rotate premises to achieve further simplification.
  3564      For compatibility reasons there is now Asm_lr_simp_tac which is like the
  3565      old Asm_full_simp_tac in that it does not rotate premises.
  3566   2. The simplifier now knows a little bit about nat-arithmetic.
  3567 
  3568 * Classical reasoner: wrapper mechanism for the classical reasoner now
  3569 allows for selected deletion of wrappers, by introduction of names for
  3570 wrapper functionals.  This implies that addbefore, addSbefore,
  3571 addaltern, and addSaltern now take a pair (name, tactic) as argument,
  3572 and that adding two tactics with the same name overwrites the first
  3573 one (emitting a warning).
  3574   type wrapper = (int -> tactic) -> (int -> tactic)
  3575   setWrapper, setSWrapper, compWrapper and compSWrapper are replaced by
  3576   addWrapper, addSWrapper: claset * (string * wrapper) -> claset
  3577   delWrapper, delSWrapper: claset *  string            -> claset
  3578   getWrapper is renamed to appWrappers, getSWrapper to appSWrappers;
  3579 
  3580 * Classical reasoner: addbefore/addSbefore now have APPEND/ORELSE
  3581 semantics; addbefore now affects only the unsafe part of step_tac
  3582 etc.; this affects addss/auto_tac/force_tac, so EXISTING PROOFS MAY
  3583 FAIL, but proofs should be fixable easily, e.g. by replacing Auto_tac
  3584 by Force_tac;
  3585 
  3586 * Classical reasoner: setwrapper to setWrapper and compwrapper to
  3587 compWrapper; added safe wrapper (and access functions for it);
  3588 
  3589 * HOL/split_all_tac is now much faster and fails if there is nothing
  3590 to split.  Some EXISTING PROOFS MAY REQUIRE ADAPTION because the order
  3591 and the names of the automatically generated variables have changed.
  3592 split_all_tac has moved within claset() from unsafe wrappers to safe
  3593 wrappers, which means that !!-bound variables are split much more
  3594 aggressively, and safe_tac and clarify_tac now split such variables.
  3595 If this splitting is not appropriate, use delSWrapper "split_all_tac".
  3596 Note: the same holds for record_split_tac, which does the job of
  3597 split_all_tac for record fields.
  3598 
  3599 * HOL/Simplifier: Rewrite rules for case distinctions can now be added
  3600 permanently to the default simpset using Addsplits just like
  3601 Addsimps. They can be removed via Delsplits just like
  3602 Delsimps. Lower-case versions are also available.
  3603 
  3604 * HOL/Simplifier: The rule split_if is now part of the default
  3605 simpset. This means that the simplifier will eliminate all occurrences
  3606 of if-then-else in the conclusion of a goal. To prevent this, you can
  3607 either remove split_if completely from the default simpset by
  3608 `Delsplits [split_if]' or remove it in a specific call of the
  3609 simplifier using `... delsplits [split_if]'.  You can also add/delete
  3610 other case splitting rules to/from the default simpset: every datatype
  3611 generates suitable rules `split_t_case' and `split_t_case_asm' (where
  3612 t is the name of the datatype).
  3613 
  3614 * Classical reasoner / Simplifier combination: new force_tac (and
  3615 derivatives Force_tac, force) combines rewriting and classical
  3616 reasoning (and whatever other tools) similarly to auto_tac, but is
  3617 aimed to solve the given subgoal completely.
  3618 
  3619 
  3620 *** General ***
  3621 
  3622 * new top-level commands `Goal' and `Goalw' that improve upon `goal'
  3623 and `goalw': the theory is no longer needed as an explicit argument -
  3624 the current theory context is used; assumptions are no longer returned
  3625 at the ML-level unless one of them starts with ==> or !!; it is
  3626 recommended to convert to these new commands using isatool fixgoal
  3627 (backup your sources first!);
  3628 
  3629 * new top-level commands 'thm' and 'thms' for retrieving theorems from
  3630 the current theory context, and 'theory' to lookup stored theories;
  3631 
  3632 * new theory section 'locale' for declaring constants, assumptions and
  3633 definitions that have local scope;
  3634 
  3635 * new theory section 'nonterminals' for purely syntactic types;
  3636 
  3637 * new theory section 'setup' for generic ML setup functions
  3638 (e.g. package initialization);
  3639 
  3640 * the distribution now includes Isabelle icons: see
  3641 lib/logo/isabelle-{small,tiny}.xpm;
  3642 
  3643 * isatool install - install binaries with absolute references to
  3644 ISABELLE_HOME/bin;
  3645 
  3646 * isatool logo -- create instances of the Isabelle logo (as EPS);
  3647 
  3648 * print mode 'emacs' reserved for Isamode;
  3649 
  3650 * support multiple print (ast) translations per constant name;
  3651 
  3652 * theorems involving oracles are now printed with a suffixed [!];
  3653 
  3654 
  3655 *** HOL ***
  3656 
  3657 * there is now a tutorial on Isabelle/HOL (do 'isatool doc tutorial');
  3658 
  3659 * HOL/inductive package reorganized and improved: now supports mutual
  3660 definitions such as
  3661 
  3662   inductive EVEN ODD
  3663     intrs
  3664       null "0 : EVEN"
  3665       oddI "n : EVEN ==> Suc n : ODD"
  3666       evenI "n : ODD ==> Suc n : EVEN"
  3667 
  3668 new theorem list "elims" contains an elimination rule for each of the
  3669 recursive sets; inductive definitions now handle disjunctive premises
  3670 correctly (also ZF);
  3671 
  3672 INCOMPATIBILITIES: requires Inductive as an ancestor; component
  3673 "mutual_induct" no longer exists - the induction rule is always
  3674 contained in "induct";
  3675 
  3676 
  3677 * HOL/datatype package re-implemented and greatly improved: now
  3678 supports mutually recursive datatypes such as
  3679 
  3680   datatype
  3681     'a aexp = IF_THEN_ELSE ('a bexp) ('a aexp) ('a aexp)
  3682             | SUM ('a aexp) ('a aexp)
  3683             | DIFF ('a aexp) ('a aexp)
  3684             | NUM 'a
  3685   and
  3686     'a bexp = LESS ('a aexp) ('a aexp)
  3687             | AND ('a bexp) ('a bexp)
  3688             | OR ('a bexp) ('a bexp)
  3689 
  3690 as well as indirectly recursive datatypes such as
  3691 
  3692   datatype
  3693     ('a, 'b) term = Var 'a
  3694                   | App 'b ((('a, 'b) term) list)
  3695 
  3696 The new tactic  mutual_induct_tac [<var_1>, ..., <var_n>] i  performs
  3697 induction on mutually / indirectly recursive datatypes.
  3698 
  3699 Primrec equations are now stored in theory and can be accessed via
  3700 <function_name>.simps.
  3701 
  3702 INCOMPATIBILITIES:
  3703 
  3704   - Theories using datatypes must now have theory Datatype as an
  3705     ancestor.
  3706   - The specific <typename>.induct_tac no longer exists - use the
  3707     generic induct_tac instead.
  3708   - natE has been renamed to nat.exhaust - use exhaust_tac
  3709     instead of res_inst_tac ... natE. Note that the variable
  3710     names in nat.exhaust differ from the names in natE, this
  3711     may cause some "fragile" proofs to fail.
  3712   - The theorems split_<typename>_case and split_<typename>_case_asm
  3713     have been renamed to <typename>.split and <typename>.split_asm.
  3714   - Since default sorts of type variables are now handled correctly,
  3715     some datatype definitions may have to be annotated with explicit
  3716     sort constraints.
  3717   - Primrec definitions no longer require function name and type
  3718     of recursive argument.
  3719 
  3720 Consider using isatool fixdatatype to adapt your theories and proof
  3721 scripts to the new package (backup your sources first!).
  3722 
  3723 
  3724 * HOL/record package: considerably improved implementation; now
  3725 includes concrete syntax for record types, terms, updates; theorems
  3726 for surjective pairing and splitting !!-bound record variables; proof
  3727 support is as follows:
  3728 
  3729   1) standard conversions (selectors or updates applied to record
  3730 constructor terms) are part of the standard simpset;
  3731 
  3732   2) inject equations of the form ((x, y) = (x', y')) == x=x' & y=y' are
  3733 made part of standard simpset and claset via addIffs;
  3734 
  3735   3) a tactic for record field splitting (record_split_tac) is part of
  3736 the standard claset (addSWrapper);
  3737 
  3738 To get a better idea about these rules you may retrieve them via
  3739 something like 'thms "foo.simps"' or 'thms "foo.iffs"', where "foo" is
  3740 the name of your record type.
  3741 
  3742 The split tactic 3) conceptually simplifies by the following rule:
  3743 
  3744   "(!!x. PROP ?P x) == (!!a b. PROP ?P (a, b))"
  3745 
  3746 Thus any record variable that is bound by meta-all will automatically
  3747 blow up into some record constructor term, consequently the
  3748 simplifications of 1), 2) apply.  Thus force_tac, auto_tac etc. shall
  3749 solve record problems automatically.
  3750 
  3751 
  3752 * reorganized the main HOL image: HOL/Integ and String loaded by
  3753 default; theory Main includes everything;
  3754 
  3755 * automatic simplification of integer sums and comparisons, using cancellation;
  3756 
  3757 * added option_map_eq_Some and not_Some_eq to the default simpset and claset;
  3758 
  3759 * added disj_not1 = "(~P | Q) = (P --> Q)" to the default simpset;
  3760 
  3761 * many new identities for unions, intersections, set difference, etc.;
  3762 
  3763 * expand_if, expand_split, expand_sum_case and expand_nat_case are now
  3764 called split_if, split_split, split_sum_case and split_nat_case (to go
  3765 with add/delsplits);
  3766 
  3767 * HOL/Prod introduces simplification procedure unit_eq_proc rewriting
  3768 (?x::unit) = (); this is made part of the default simpset, which COULD
  3769 MAKE EXISTING PROOFS FAIL under rare circumstances (consider
  3770 'Delsimprocs [unit_eq_proc];' as last resort); also note that
  3771 unit_abs_eta_conv is added in order to counter the effect of
  3772 unit_eq_proc on (%u::unit. f u), replacing it by f rather than by
  3773 %u.f();
  3774 
  3775 * HOL/Fun INCOMPATIBILITY: `inj_onto' is now called `inj_on' (which
  3776 makes more sense);
  3777 
  3778 * HOL/Set INCOMPATIBILITY: rule `equals0D' is now a well-formed destruct rule;
  3779   It and 'sym RS equals0D' are now in the default  claset, giving automatic
  3780   disjointness reasoning but breaking a few old proofs.
  3781 
  3782 * HOL/Relation INCOMPATIBILITY: renamed the relational operator r^-1
  3783 to 'converse' from 'inverse' (for compatibility with ZF and some
  3784 literature);
  3785 
  3786 * HOL/recdef can now declare non-recursive functions, with {} supplied as
  3787 the well-founded relation;
  3788 
  3789 * HOL/Set INCOMPATIBILITY: the complement of set A is now written -A instead of
  3790     Compl A.  The "Compl" syntax remains available as input syntax for this
  3791     release ONLY.
  3792 
  3793 * HOL/Update: new theory of function updates:
  3794     f(a:=b) == %x. if x=a then b else f x
  3795 may also be iterated as in f(a:=b,c:=d,...);
  3796 
  3797 * HOL/Vimage: new theory for inverse image of a function, syntax f-``B;
  3798 
  3799 * HOL/List:
  3800   - new function list_update written xs[i:=v] that updates the i-th
  3801     list position. May also be iterated as in xs[i:=a,j:=b,...].
  3802   - new function `upt' written [i..j(] which generates the list
  3803     [i,i+1,...,j-1], i.e. the upper bound is excluded. To include the upper
  3804     bound write [i..j], which is a shorthand for [i..j+1(].
  3805   - new lexicographic orderings and corresponding wellfoundedness theorems.
  3806 
  3807 * HOL/Arith:
  3808   - removed 'pred' (predecessor) function;
  3809   - generalized some theorems about n-1;
  3810   - many new laws about "div" and "mod";
  3811   - new laws about greatest common divisors (see theory ex/Primes);
  3812 
  3813 * HOL/Relation: renamed the relational operator r^-1 "converse"
  3814 instead of "inverse";
  3815 
  3816 * HOL/Induct/Multiset: a theory of multisets, including the wellfoundedness
  3817   of the multiset ordering;
  3818 
  3819 * directory HOL/Real: a construction of the reals using Dedekind cuts
  3820   (not included by default);
  3821 
  3822 * directory HOL/UNITY: Chandy and Misra's UNITY formalism;
  3823 
  3824 * directory HOL/Hoare: a new version of Hoare logic which permits many-sorted
  3825   programs, i.e. different program variables may have different types.
  3826 
  3827 * calling (stac rew i) now fails if "rew" has no effect on the goal
  3828   [previously, this check worked only if the rewrite rule was unconditional]
  3829   Now rew can involve either definitions or equalities (either == or =).
  3830 
  3831 
  3832 *** ZF ***
  3833 
  3834 * theory Main includes everything; INCOMPATIBILITY: theory ZF.thy contains
  3835   only the theorems proved on ZF.ML;
  3836 
  3837 * ZF INCOMPATIBILITY: rule `equals0D' is now a well-formed destruct rule;
  3838   It and 'sym RS equals0D' are now in the default  claset, giving automatic
  3839   disjointness reasoning but breaking a few old proofs.
  3840 
  3841 * ZF/Update: new theory of function updates
  3842     with default rewrite rule  f(x:=y) ` z = if(z=x, y, f`z)
  3843   may also be iterated as in f(a:=b,c:=d,...);
  3844 
  3845 * in  let x=t in u(x), neither t nor u(x) has to be an FOL term.
  3846 
  3847 * calling (stac rew i) now fails if "rew" has no effect on the goal
  3848   [previously, this check worked only if the rewrite rule was unconditional]
  3849   Now rew can involve either definitions or equalities (either == or =).
  3850 
  3851 * case_tac provided for compatibility with HOL
  3852     (like the old excluded_middle_tac, but with subgoals swapped)
  3853 
  3854 
  3855 *** Internal programming interfaces ***
  3856 
  3857 * Pure: several new basic modules made available for general use, see
  3858 also src/Pure/README;
  3859 
  3860 * improved the theory data mechanism to support encapsulation (data
  3861 kind name replaced by private Object.kind, acting as authorization
  3862 key); new type-safe user interface via functor TheoryDataFun; generic
  3863 print_data function becomes basically useless;
  3864 
  3865 * removed global_names compatibility flag -- all theory declarations
  3866 are qualified by default;
  3867 
  3868 * module Pure/Syntax now offers quote / antiquote translation
  3869 functions (useful for Hoare logic etc. with implicit dependencies);
  3870 see HOL/ex/Antiquote for an example use;
  3871 
  3872 * Simplifier now offers conversions (asm_)(full_)rewrite: simpset ->
  3873 cterm -> thm;
  3874 
  3875 * new tactical CHANGED_GOAL for checking that a tactic modifies a
  3876 subgoal;
  3877 
  3878 * Display.print_goals function moved to Locale.print_goals;
  3879 
  3880 * standard print function for goals supports current_goals_markers
  3881 variable for marking begin of proof, end of proof, start of goal; the
  3882 default is ("", "", ""); setting current_goals_markers := ("<proof>",
  3883 "</proof>", "<goal>") causes SGML like tagged proof state printing,
  3884 for example;
  3885 
  3886 
  3887 
  3888 New in Isabelle98 (January 1998)
  3889 --------------------------------
  3890 
  3891 *** Overview of INCOMPATIBILITIES (see below for more details) ***
  3892 
  3893 * changed lexical syntax of terms / types: dots made part of long
  3894 identifiers, e.g. "%x.x" no longer possible, should be "%x. x";
  3895 
  3896 * simpset (and claset) reference variable replaced by functions
  3897 simpset / simpset_ref;
  3898 
  3899 * no longer supports theory aliases (via merge) and non-trivial
  3900 implicit merge of thms' signatures;
  3901 
  3902 * most internal names of constants changed due to qualified names;
  3903 
  3904 * changed Pure/Sequence interface (see Pure/seq.ML);
  3905 
  3906 
  3907 *** General Changes ***
  3908 
  3909 * hierachically structured name spaces (for consts, types, axms, thms
  3910 etc.); new lexical class 'longid' (e.g. Foo.bar.x) may render much of
  3911 old input syntactically incorrect (e.g. "%x.x"); COMPATIBILITY:
  3912 isatool fixdots ensures space after dots (e.g. "%x. x"); set
  3913 long_names for fully qualified output names; NOTE: ML programs
  3914 (special tactics, packages etc.) referring to internal names may have
  3915 to be adapted to cope with fully qualified names; in case of severe
  3916 backward campatibility problems try setting 'global_names' at compile
  3917 time to have enrything declared within a flat name space; one may also
  3918 fine tune name declarations in theories via the 'global' and 'local'
  3919 section;
  3920 
  3921 * reimplemented the implicit simpset and claset using the new anytype
  3922 data filed in signatures; references simpset:simpset ref etc. are
  3923 replaced by functions simpset:unit->simpset and
  3924 simpset_ref:unit->simpset ref; COMPATIBILITY: use isatool fixclasimp
  3925 to patch your ML files accordingly;
  3926 
  3927 * HTML output now includes theory graph data for display with Java
  3928 applet or isatool browser; data generated automatically via isatool
  3929 usedir (see -i option, ISABELLE_USEDIR_OPTIONS);
  3930 
  3931 * defs may now be conditional; improved rewrite_goals_tac to handle
  3932 conditional equations;
  3933 
  3934 * defs now admits additional type arguments, using TYPE('a) syntax;
  3935 
  3936 * theory aliases via merge (e.g. M=A+B+C) no longer supported, always
  3937 creates a new theory node; implicit merge of thms' signatures is
  3938 restricted to 'trivial' ones; COMPATIBILITY: one may have to use
  3939 transfer:theory->thm->thm in (rare) cases;
  3940 
  3941 * improved handling of draft signatures / theories; draft thms (and
  3942 ctyps, cterms) are automatically promoted to real ones;
  3943 
  3944 * slightly changed interfaces for oracles: admit many per theory, named
  3945 (e.g. oracle foo = mlfun), additional name argument for invoke_oracle;
  3946 
  3947 * print_goals: optional output of const types (set show_consts and
  3948 show_types);
  3949 
  3950 * improved output of warnings (###) and errors (***);
  3951 
  3952 * subgoal_tac displays a warning if the new subgoal has type variables;
  3953 
  3954 * removed old README and Makefiles;
  3955 
  3956 * replaced print_goals_ref hook by print_current_goals_fn and result_error_fn;
  3957 
  3958 * removed obsolete init_pps and init_database;
  3959 
  3960 * deleted the obsolete tactical STATE, which was declared by
  3961     fun STATE tacfun st = tacfun st st;
  3962 
  3963 * cd and use now support path variables, e.g. $ISABELLE_HOME, or ~
  3964 (which abbreviates $HOME);
  3965 
  3966 * changed Pure/Sequence interface (see Pure/seq.ML); COMPATIBILITY:
  3967 use isatool fixseq to adapt your ML programs (this works for fully
  3968 qualified references to the Sequence structure only!);
  3969 
  3970 * use_thy no longer requires writable current directory; it always
  3971 reloads .ML *and* .thy file, if either one is out of date;
  3972 
  3973 
  3974 *** Classical Reasoner ***
  3975 
  3976 * Clarify_tac, clarify_tac, clarify_step_tac, Clarify_step_tac: new
  3977 tactics that use classical reasoning to simplify a subgoal without
  3978 splitting it into several subgoals;
  3979 
  3980 * Safe_tac: like safe_tac but uses the default claset;
  3981 
  3982 
  3983 *** Simplifier ***
  3984 
  3985 * added simplification meta rules:
  3986     (asm_)(full_)simplify: simpset -> thm -> thm;
  3987 
  3988 * simplifier.ML no longer part of Pure -- has to be loaded by object
  3989 logics (again);
  3990 
  3991 * added prems argument to simplification procedures;
  3992 
  3993 * HOL, FOL, ZF: added infix function `addsplits':
  3994   instead of `<simpset> setloop (split_tac <thms>)'
  3995   you can simply write `<simpset> addsplits <thms>'
  3996 
  3997 
  3998 *** Syntax ***
  3999 
  4000 * TYPE('a) syntax for type reflection terms;
  4001 
  4002 * no longer handles consts with name "" -- declare as 'syntax' instead;
  4003 
  4004 * pretty printer: changed order of mixfix annotation preference (again!);
  4005 
  4006 * Pure: fixed idt/idts vs. pttrn/pttrns syntactic categories;
  4007 
  4008 
  4009 *** HOL ***
  4010 
  4011 * HOL: there is a new splitter `split_asm_tac' that can be used e.g.
  4012   with `addloop' of the simplifier to faciliate case splitting in premises.
  4013 
  4014 * HOL/TLA: Stephan Merz's formalization of Lamport's Temporal Logic of Actions;
  4015 
  4016 * HOL/Auth: new protocol proofs including some for the Internet
  4017   protocol TLS;
  4018 
  4019 * HOL/Map: new theory of `maps' a la VDM;
  4020 
  4021 * HOL/simplifier: simplification procedures nat_cancel_sums for
  4022 cancelling out common nat summands from =, <, <= (in)equalities, or
  4023 differences; simplification procedures nat_cancel_factor for
  4024 cancelling common factor from =, <, <= (in)equalities over natural
  4025 sums; nat_cancel contains both kinds of procedures, it is installed by
  4026 default in Arith.thy -- this COULD MAKE EXISTING PROOFS FAIL;
  4027 
  4028 * HOL/simplifier: terms of the form
  4029   `? x. P1(x) & ... & Pn(x) & x=t & Q1(x) & ... Qn(x)'  (or t=x)
  4030   are rewritten to
  4031   `P1(t) & ... & Pn(t) & Q1(t) & ... Qn(t)',
  4032   and those of the form
  4033   `! x. P1(x) & ... & Pn(x) & x=t & Q1(x) & ... Qn(x) --> R(x)'  (or t=x)
  4034   are rewritten to
  4035   `P1(t) & ... & Pn(t) & Q1(t) & ... Qn(t) --> R(t)',
  4036 
  4037 * HOL/datatype
  4038   Each datatype `t' now comes with a theorem `split_t_case' of the form
  4039 
  4040   P(t_case f1 ... fn x) =
  4041      ( (!y1 ... ym1. x = C1 y1 ... ym1 --> P(f1 y1 ... ym1)) &
  4042         ...
  4043        (!y1 ... ymn. x = Cn y1 ... ymn --> P(f1 y1 ... ymn))
  4044      )
  4045 
  4046   and a theorem `split_t_case_asm' of the form
  4047 
  4048   P(t_case f1 ... fn x) =
  4049     ~( (? y1 ... ym1. x = C1 y1 ... ym1 & ~P(f1 y1 ... ym1)) |
  4050         ...
  4051        (? y1 ... ymn. x = Cn y1 ... ymn & ~P(f1 y1 ... ymn))
  4052      )
  4053   which can be added to a simpset via `addsplits'. The existing theorems
  4054   expand_list_case and expand_option_case have been renamed to
  4055   split_list_case and split_option_case.
  4056 
  4057 * HOL/Arithmetic:
  4058   - `pred n' is automatically converted to `n-1'.
  4059     Users are strongly encouraged not to use `pred' any longer,
  4060     because it will disappear altogether at some point.
  4061   - Users are strongly encouraged to write "0 < n" rather than
  4062     "n ~= 0". Theorems and proof tools have been modified towards this
  4063     `standard'.
  4064 
  4065 * HOL/Lists:
  4066   the function "set_of_list" has been renamed "set" (and its theorems too);
  4067   the function "nth" now takes its arguments in the reverse order and
  4068   has acquired the infix notation "!" as in "xs!n".
  4069 
  4070 * HOL/Set: UNIV is now a constant and is no longer translated to Compl{};
  4071 
  4072 * HOL/Set: The operator (UN x.B x) now abbreviates (UN x:UNIV. B x) and its
  4073   specialist theorems (like UN1_I) are gone.  Similarly for (INT x.B x);
  4074 
  4075 * HOL/record: extensible records with schematic structural subtyping
  4076 (single inheritance); EXPERIMENTAL version demonstrating the encoding,
  4077 still lacks various theorems and concrete record syntax;
  4078 
  4079 
  4080 *** HOLCF ***
  4081 
  4082 * removed "axioms" and "generated by" sections;
  4083 
  4084 * replaced "ops" section by extended "consts" section, which is capable of
  4085   handling the continuous function space "->" directly;
  4086 
  4087 * domain package:
  4088   . proves theorems immediately and stores them in the theory,
  4089   . creates hierachical name space,
  4090   . now uses normal mixfix annotations (instead of cinfix...),
  4091   . minor changes to some names and values (for consistency),
  4092   . e.g. cases -> casedist, dists_eq -> dist_eqs, [take_lemma] -> take_lemmas,
  4093   . separator between mutual domain defs: changed "," to "and",
  4094   . improved handling of sort constraints;  now they have to
  4095     appear on the left-hand side of the equations only;
  4096 
  4097 * fixed LAM <x,y,zs>.b syntax;
  4098 
  4099 * added extended adm_tac to simplifier in HOLCF -- can now discharge
  4100 adm (%x. P (t x)), where P is chainfinite and t continuous;
  4101 
  4102 
  4103 *** FOL and ZF ***
  4104 
  4105 * FOL: 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 * qed_spec_mp, qed_goal_spec_mp, qed_goalw_spec_mp are available, as
  4109 in HOL, they strip ALL and --> from proved theorems;
  4110 
  4111 
  4112 
  4113 New in Isabelle94-8 (May 1997)
  4114 ------------------------------
  4115 
  4116 *** General Changes ***
  4117 
  4118 * new utilities to build / run / maintain Isabelle etc. (in parts
  4119 still somewhat experimental); old Makefiles etc. still functional;
  4120 
  4121 * new 'Isabelle System Manual';
  4122 
  4123 * INSTALL text, together with ./configure and ./build scripts;
  4124 
  4125 * reimplemented type inference for greater efficiency, better error
  4126 messages and clean internal interface;
  4127 
  4128 * prlim command for dealing with lots of subgoals (an easier way of
  4129 setting goals_limit);
  4130 
  4131 
  4132 *** Syntax ***
  4133 
  4134 * supports alternative (named) syntax tables (parser and pretty
  4135 printer); internal interface is provided by add_modesyntax(_i);
  4136 
  4137 * Pure, FOL, ZF, HOL, HOLCF now support symbolic input and output; to
  4138 be used in conjunction with the Isabelle symbol font; uses the
  4139 "symbols" syntax table;
  4140 
  4141 * added token_translation interface (may translate name tokens in
  4142 arbitrary ways, dependent on their type (free, bound, tfree, ...) and
  4143 the current print_mode); IMPORTANT: user print translation functions
  4144 are responsible for marking newly introduced bounds
  4145 (Syntax.mark_boundT);
  4146 
  4147 * token translations for modes "xterm" and "xterm_color" that display
  4148 names in bold, underline etc. or colors (which requires a color
  4149 version of xterm);
  4150 
  4151 * infixes may now be declared with names independent of their syntax;
  4152 
  4153 * added typed_print_translation (like print_translation, but may
  4154 access type of constant);
  4155 
  4156 
  4157 *** Classical Reasoner ***
  4158 
  4159 Blast_tac: a new tactic!  It is often more powerful than fast_tac, but has
  4160 some limitations.  Blast_tac...
  4161   + ignores addss, addbefore, addafter; this restriction is intrinsic
  4162   + ignores elimination rules that don't have the correct format
  4163         (the conclusion MUST be a formula variable)
  4164   + ignores types, which can make HOL proofs fail
  4165   + rules must not require higher-order unification, e.g. apply_type in ZF
  4166     [message "Function Var's argument not a bound variable" relates to this]
  4167   + its proof strategy is more general but can actually be slower
  4168 
  4169 * substitution with equality assumptions no longer permutes other
  4170 assumptions;
  4171 
  4172 * minor changes in semantics of addafter (now called addaltern); renamed
  4173 setwrapper to setWrapper and compwrapper to compWrapper; added safe wrapper
  4174 (and access functions for it);
  4175 
  4176 * improved combination of classical reasoner and simplifier:
  4177   + functions for handling clasimpsets
  4178   + improvement of addss: now the simplifier is called _after_ the
  4179     safe steps.
  4180   + safe variant of addss called addSss: uses safe simplifications
  4181     _during_ the safe steps. It is more complete as it allows multiple
  4182     instantiations of unknowns (e.g. with slow_tac).
  4183 
  4184 *** Simplifier ***
  4185 
  4186 * added interface for simplification procedures (functions that
  4187 produce *proven* rewrite rules on the fly, depending on current
  4188 redex);
  4189 
  4190 * ordering on terms as parameter (used for ordered rewriting);
  4191 
  4192 * new functions delcongs, deleqcongs, and Delcongs. richer rep_ss;
  4193 
  4194 * the solver is now split into a safe and an unsafe part.
  4195 This should be invisible for the normal user, except that the
  4196 functions setsolver and addsolver have been renamed to setSolver and
  4197 addSolver; added safe_asm_full_simp_tac;
  4198 
  4199 
  4200 *** HOL ***
  4201 
  4202 * a generic induction tactic `induct_tac' which works for all datatypes and
  4203 also for type `nat';
  4204 
  4205 * a generic case distinction tactic `exhaust_tac' which works for all
  4206 datatypes and also for type `nat';
  4207 
  4208 * each datatype comes with a function `size';
  4209 
  4210 * patterns in case expressions allow tuple patterns as arguments to
  4211 constructors, for example `case x of [] => ... | (x,y,z)#ps => ...';
  4212 
  4213 * primrec now also works with type nat;
  4214 
  4215 * recdef: a new declaration form, allows general recursive functions to be
  4216 defined in theory files.  See HOL/ex/Fib, HOL/ex/Primes, HOL/Subst/Unify.
  4217 
  4218 * the constant for negation has been renamed from "not" to "Not" to
  4219 harmonize with FOL, ZF, LK, etc.;
  4220 
  4221 * HOL/ex/LFilter theory of a corecursive "filter" functional for
  4222 infinite lists;
  4223 
  4224 * HOL/Modelcheck demonstrates invocation of model checker oracle;
  4225 
  4226 * HOL/ex/Ring.thy declares cring_simp, which solves equational
  4227 problems in commutative rings, using axiomatic type classes for + and *;
  4228 
  4229 * more examples in HOL/MiniML and HOL/Auth;
  4230 
  4231 * more default rewrite rules for quantifiers, union/intersection;
  4232 
  4233 * a new constant `arbitrary == @x.False';
  4234 
  4235 * HOLCF/IOA replaces old HOL/IOA;
  4236 
  4237 * HOLCF changes: derived all rules and arities
  4238   + axiomatic type classes instead of classes
  4239   + typedef instead of faking type definitions
  4240   + eliminated the internal constants less_fun, less_cfun, UU_fun, UU_cfun etc.
  4241   + new axclasses cpo, chfin, flat with flat < chfin < pcpo < cpo < po
  4242   + eliminated the types void, one, tr
  4243   + use unit lift and bool lift (with translations) instead of one and tr
  4244   + eliminated blift from Lift3.thy (use Def instead of blift)
  4245   all eliminated rules are derived as theorems --> no visible changes ;
  4246 
  4247 
  4248 *** ZF ***
  4249 
  4250 * ZF now has Fast_tac, Simp_tac and Auto_tac.  Union_iff is a now a default
  4251 rewrite rule; this may affect some proofs.  eq_cs is gone but can be put back
  4252 as ZF_cs addSIs [equalityI];
  4253 
  4254 
  4255 
  4256 New in Isabelle94-7 (November 96)
  4257 ---------------------------------
  4258 
  4259 * allowing negative levels (as offsets) in prlev and choplev;
  4260 
  4261 * super-linear speedup for large simplifications;
  4262 
  4263 * FOL, ZF and HOL now use miniscoping: rewriting pushes
  4264 quantifications in as far as possible (COULD MAKE EXISTING PROOFS
  4265 FAIL); can suppress it using the command Delsimps (ex_simps @
  4266 all_simps); De Morgan laws are also now included, by default;
  4267 
  4268 * improved printing of ==>  :  ~:
  4269 
  4270 * new object-logic "Sequents" adds linear logic, while replacing LK
  4271 and Modal (thanks to Sara Kalvala);
  4272 
  4273 * HOL/Auth: correctness proofs for authentication protocols;
  4274 
  4275 * HOL: new auto_tac combines rewriting and classical reasoning (many
  4276 examples on HOL/Auth);
  4277 
  4278 * HOL: new command AddIffs for declaring theorems of the form P=Q to
  4279 the rewriter and classical reasoner simultaneously;
  4280 
  4281 * function uresult no longer returns theorems in "standard" format;
  4282 regain previous version by: val uresult = standard o uresult;
  4283 
  4284 
  4285 
  4286 New in Isabelle94-6
  4287 -------------------
  4288 
  4289 * oracles -- these establish an interface between Isabelle and trusted
  4290 external reasoners, which may deliver results as theorems;
  4291 
  4292 * proof objects (in particular record all uses of oracles);
  4293 
  4294 * Simp_tac, Fast_tac, etc. that refer to implicit simpset / claset;
  4295 
  4296 * "constdefs" section in theory files;
  4297 
  4298 * "primrec" section (HOL) no longer requires names;
  4299 
  4300 * internal type "tactic" now simply "thm -> thm Sequence.seq";
  4301 
  4302 
  4303 
  4304 New in Isabelle94-5
  4305 -------------------
  4306 
  4307 * reduced space requirements;
  4308 
  4309 * automatic HTML generation from theories;
  4310 
  4311 * theory files no longer require "..." (quotes) around most types;
  4312 
  4313 * new examples, including two proofs of the Church-Rosser theorem;
  4314 
  4315 * non-curried (1994) version of HOL is no longer distributed;
  4316 
  4317 
  4318 
  4319 New in Isabelle94-4
  4320 -------------------
  4321 
  4322 * greatly reduced space requirements;
  4323 
  4324 * theory files (.thy) no longer require \...\ escapes at line breaks;
  4325 
  4326 * searchable theorem database (see the section "Retrieving theorems" on
  4327 page 8 of the Reference Manual);
  4328 
  4329 * new examples, including Grabczewski's monumental case study of the
  4330 Axiom of Choice;
  4331 
  4332 * The previous version of HOL renamed to Old_HOL;
  4333 
  4334 * The new version of HOL (previously called CHOL) uses a curried syntax
  4335 for functions.  Application looks like f a b instead of f(a,b);
  4336 
  4337 * Mutually recursive inductive definitions finally work in HOL;
  4338 
  4339 * In ZF, pattern-matching on tuples is now available in all abstractions and
  4340 translates to the operator "split";
  4341 
  4342 
  4343 
  4344 New in Isabelle94-3
  4345 -------------------
  4346 
  4347 * new infix operator, addss, allowing the classical reasoner to
  4348 perform simplification at each step of its search.  Example:
  4349         fast_tac (cs addss ss)
  4350 
  4351 * a new logic, CHOL, the same as HOL, but with a curried syntax
  4352 for functions.  Application looks like f a b instead of f(a,b).  Also pairs
  4353 look like (a,b) instead of <a,b>;
  4354 
  4355 * PLEASE NOTE: CHOL will eventually replace HOL!
  4356 
  4357 * In CHOL, pattern-matching on tuples is now available in all abstractions.
  4358 It translates to the operator "split".  A new theory of integers is available;
  4359 
  4360 * In ZF, integer numerals now denote two's-complement binary integers.
  4361 Arithmetic operations can be performed by rewriting.  See ZF/ex/Bin.ML;
  4362 
  4363 * Many new examples: I/O automata, Church-Rosser theorem, equivalents
  4364 of the Axiom of Choice;
  4365 
  4366 
  4367 
  4368 New in Isabelle94-2
  4369 -------------------
  4370 
  4371 * Significantly faster resolution;
  4372 
  4373 * the different sections in a .thy file can now be mixed and repeated
  4374 freely;
  4375 
  4376 * Database of theorems for FOL, HOL and ZF.  New
  4377 commands including qed, qed_goal and bind_thm store theorems in the database.
  4378 
  4379 * Simple database queries: return a named theorem (get_thm) or all theorems of
  4380 a given theory (thms_of), or find out what theory a theorem was proved in
  4381 (theory_of_thm);
  4382 
  4383 * Bugs fixed in the inductive definition and datatype packages;
  4384 
  4385 * The classical reasoner provides deepen_tac and depth_tac, making FOL_dup_cs
  4386 and HOL_dup_cs obsolete;
  4387 
  4388 * Syntactic ambiguities caused by the new treatment of syntax in Isabelle94-1
  4389 have been removed;
  4390 
  4391 * Simpler definition of function space in ZF;
  4392 
  4393 * new results about cardinal and ordinal arithmetic in ZF;
  4394 
  4395 * 'subtype' facility in HOL for introducing new types as subsets of existing
  4396 types;
  4397 
  4398 
  4399 $Id$