NEWS
author ballarin
Mon May 12 15:51:34 2003 +0200 (2003-05-12)
changeset 14018 3ef6c509f29c
parent 14011 5fa103f5618c
child 14119 fb9c392644a1
permissions -rw-r--r--
Improved entry on Algebra.
     1 Isabelle NEWS -- history user-relevant changes
     2 ==============================================
     3 
     4 New in this Isabelle release
     5 ----------------------------
     6 
     7 *** General ***
     8 
     9 * Provers/simplifier:
    10 
    11   - Completely reimplemented method simp (ML: Asm_full_simp_tac):
    12     Assumptions are now subject to complete mutual simplification,
    13     not just from left to right. The simplifier now preserves
    14     the order of assumptions.
    15 
    16     Potential INCOMPATIBILITY:
    17 
    18     -- simp sometimes diverges where the old version did
    19        not, e.g. invoking simp on the goal
    20 
    21         [| P (f x); y = x; f x = f y |] ==> Q
    22 
    23        now gives rise to the infinite reduction sequence
    24 
    25         P(f x) --(f x = f y)--> P(f y) --(y = x)--> P(f x) --(f x = f y)--> ...
    26 
    27        Using "simp (asm_lr)" (ML: Asm_lr_simp_tac) instead often solves this
    28        kind of problem.
    29 
    30     -- Tactics combining classical reasoner and simplification (such as auto)
    31        are also affected by this change, because many of them rely on
    32        simp. They may sometimes diverge as well or yield a different numbers
    33        of subgoals. Try to use e.g. force, fastsimp, or safe instead of auto
    34        in case of problems. Sometimes subsequent calls to the classical
    35        reasoner will fail because a preceeding call to the simplifier too
    36        eagerly simplified the goal, e.g. deleted redundant premises.
    37 
    38   - The simplifier trace now shows the names of the applied rewrite rules
    39 
    40   - You can limit the number of recursive invocations of the simplifier
    41     during conditional rewriting (where the simplifie tries to solve the
    42     conditions before applying the rewrite rule):
    43     ML "simp_depth_limit := n"
    44     where n is an integer. Thus you can force termination where previously
    45     the simplifier would diverge.
    46 
    47   - Accepts free variables as head terms in congruence rules.  Useful in Isar.
    48 
    49   - No longer aborts on failed congruence proof.  Instead, the
    50     congruence is ignored.
    51 
    52 * Pure: New generic framework for extracting programs from constructive
    53   proofs. See HOL/Extraction.thy for an example instantiation, as well
    54   as HOL/Extraction for some case studies.
    55 
    56 * Pure: The main goal of the proof state is no longer shown by default, only
    57 the subgoals. This behaviour is controlled by a new flag.
    58    PG menu: Isabelle/Isar -> Settings -> Show Main Goal
    59 (ML: Proof.show_main_goal).
    60 
    61 * Pure: You can find all matching introduction rules for subgoal 1, i.e. all
    62 rules whose conclusion matches subgoal 1:
    63       PG menu: Isabelle/Isar -> Show me -> matching rules
    64 The rules are ordered by how closely they match the subgoal.
    65 In particular, rules that solve a subgoal outright are displayed first
    66 (or rather last, the way they are printed).
    67 (ML: ProofGeneral.print_intros())
    68 
    69 * Pure: New flag trace_unify_fail causes unification to print
    70 diagnostic information (PG: in trace buffer) when it fails. This is
    71 useful for figuring out why single step proofs like rule, erule or
    72 assumption failed.
    73 
    74 * Pure: Locale specifications now produce predicate definitions
    75 according to the body of text (covering assumptions modulo local
    76 definitions); predicate "loc_axioms" covers newly introduced text,
    77 while "loc" is cumulative wrt. all included locale expressions; the
    78 latter view is presented only on export into the global theory
    79 context; potential INCOMPATIBILITY, use "(open)" option to fall back
    80 on the old view without predicates;
    81 
    82 * Pure: predefined locales "var" and "struct" are useful for sharing
    83 parameters (as in CASL, for example); just specify something like
    84 ``var x + var y + struct M'' as import;
    85 
    86 * Pure: improved thms_containing: proper indexing of facts instead of
    87 raw theorems; check validity of results wrt. current name space;
    88 include local facts of proof configuration (also covers active
    89 locales), cover fixed variables in index; may use "_" in term
    90 specification; an optional limit for the number of printed facts may
    91 be given (the default is 40);
    92 
    93 * Pure: disallow duplicate fact bindings within new-style theory files
    94 (batch-mode only);
    95 
    96 * Provers: improved induct method: assumptions introduced by case
    97 "foo" are split into "foo.hyps" (from the rule) and "foo.prems" (from
    98 the goal statement); "foo" still refers to all facts collectively;
    99 
   100 * Provers: the function blast.overloaded has been removed: all constants
   101 are regarded as potentially overloaded, which improves robustness in exchange
   102 for slight decrease in efficiency;
   103 
   104 * Provers/linorder: New generic prover for transitivity reasoning over
   105 linear orders.  Note: this prover is not efficient!
   106 
   107 * Isar: preview of problems to finish 'show' now produce an error
   108 rather than just a warning (in interactive mode);
   109 
   110 
   111 *** HOL ***
   112 
   113 * arith(_tac)
   114 
   115  - Produces a counter example if it cannot prove a goal.
   116    Note that the counter example may be spurious if the goal is not a formula
   117    of quantifier-free linear arithmetic.
   118    In ProofGeneral the counter example appears in the trace buffer.
   119 
   120  - Knows about div k and mod k where k is a numeral of type nat or int.
   121 
   122  - Calls full Presburger arithmetic (by Amine Chaieb) if quantifier-free
   123    linear arithmetic fails. This takes account of quantifiers and divisibility.
   124    Presburger arithmetic can also be called explicitly via presburger(_tac). 
   125 
   126 * simp's arithmetic capabilities have been enhanced a bit: it now
   127 takes ~= in premises into account (by performing a case split);
   128 
   129 * simp reduces "m*(n div m) + n mod m" to n, even if the two summands
   130 are distributed over a sum of terms;
   131 
   132 * New tactic "trans_tac" and method "trans" instantiate
   133 Provers/linorder.ML for axclasses "order" and "linorder" (predicates
   134 "<=", "<" and "="). 
   135 
   136 * function INCOMPATIBILITIES: Pi-sets have been redefined and moved from main 
   137 HOL to Library/FuncSet; constant "Fun.op o" is now called "Fun.comp";
   138 
   139 * 'typedef' command has new option "open" to suppress the set
   140 definition;
   141 
   142 * functions Min and Max on finite sets have been introduced (theory
   143 Finite_Set);
   144 
   145 * attribute [symmetric] now works for relations as well; it turns
   146 (x,y) : R^-1 into (y,x) : R, and vice versa;
   147 
   148 * induct over a !!-quantified statement (say !!x1..xn):
   149   each "case" automatically performs "fix x1 .. xn" with exactly those names.
   150 
   151 * Map: `empty' is no longer a constant but a syntactic abbreviation for
   152 %x. None. Warning: empty_def now refers to the previously hidden definition
   153 of the empty set.
   154 
   155 * Algebra: formalization of classical algebra.  Intended as base for
   156 any algebraic development in Isabelle.  Currently covers group theory
   157 (up to Sylow's theorem) and ring theory (Universal Property of
   158 Univariate Polynomials).  Contributions welcome;
   159 
   160 * GroupTheory: deleted, since its material has been moved to Algebra;
   161 
   162 * Complex: new directory of the complex numbers with numeric constants, 
   163 nonstandard complex numbers, and some complex analysis, standard and 
   164 nonstandard (Jacques Fleuriot);
   165 
   166 * HOL-Complex: new image for analysis, replacing HOL-Real and HOL-Hyperreal;
   167 
   168 * Hyperreal: introduced Gauge integration and hyperreal logarithms (Jacques 
   169 Fleuriot);
   170 
   171 * Real/HahnBanach: updated and adapted to locales;
   172 
   173 * NumberTheory: added Gauss's law of quadratic reciprocity (by Avigad,
   174 Gray and Kramer);
   175 
   176 * UNITY: added the Meier-Sanders theory of progress sets;
   177 
   178 * MicroJava: bytecode verifier and lightweight bytecode verifier
   179 as abstract algorithms, instantiated to the JVM;
   180 
   181 * Bali: Java source language formalization. Type system, operational
   182 semantics, axiomatic semantics. Supported language features:
   183 classes, interfaces, objects,virtual methods, static methods,
   184 static/instance fields, arrays, access modifiers, definite
   185 assignment, exceptions.
   186 
   187 
   188 *** ZF ***
   189 
   190 * ZF/Constructible: consistency proof for AC (Gödel's constructible
   191 universe, etc.);
   192 
   193 * Main ZF: virtually all theories converted to new-style format;
   194 
   195 
   196 *** ML ***
   197 
   198 * Pure: Tactic.prove provides sane interface for internal proofs;
   199 omits the infamous "standard" operation, so this is more appropriate
   200 than prove_goalw_cterm in many situations (e.g. in simprocs);
   201 
   202 * Pure: improved error reporting of simprocs;
   203 
   204 * Provers: Simplifier.simproc(_i) provides sane interface for setting
   205 up simprocs;
   206 
   207 
   208 *** Document preparation ***
   209 
   210 * uses \par instead of \\ for line breaks in theory text. This may
   211 shift some page breaks in large documents. To get the old behaviour
   212 use \renewcommand{\isanewline}{\mbox{}\\\mbox{}} in root.tex.
   213 
   214 * minimized dependencies of isabelle.sty and isabellesym.sty on 
   215 other packages
   216 
   217 * \<euro> now needs package babel/greek instead of marvosym (which
   218 broke \Rightarrow)
   219 
   220 * normal size for \<zero>...\<nine> (uses \mathbf instead of 
   221 textcomp package)
   222 
   223 
   224 New in Isabelle2002 (March 2002)
   225 --------------------------------
   226 
   227 *** Document preparation ***
   228 
   229 * greatly simplified document preparation setup, including more
   230 graceful interpretation of isatool usedir -i/-d/-D options, and more
   231 instructive isatool mkdir; users should basically be able to get
   232 started with "isatool mkdir HOL Test && isatool make"; alternatively,
   233 users may run a separate document processing stage manually like this:
   234 "isatool usedir -D output HOL Test && isatool document Test/output";
   235 
   236 * theory dependency graph may now be incorporated into documents;
   237 isatool usedir -g true will produce session_graph.eps/.pdf for use
   238 with \includegraphics of LaTeX;
   239 
   240 * proper spacing of consecutive markup elements, especially text
   241 blocks after section headings;
   242 
   243 * support bold style (for single symbols only), input syntax is like
   244 this: "\<^bold>\<alpha>" or "\<^bold>A";
   245 
   246 * \<bullet> is now output as bold \cdot by default, which looks much
   247 better in printed text;
   248 
   249 * added default LaTeX bindings for \<tturnstile> and \<TTurnstile>;
   250 note that these symbols are currently unavailable in Proof General /
   251 X-Symbol; new symbols \<zero>, \<one>, ..., \<nine>, and \<euro>;
   252 
   253 * isatool latex no longer depends on changed TEXINPUTS, instead
   254 isatool document copies the Isabelle style files to the target
   255 location;
   256 
   257 
   258 *** Isar ***
   259 
   260 * Pure/Provers: improved proof by cases and induction;
   261   - 'case' command admits impromptu naming of parameters (such as
   262     "case (Suc n)");
   263   - 'induct' method divinates rule instantiation from the inductive
   264     claim; no longer requires excessive ?P bindings for proper
   265     instantiation of cases;
   266   - 'induct' method properly enumerates all possibilities of set/type
   267     rules; as a consequence facts may be also passed through *type*
   268     rules without further ado;
   269   - 'induct' method now derives symbolic cases from the *rulified*
   270     rule (before it used to rulify cases stemming from the internal
   271     atomized version); this means that the context of a non-atomic
   272     statement becomes is included in the hypothesis, avoiding the
   273     slightly cumbersome show "PROP ?case" form;
   274   - 'induct' may now use elim-style induction rules without chaining
   275     facts, using ``missing'' premises from the goal state; this allows
   276     rules stemming from inductive sets to be applied in unstructured
   277     scripts, while still benefitting from proper handling of non-atomic
   278     statements; NB: major inductive premises need to be put first, all
   279     the rest of the goal is passed through the induction;
   280   - 'induct' proper support for mutual induction involving non-atomic
   281     rule statements (uses the new concept of simultaneous goals, see
   282     below);
   283   - append all possible rule selections, but only use the first
   284     success (no backtracking);
   285   - removed obsolete "(simplified)" and "(stripped)" options of methods;
   286   - undeclared rule case names default to numbers 1, 2, 3, ...;
   287   - added 'print_induct_rules' (covered by help item in recent Proof
   288     General versions);
   289   - moved induct/cases attributes to Pure, methods to Provers;
   290   - generic method setup instantiated for FOL and HOL;
   291 
   292 * Pure: support multiple simultaneous goal statements, for example
   293 "have a: A and b: B" (same for 'theorem' etc.); being a pure
   294 meta-level mechanism, this acts as if several individual goals had
   295 been stated separately; in particular common proof methods need to be
   296 repeated in order to cover all claims; note that a single elimination
   297 step is *not* sufficient to establish the two conjunctions, so this
   298 fails:
   299 
   300   assume "A & B" then have A and B ..   (*".." fails*)
   301 
   302 better use "obtain" in situations as above; alternative refer to
   303 multi-step methods like 'auto', 'simp_all', 'blast+' etc.;
   304 
   305 * Pure: proper integration with ``locales''; unlike the original
   306 version by Florian Kammüller, Isar locales package high-level proof
   307 contexts rather than raw logical ones (e.g. we admit to include
   308 attributes everywhere); operations on locales include merge and
   309 rename; support for implicit arguments (``structures''); simultaneous
   310 type-inference over imports and text; see also HOL/ex/Locales.thy for
   311 some examples;
   312 
   313 * Pure: the following commands have been ``localized'', supporting a
   314 target locale specification "(in name)": 'lemma', 'theorem',
   315 'corollary', 'lemmas', 'theorems', 'declare'; the results will be
   316 stored both within the locale and at the theory level (exported and
   317 qualified by the locale name);
   318 
   319 * Pure: theory goals may now be specified in ``long'' form, with
   320 ad-hoc contexts consisting of arbitrary locale elements. for example
   321 ``lemma foo: fixes x assumes "A x" shows "B x"'' (local syntax and
   322 definitions may be given, too); the result is a meta-level rule with
   323 the context elements being discharged in the obvious way;
   324 
   325 * Pure: new proof command 'using' allows to augment currently used
   326 facts after a goal statement ('using' is syntactically analogous to
   327 'apply', but acts on the goal's facts only); this allows chained facts
   328 to be separated into parts given before and after a claim, as in
   329 ``from a and b have C using d and e <proof>'';
   330 
   331 * Pure: renamed "antecedent" case to "rule_context";
   332 
   333 * Pure: new 'judgment' command records explicit information about the
   334 object-logic embedding (used by several tools internally); no longer
   335 use hard-wired "Trueprop";
   336 
   337 * Pure: added 'corollary' command;
   338 
   339 * Pure: fixed 'token_translation' command;
   340 
   341 * Pure: removed obsolete 'exported' attribute;
   342 
   343 * Pure: dummy pattern "_" in is/let is now automatically lifted over
   344 bound variables: "ALL x. P x --> Q x" (is "ALL x. _ --> ?C x")
   345 supersedes more cumbersome ... (is "ALL x. _ x --> ?C x");
   346 
   347 * Pure: method 'atomize' presents local goal premises as object-level
   348 statements (atomic meta-level propositions); setup controlled via
   349 rewrite rules declarations of 'atomize' attribute; example
   350 application: 'induct' method with proper rule statements in improper
   351 proof *scripts*;
   352 
   353 * Pure: emulation of instantiation tactics (rule_tac, cut_tac, etc.)
   354 now consider the syntactic context of assumptions, giving a better
   355 chance to get type-inference of the arguments right (this is
   356 especially important for locales);
   357 
   358 * Pure: "sorry" no longer requires quick_and_dirty in interactive
   359 mode;
   360 
   361 * Pure/obtain: the formal conclusion "thesis", being marked as
   362 ``internal'', may no longer be reference directly in the text;
   363 potential INCOMPATIBILITY, may need to use "?thesis" in rare
   364 situations;
   365 
   366 * Pure: generic 'sym' attribute which declares a rule both as pure
   367 'elim?' and for the 'symmetric' operation;
   368 
   369 * Pure: marginal comments ``--'' may now occur just anywhere in the
   370 text; the fixed correlation with particular command syntax has been
   371 discontinued;
   372 
   373 * Pure: new method 'rules' is particularly well-suited for proof
   374 search in intuitionistic logic; a bit slower than 'blast' or 'fast',
   375 but often produces more compact proof terms with less detours;
   376 
   377 * Pure/Provers/classical: simplified integration with pure rule
   378 attributes and methods; the classical "intro?/elim?/dest?"
   379 declarations coincide with the pure ones; the "rule" method no longer
   380 includes classically swapped intros; "intro" and "elim" methods no
   381 longer pick rules from the context; also got rid of ML declarations
   382 AddXIs/AddXEs/AddXDs; all of this has some potential for
   383 INCOMPATIBILITY;
   384 
   385 * Provers/classical: attribute 'swapped' produces classical inversions
   386 of introduction rules;
   387 
   388 * Provers/simplifier: 'simplified' attribute may refer to explicit
   389 rules instead of full simplifier context; 'iff' attribute handles
   390 conditional rules;
   391 
   392 * HOL: 'typedef' now allows alternative names for Rep/Abs morphisms;
   393 
   394 * HOL: 'recdef' now fails on unfinished automated proofs, use
   395 "(permissive)" option to recover old behavior;
   396 
   397 * HOL: 'inductive' no longer features separate (collective) attributes
   398 for 'intros' (was found too confusing);
   399 
   400 * HOL: properly declared induction rules less_induct and
   401 wf_induct_rule;
   402 
   403 
   404 *** HOL ***
   405 
   406 * HOL: moved over to sane numeral syntax; the new policy is as
   407 follows:
   408 
   409   - 0 and 1 are polymorphic constants, which are defined on any
   410   numeric type (nat, int, real etc.);
   411 
   412   - 2, 3, 4, ... and -1, -2, -3, ... are polymorphic numerals, based
   413   binary representation internally;
   414 
   415   - type nat has special constructor Suc, and generally prefers Suc 0
   416   over 1::nat and Suc (Suc 0) over 2::nat;
   417 
   418 This change may cause significant problems of INCOMPATIBILITY; here
   419 are some hints on converting existing sources:
   420 
   421   - due to the new "num" token, "-0" and "-1" etc. are now atomic
   422   entities, so expressions involving "-" (unary or binary minus) need
   423   to be spaced properly;
   424 
   425   - existing occurrences of "1" may need to be constraint "1::nat" or
   426   even replaced by Suc 0; similar for old "2";
   427 
   428   - replace "#nnn" by "nnn", and "#-nnn" by "-nnn";
   429 
   430   - remove all special provisions on numerals in proofs;
   431 
   432 * HOL: simp rules nat_number expand numerals on nat to Suc/0
   433 representation (depends on bin_arith_simps in the default context);
   434 
   435 * HOL: symbolic syntax for x^2 (numeral 2);
   436 
   437 * HOL: the class of all HOL types is now called "type" rather than
   438 "term"; INCOMPATIBILITY, need to adapt references to this type class
   439 in axclass/classes, instance/arities, and (usually rare) occurrences
   440 in typings (of consts etc.); internally the class is called
   441 "HOL.type", ML programs should refer to HOLogic.typeS;
   442 
   443 * HOL/record package improvements:
   444   - new derived operations "fields" to build a partial record section,
   445     "extend" to promote a fixed record to a record scheme, and
   446     "truncate" for the reverse; cf. theorems "xxx.defs", which are *not*
   447     declared as simp by default;
   448   - shared operations ("more", "fields", etc.) now need to be always
   449     qualified) --- potential INCOMPATIBILITY;
   450   - removed "make_scheme" operations (use "make" with "extend") --
   451     INCOMPATIBILITY;
   452   - removed "more" class (simply use "term") -- INCOMPATIBILITY;
   453   - provides cases/induct rules for use with corresponding Isar
   454     methods (for concrete records, record schemes, concrete more
   455     parts, and schematic more parts -- in that order);
   456   - internal definitions directly based on a light-weight abstract
   457     theory of product types over typedef rather than datatype;
   458 
   459 * HOL: generic code generator for generating executable ML code from
   460 specifications; specific support for HOL constructs such as inductive
   461 datatypes and sets, as well as recursive functions; can be invoked
   462 via 'generate_code' theory section;
   463 
   464 * HOL: canonical cases/induct rules for n-tuples (n = 3..7);
   465 
   466 * HOL: consolidated and renamed several theories.  In particular:
   467 	Ord.thy has been absorbed into HOL.thy
   468 	String.thy has been absorbed into List.thy
   469  
   470 * HOL: concrete setsum syntax "\<Sum>i:A. b" == "setsum (%i. b) A"
   471 (beware of argument permutation!);
   472 
   473 * HOL: linorder_less_split superseded by linorder_cases;
   474 
   475 * HOL/List: "nodups" renamed to "distinct";
   476 
   477 * HOL: added "The" definite description operator; move Hilbert's "Eps"
   478 to peripheral theory "Hilbert_Choice"; some INCOMPATIBILITIES:
   479   - Ex_def has changed, now need to use some_eq_ex
   480 
   481 * HOL: made split_all_tac safe; EXISTING PROOFS MAY FAIL OR LOOP, so
   482 in this (rare) case use:
   483 
   484   delSWrapper "split_all_tac"
   485   addSbefore ("unsafe_split_all_tac", unsafe_split_all_tac)
   486 
   487 * HOL: added safe wrapper "split_conv_tac" to claset; EXISTING PROOFS
   488 MAY FAIL;
   489 
   490 * HOL: introduced f^n = f o ... o f; warning: due to the limits of
   491 Isabelle's type classes, ^ on functions and relations has too general
   492 a domain, namely ('a * 'b) set and 'a => 'b; this means that it may be
   493 necessary to attach explicit type constraints;
   494 
   495 * HOL/Relation: the prefix name of the infix "O" has been changed from
   496 "comp" to "rel_comp"; INCOMPATIBILITY: a few theorems have been
   497 renamed accordingly (eg "compI" -> "rel_compI").
   498 
   499 * HOL: syntax translations now work properly with numerals and records
   500 expressions;
   501 
   502 * HOL: bounded abstraction now uses syntax "%" / "\<lambda>" instead
   503 of "lam" -- INCOMPATIBILITY;
   504 
   505 * HOL: got rid of some global declarations (potential INCOMPATIBILITY
   506 for ML tools): const "()" renamed "Product_Type.Unity", type "unit"
   507 renamed "Product_Type.unit";
   508 
   509 * HOL: renamed rtrancl_into_rtrancl2 to converse_rtrancl_into_rtrancl
   510 
   511 * HOL: removed obsolete theorem "optionE" (use "option.exhaust", or
   512 the "cases" method);
   513 
   514 * HOL/GroupTheory: group theory examples including Sylow's theorem (by
   515 Florian Kammüller);
   516 
   517 * HOL/IMP: updated and converted to new-style theory format; several
   518 parts turned into readable document, with proper Isar proof texts and
   519 some explanations (by Gerwin Klein);
   520 
   521 * HOL-Real: added Complex_Numbers (by Gertrud Bauer);
   522 
   523 * HOL-Hyperreal is now a logic image;
   524 
   525 
   526 *** HOLCF ***
   527 
   528 * Isar: consts/constdefs supports mixfix syntax for continuous
   529 operations;
   530 
   531 * Isar: domain package adapted to new-style theory format, e.g. see
   532 HOLCF/ex/Dnat.thy;
   533 
   534 * theory Lift: proper use of rep_datatype lift instead of ML hacks --
   535 potential INCOMPATIBILITY; now use plain induct_tac instead of former
   536 lift.induct_tac, always use UU instead of Undef;
   537 
   538 * HOLCF/IMP: updated and converted to new-style theory;
   539 
   540 
   541 *** ZF ***
   542 
   543 * Isar: proper integration of logic-specific tools and packages,
   544 including theory commands '(co)inductive', '(co)datatype',
   545 'rep_datatype', 'inductive_cases', as well as methods 'ind_cases',
   546 'induct_tac', 'case_tac', and 'typecheck' (with attribute 'TC');
   547 
   548 * theory Main no longer includes AC; for the Axiom of Choice, base
   549 your theory on Main_ZFC;
   550 
   551 * the integer library now covers quotients and remainders, with many
   552 laws relating division to addition, multiplication, etc.;
   553 
   554 * ZF/UNITY: Chandy and Misra's UNITY is now available in ZF, giving a
   555 typeless version of the formalism;
   556 
   557 * ZF/AC, Coind, IMP, Resid: updated and converted to new-style theory
   558 format;
   559 
   560 * ZF/Induct: new directory for examples of inductive definitions,
   561 including theory Multiset for multiset orderings; converted to
   562 new-style theory format;
   563 
   564 * ZF: many new theorems about lists, ordinals, etc.;
   565 
   566 
   567 *** General ***
   568 
   569 * Pure/kernel: meta-level proof terms (by Stefan Berghofer); reference
   570 variable proof controls level of detail: 0 = no proofs (only oracle
   571 dependencies), 1 = lemma dependencies, 2 = compact proof terms; see
   572 also ref manual for further ML interfaces;
   573 
   574 * Pure/axclass: removed obsolete ML interface
   575 goal_subclass/goal_arity;
   576 
   577 * Pure/syntax: new token syntax "num" for plain numerals (without "#"
   578 of "xnum"); potential INCOMPATIBILITY, since -0, -1 etc. are now
   579 separate tokens, so expressions involving minus need to be spaced
   580 properly;
   581 
   582 * Pure/syntax: support non-oriented infixes, using keyword "infix"
   583 rather than "infixl" or "infixr";
   584 
   585 * Pure/syntax: concrete syntax for dummy type variables admits genuine
   586 sort constraint specifications in type inference; e.g. "x::_::foo"
   587 ensures that the type of "x" is of sort "foo" (but not necessarily a
   588 type variable);
   589 
   590 * Pure/syntax: print modes "type_brackets" and "no_type_brackets"
   591 control output of nested => (types); the default behavior is
   592 "type_brackets";
   593 
   594 * Pure/syntax: builtin parse translation for "_constify" turns valued
   595 tokens into AST constants;
   596 
   597 * Pure/syntax: prefer later declarations of translations and print
   598 translation functions; potential INCOMPATIBILITY: need to reverse
   599 multiple declarations for same syntax element constant;
   600 
   601 * Pure/show_hyps reset by default (in accordance to existing Isar
   602 practice);
   603 
   604 * Provers/classical: renamed addaltern to addafter, addSaltern to
   605 addSafter;
   606 
   607 * Provers/clasimp: ``iff'' declarations now handle conditional rules
   608 as well;
   609 
   610 * system: tested support for MacOS X; should be able to get Isabelle +
   611 Proof General to work in a plain Terminal after installing Poly/ML
   612 (e.g. from the Isabelle distribution area) and GNU bash alone
   613 (e.g. from http://www.apple.com); full X11, XEmacs and X-Symbol
   614 support requires further installations, e.g. from
   615 http://fink.sourceforge.net/);
   616 
   617 * system: support Poly/ML 4.1.1 (able to manage larger heaps);
   618 
   619 * system: reduced base memory usage by Poly/ML (approx. 20 MB instead
   620 of 40 MB), cf. ML_OPTIONS;
   621 
   622 * system: Proof General keywords specification is now part of the
   623 Isabelle distribution (see etc/isar-keywords.el);
   624 
   625 * system: support for persistent Proof General sessions (refrain from
   626 outdating all loaded theories on startup); user may create writable
   627 logic images like this: ``isabelle -q HOL Test'';
   628 
   629 * system: smart selection of Isabelle process versus Isabelle
   630 interface, accommodates case-insensitive file systems (e.g. HFS+); may
   631 run both "isabelle" and "Isabelle" even if file names are badly
   632 damaged (executable inspects the case of the first letter of its own
   633 name); added separate "isabelle-process" and "isabelle-interface";
   634 
   635 * system: refrain from any attempt at filtering input streams; no
   636 longer support ``8bit'' encoding of old isabelle font, instead proper
   637 iso-latin characters may now be used; the related isatools
   638 "symbolinput" and "nonascii" have disappeared as well;
   639 
   640 * system: removed old "xterm" interface (the print modes "xterm" and
   641 "xterm_color" are still available for direct use in a suitable
   642 terminal);
   643 
   644 
   645 
   646 New in Isabelle99-2 (February 2001)
   647 -----------------------------------
   648 
   649 *** Overview of INCOMPATIBILITIES ***
   650 
   651 * HOL: please note that theories in the Library and elsewhere often use the
   652 new-style (Isar) format; to refer to their theorems in an ML script you must
   653 bind them to ML identifers by e.g.      val thm_name = thm "thm_name";
   654 
   655 * HOL: inductive package no longer splits induction rule aggressively,
   656 but only as far as specified by the introductions given; the old
   657 format may be recovered via ML function complete_split_rule or attribute
   658 'split_rule (complete)';
   659 
   660 * HOL: induct renamed to lfp_induct, lfp_Tarski to lfp_unfold,
   661 gfp_Tarski to gfp_unfold;
   662 
   663 * HOL: contrapos, contrapos2 renamed to contrapos_nn, contrapos_pp;
   664 
   665 * HOL: infix "dvd" now has priority 50 rather than 70 (because it is a
   666 relation); infix "^^" has been renamed "``"; infix "``" has been
   667 renamed "`"; "univalent" has been renamed "single_valued";
   668 
   669 * HOL/Real: "rinv" and "hrinv" replaced by overloaded "inverse"
   670 operation;
   671 
   672 * HOLCF: infix "`" has been renamed "$"; the symbol syntax is \<cdot>;
   673 
   674 * Isar: 'obtain' no longer declares "that" fact as simp/intro;
   675 
   676 * Isar/HOL: method 'induct' now handles non-atomic goals; as a
   677 consequence, it is no longer monotonic wrt. the local goal context
   678 (which is now passed through the inductive cases);
   679 
   680 * Document preparation: renamed standard symbols \<ll> to \<lless> and
   681 \<gg> to \<ggreater>;
   682 
   683 
   684 *** Document preparation ***
   685 
   686 * \isabellestyle{NAME} selects version of Isabelle output (currently
   687 available: are "it" for near math-mode best-style output, "sl" for
   688 slanted text style, and "tt" for plain type-writer; if no
   689 \isabellestyle command is given, output is according to slanted
   690 type-writer);
   691 
   692 * support sub/super scripts (for single symbols only), input syntax is
   693 like this: "A\<^sup>*" or "A\<^sup>\<star>";
   694 
   695 * some more standard symbols; see Appendix A of the system manual for
   696 the complete list of symbols defined in isabellesym.sty;
   697 
   698 * improved isabelle style files; more abstract symbol implementation
   699 (should now use \isamath{...} and \isatext{...} in custom symbol
   700 definitions);
   701 
   702 * antiquotation @{goals} and @{subgoals} for output of *dynamic* goals
   703 state; Note that presentation of goal states does not conform to
   704 actual human-readable proof documents.  Please do not include goal
   705 states into document output unless you really know what you are doing!
   706 
   707 * proper indentation of antiquoted output with proportional LaTeX
   708 fonts;
   709 
   710 * no_document ML operator temporarily disables LaTeX document
   711 generation;
   712 
   713 * isatool unsymbolize tunes sources for plain ASCII communication;
   714 
   715 
   716 *** Isar ***
   717 
   718 * Pure: Isar now suffers initial goal statements to contain unbound
   719 schematic variables (this does not conform to actual readable proof
   720 documents, due to unpredictable outcome and non-compositional proof
   721 checking); users who know what they are doing may use schematic goals
   722 for Prolog-style synthesis of proven results;
   723 
   724 * Pure: assumption method (an implicit finishing) now handles actual
   725 rules as well;
   726 
   727 * Pure: improved 'obtain' --- moved to Pure, insert "that" into
   728 initial goal, declare "that" only as Pure intro (only for single
   729 steps); the "that" rule assumption may now be involved in implicit
   730 finishing, thus ".." becomes a feasible for trivial obtains;
   731 
   732 * Pure: default proof step now includes 'intro_classes'; thus trivial
   733 instance proofs may be performed by "..";
   734 
   735 * Pure: ?thesis / ?this / "..." now work for pure meta-level
   736 statements as well;
   737 
   738 * Pure: more robust selection of calculational rules;
   739 
   740 * Pure: the builtin notion of 'finished' goal now includes the ==-refl
   741 rule (as well as the assumption rule);
   742 
   743 * Pure: 'thm_deps' command visualizes dependencies of theorems and
   744 lemmas, using the graph browser tool;
   745 
   746 * Pure: predict failure of "show" in interactive mode;
   747 
   748 * Pure: 'thms_containing' now takes actual terms as arguments;
   749 
   750 * HOL: improved method 'induct' --- now handles non-atomic goals
   751 (potential INCOMPATIBILITY); tuned error handling;
   752 
   753 * HOL: cases and induct rules now provide explicit hints about the
   754 number of facts to be consumed (0 for "type" and 1 for "set" rules);
   755 any remaining facts are inserted into the goal verbatim;
   756 
   757 * HOL: local contexts (aka cases) may now contain term bindings as
   758 well; the 'cases' and 'induct' methods new provide a ?case binding for
   759 the result to be shown in each case;
   760 
   761 * HOL: added 'recdef_tc' command;
   762 
   763 * isatool convert assists in eliminating legacy ML scripts;
   764 
   765 
   766 *** HOL ***
   767 
   768 * HOL/Library: a collection of generic theories to be used together
   769 with main HOL; the theory loader path already includes this directory
   770 by default; the following existing theories have been moved here:
   771 HOL/Induct/Multiset, HOL/Induct/Acc (as Accessible_Part), HOL/While
   772 (as While_Combinator), HOL/Lex/Prefix (as List_Prefix);
   773 
   774 * HOL/Unix: "Some aspects of Unix file-system security", a typical
   775 modelling and verification task performed in Isabelle/HOL +
   776 Isabelle/Isar + Isabelle document preparation (by Markus Wenzel).
   777 
   778 * HOL/Algebra: special summation operator SUM no longer exists, it has
   779 been replaced by setsum; infix 'assoc' now has priority 50 (like
   780 'dvd'); axiom 'one_not_zero' has been moved from axclass 'ring' to
   781 'domain', this makes the theory consistent with mathematical
   782 literature;
   783 
   784 * HOL basics: added overloaded operations "inverse" and "divide"
   785 (infix "/"), syntax for generic "abs" operation, generic summation
   786 operator \<Sum>;
   787 
   788 * HOL/typedef: simplified package, provide more useful rules (see also
   789 HOL/subset.thy);
   790 
   791 * HOL/datatype: induction rule for arbitrarily branching datatypes is
   792 now expressed as a proper nested rule (old-style tactic scripts may
   793 require atomize_strip_tac to cope with non-atomic premises);
   794 
   795 * HOL: renamed theory "Prod" to "Product_Type", renamed "split" rule
   796 to "split_conv" (old name still available for compatibility);
   797 
   798 * HOL: improved concrete syntax for strings (e.g. allows translation
   799 rules with string literals);
   800 
   801 * HOL-Real-Hyperreal: this extends HOL-Real with the hyperreals
   802  and Fleuriot's mechanization of analysis, including the transcendental
   803  functions for the reals;
   804 
   805 * HOL/Real, HOL/Hyperreal: improved arithmetic simplification;
   806 
   807 
   808 *** CTT ***
   809 
   810 * CTT: x-symbol support for Pi, Sigma, -->, : (membership); note that
   811 "lam" is displayed as TWO lambda-symbols
   812 
   813 * CTT: theory Main now available, containing everything (that is, Bool
   814 and Arith);
   815 
   816 
   817 *** General ***
   818 
   819 * Pure: the Simplifier has been implemented properly as a derived rule
   820 outside of the actual kernel (at last!); the overall performance
   821 penalty in practical applications is about 50%, while reliability of
   822 the Isabelle inference kernel has been greatly improved;
   823 
   824 * print modes "brackets" and "no_brackets" control output of nested =>
   825 (types) and ==> (props); the default behaviour is "brackets";
   826 
   827 * Provers: fast_tac (and friends) now handle actual object-logic rules
   828 as assumptions as well;
   829 
   830 * system: support Poly/ML 4.0;
   831 
   832 * system: isatool install handles KDE version 1 or 2;
   833 
   834 
   835 
   836 New in Isabelle99-1 (October 2000)
   837 ----------------------------------
   838 
   839 *** Overview of INCOMPATIBILITIES ***
   840 
   841 * HOL: simplification of natural numbers is much changed; to partly
   842 recover the old behaviour (e.g. to prevent n+n rewriting to #2*n)
   843 issue the following ML commands:
   844 
   845   Delsimprocs Nat_Numeral_Simprocs.cancel_numerals;
   846   Delsimprocs [Nat_Numeral_Simprocs.combine_numerals];
   847 
   848 * HOL: simplification no longer dives into case-expressions; this is
   849 controlled by "t.weak_case_cong" for each datatype t;
   850 
   851 * HOL: nat_less_induct renamed to less_induct;
   852 
   853 * HOL: systematic renaming of the SOME (Eps) rules, may use isatool
   854 fixsome to patch .thy and .ML sources automatically;
   855 
   856   select_equality  -> some_equality
   857   select_eq_Ex     -> some_eq_ex
   858   selectI2EX       -> someI2_ex
   859   selectI2         -> someI2
   860   selectI          -> someI
   861   select1_equality -> some1_equality
   862   Eps_sym_eq       -> some_sym_eq_trivial
   863   Eps_eq           -> some_eq_trivial
   864 
   865 * HOL: exhaust_tac on datatypes superceded by new generic case_tac;
   866 
   867 * HOL: removed obsolete theorem binding expand_if (refer to split_if
   868 instead);
   869 
   870 * HOL: the recursion equations generated by 'recdef' are now called
   871 f.simps instead of f.rules;
   872 
   873 * HOL: qed_spec_mp now also handles bounded ALL as well;
   874 
   875 * HOL: 0 is now overloaded, so the type constraint ":: nat" may
   876 sometimes be needed;
   877 
   878 * HOL: the constant for "f``x" is now "image" rather than "op ``";
   879 
   880 * HOL: the constant for "f-``x" is now "vimage" rather than "op -``";
   881 
   882 * HOL: the disjoint sum is now "<+>" instead of "Plus"; the cartesian
   883 product is now "<*>" instead of "Times"; the lexicographic product is
   884 now "<*lex*>" instead of "**";
   885 
   886 * HOL: theory Sexp is now in HOL/Induct examples (it used to be part
   887 of main HOL, but was unused); better use HOL's datatype package;
   888 
   889 * HOL: removed "symbols" syntax for constant "override" of theory Map;
   890 the old syntax may be recovered as follows:
   891 
   892   syntax (symbols)
   893     override  :: "('a ~=> 'b) => ('a ~=> 'b) => ('a ~=> 'b)"
   894       (infixl "\\<oplus>" 100)
   895 
   896 * HOL/Real: "rabs" replaced by overloaded "abs" function;
   897 
   898 * HOL/ML: even fewer consts are declared as global (see theories Ord,
   899 Lfp, Gfp, WF); this only affects ML packages that refer to const names
   900 internally;
   901 
   902 * HOL and ZF: syntax for quotienting wrt an equivalence relation
   903 changed from A/r to A//r;
   904 
   905 * ZF: new treatment of arithmetic (nat & int) may break some old
   906 proofs;
   907 
   908 * Isar: renamed some attributes (RS -> THEN, simplify -> simplified,
   909 rulify -> rule_format, elimify -> elim_format, ...);
   910 
   911 * Isar/Provers: intro/elim/dest attributes changed; renamed
   912 intro/intro!/intro!! flags to intro!/intro/intro? (in most cases, one
   913 should have to change intro!! to intro? only); replaced "delrule" by
   914 "rule del";
   915 
   916 * Isar/HOL: renamed "intrs" to "intros" in inductive definitions;
   917 
   918 * Provers: strengthened force_tac by using new first_best_tac;
   919 
   920 * LaTeX document preparation: several changes of isabelle.sty (see
   921 lib/texinputs);
   922 
   923 
   924 *** Document preparation ***
   925 
   926 * formal comments (text blocks etc.) in new-style theories may now
   927 contain antiquotations of thm/prop/term/typ/text to be presented
   928 according to latex print mode; concrete syntax is like this:
   929 @{term[show_types] "f(x) = a + x"};
   930 
   931 * isatool mkdir provides easy setup of Isabelle session directories,
   932 including proper document sources;
   933 
   934 * generated LaTeX sources are now deleted after successful run
   935 (isatool document -c); may retain a copy somewhere else via -D option
   936 of isatool usedir;
   937 
   938 * isatool usedir -D now lets isatool latex -o sty update the Isabelle
   939 style files, achieving self-contained LaTeX sources and simplifying
   940 LaTeX debugging;
   941 
   942 * old-style theories now produce (crude) LaTeX output as well;
   943 
   944 * browser info session directories are now self-contained (may be put
   945 on WWW server seperately); improved graphs of nested sessions; removed
   946 graph for 'all sessions';
   947 
   948 * several improvements in isabelle style files; \isabellestyle{it}
   949 produces fake math mode output; \isamarkupheader is now \section by
   950 default; see lib/texinputs/isabelle.sty etc.;
   951 
   952 
   953 *** Isar ***
   954 
   955 * Isar/Pure: local results and corresponding term bindings are now
   956 subject to Hindley-Milner polymorphism (similar to ML); this
   957 accommodates incremental type-inference very nicely;
   958 
   959 * Isar/Pure: new derived language element 'obtain' supports
   960 generalized existence reasoning;
   961 
   962 * Isar/Pure: new calculational elements 'moreover' and 'ultimately'
   963 support accumulation of results, without applying any rules yet;
   964 useful to collect intermediate results without explicit name
   965 references, and for use with transitivity rules with more than 2
   966 premises;
   967 
   968 * Isar/Pure: scalable support for case-analysis type proofs: new
   969 'case' language element refers to local contexts symbolically, as
   970 produced by certain proof methods; internally, case names are attached
   971 to theorems as "tags";
   972 
   973 * Isar/Pure: theory command 'hide' removes declarations from
   974 class/type/const name spaces;
   975 
   976 * Isar/Pure: theory command 'defs' supports option "(overloaded)" to
   977 indicate potential overloading;
   978 
   979 * Isar/Pure: changed syntax of local blocks from {{ }} to { };
   980 
   981 * Isar/Pure: syntax of sorts made 'inner', i.e. have to write
   982 "{a,b,c}" instead of {a,b,c};
   983 
   984 * Isar/Pure now provides its own version of intro/elim/dest
   985 attributes; useful for building new logics, but beware of confusion
   986 with the version in Provers/classical;
   987 
   988 * Isar/Pure: the local context of (non-atomic) goals is provided via
   989 case name 'antecedent';
   990 
   991 * Isar/Pure: removed obsolete 'transfer' attribute (transfer of thms
   992 to the current context is now done automatically);
   993 
   994 * Isar/Pure: theory command 'method_setup' provides a simple interface
   995 for definining proof methods in ML;
   996 
   997 * Isar/Provers: intro/elim/dest attributes changed; renamed
   998 intro/intro!/intro!! flags to intro!/intro/intro? (INCOMPATIBILITY, in
   999 most cases, one should have to change intro!! to intro? only);
  1000 replaced "delrule" by "rule del";
  1001 
  1002 * Isar/Provers: new 'hypsubst' method, plain 'subst' method and
  1003 'symmetric' attribute (the latter supercedes [RS sym]);
  1004 
  1005 * Isar/Provers: splitter support (via 'split' attribute and 'simp'
  1006 method modifier); 'simp' method: 'only:' modifier removes loopers as
  1007 well (including splits);
  1008 
  1009 * Isar/Provers: Simplifier and Classical methods now support all kind
  1010 of modifiers used in the past, including 'cong', 'iff', etc.
  1011 
  1012 * Isar/Provers: added 'fastsimp' and 'clarsimp' methods (combination
  1013 of Simplifier and Classical reasoner);
  1014 
  1015 * Isar/HOL: new proof method 'cases' and improved version of 'induct'
  1016 now support named cases; major packages (inductive, datatype, primrec,
  1017 recdef) support case names and properly name parameters;
  1018 
  1019 * Isar/HOL: new transitivity rules for substitution in inequalities --
  1020 monotonicity conditions are extracted to be proven at end of
  1021 calculations;
  1022 
  1023 * Isar/HOL: removed 'case_split' thm binding, should use 'cases' proof
  1024 method anyway;
  1025 
  1026 * Isar/HOL: removed old expand_if = split_if; theorems if_splits =
  1027 split_if split_if_asm; datatype package provides theorems foo.splits =
  1028 foo.split foo.split_asm for each datatype;
  1029 
  1030 * Isar/HOL: tuned inductive package, rename "intrs" to "intros"
  1031 (potential INCOMPATIBILITY), emulation of mk_cases feature for proof
  1032 scripts: new 'inductive_cases' command and 'ind_cases' method; (Note:
  1033 use "(cases (simplified))" method in proper proof texts);
  1034 
  1035 * Isar/HOL: added global 'arith_split' attribute for 'arith' method;
  1036 
  1037 * Isar: names of theorems etc. may be natural numbers as well;
  1038 
  1039 * Isar: 'pr' command: optional arguments for goals_limit and
  1040 ProofContext.prems_limit; no longer prints theory contexts, but only
  1041 proof states;
  1042 
  1043 * Isar: diagnostic commands 'pr', 'thm', 'prop', 'term', 'typ' admit
  1044 additional print modes to be specified; e.g. "pr(latex)" will print
  1045 proof state according to the Isabelle LaTeX style;
  1046 
  1047 * Isar: improved support for emulating tactic scripts, including proof
  1048 methods 'rule_tac' etc., 'cut_tac', 'thin_tac', 'subgoal_tac',
  1049 'rename_tac', 'rotate_tac', 'tactic', and 'case_tac' / 'induct_tac'
  1050 (for HOL datatypes);
  1051 
  1052 * Isar: simplified (more robust) goal selection of proof methods: 1st
  1053 goal, all goals, or explicit goal specifier (tactic emulation); thus
  1054 'proof method scripts' have to be in depth-first order;
  1055 
  1056 * Isar: tuned 'let' syntax: replaced 'as' keyword by 'and';
  1057 
  1058 * Isar: removed 'help' command, which hasn't been too helpful anyway;
  1059 should instead use individual commands for printing items
  1060 (print_commands, print_methods etc.);
  1061 
  1062 * Isar: added 'nothing' --- the empty list of theorems;
  1063 
  1064 
  1065 *** HOL ***
  1066 
  1067 * HOL/MicroJava: formalization of a fragment of Java, together with a
  1068 corresponding virtual machine and a specification of its bytecode
  1069 verifier and a lightweight bytecode verifier, including proofs of
  1070 type-safety; by Gerwin Klein, Tobias Nipkow, David von Oheimb, and
  1071 Cornelia Pusch (see also the homepage of project Bali at
  1072 http://isabelle.in.tum.de/Bali/);
  1073 
  1074 * HOL/Algebra: new theory of rings and univariate polynomials, by
  1075 Clemens Ballarin;
  1076 
  1077 * HOL/NumberTheory: fundamental Theorem of Arithmetic, Chinese
  1078 Remainder Theorem, Fermat/Euler Theorem, Wilson's Theorem, by Thomas M
  1079 Rasmussen;
  1080 
  1081 * HOL/Lattice: fundamental concepts of lattice theory and order
  1082 structures, including duals, properties of bounds versus algebraic
  1083 laws, lattice operations versus set-theoretic ones, the Knaster-Tarski
  1084 Theorem for complete lattices etc.; may also serve as a demonstration
  1085 for abstract algebraic reasoning using axiomatic type classes, and
  1086 mathematics-style proof in Isabelle/Isar; by Markus Wenzel;
  1087 
  1088 * HOL/Prolog: a (bare-bones) implementation of Lambda-Prolog, by David
  1089 von Oheimb;
  1090 
  1091 * HOL/IMPP: extension of IMP with local variables and mutually
  1092 recursive procedures, by David von Oheimb;
  1093 
  1094 * HOL/Lambda: converted into new-style theory and document;
  1095 
  1096 * HOL/ex/Multiquote: example of multiple nested quotations and
  1097 anti-quotations -- basically a generalized version of de-Bruijn
  1098 representation; very useful in avoiding lifting of operations;
  1099 
  1100 * HOL/record: added general record equality rule to simpset; fixed
  1101 select-update simplification procedure to handle extended records as
  1102 well; admit "r" as field name;
  1103 
  1104 * HOL: 0 is now overloaded over the new sort "zero", allowing its use with
  1105 other numeric types and also as the identity of groups, rings, etc.;
  1106 
  1107 * HOL: new axclass plus_ac0 for addition with the AC-laws and 0 as identity.
  1108 Types nat and int belong to this axclass;
  1109 
  1110 * HOL: greatly improved simplification involving numerals of type nat, int, real:
  1111    (i + #8 + j) = Suc k simplifies to  #7 + (i + j) = k
  1112    i*j + k + j*#3*i     simplifies to  #4*(i*j) + k
  1113   two terms #m*u and #n*u are replaced by #(m+n)*u
  1114     (where #m, #n and u can implicitly be 1; this is simproc combine_numerals)
  1115   and the term/formula #m*u+x ~~ #n*u+y simplifies simplifies to #(m-n)+x ~~ y
  1116     or x ~~ #(n-m)+y, where ~~ is one of = < <= or - (simproc cancel_numerals);
  1117 
  1118 * HOL: meson_tac is available (previously in ex/meson.ML); it is a
  1119 powerful prover for predicate logic but knows nothing of clasets; see
  1120 ex/mesontest.ML and ex/mesontest2.ML for example applications;
  1121 
  1122 * HOL: new version of "case_tac" subsumes both boolean case split and
  1123 "exhaust_tac" on datatypes; INCOMPATIBILITY: exhaust_tac no longer
  1124 exists, may define val exhaust_tac = case_tac for ad-hoc portability;
  1125 
  1126 * HOL: simplification no longer dives into case-expressions: only the
  1127 selector expression is simplified, but not the remaining arms; to
  1128 enable full simplification of case-expressions for datatype t, you may
  1129 remove t.weak_case_cong from the simpset, either globally (Delcongs
  1130 [thm"t.weak_case_cong"];) or locally (delcongs [...]).
  1131 
  1132 * HOL/recdef: the recursion equations generated by 'recdef' for
  1133 function 'f' are now called f.simps instead of f.rules; if all
  1134 termination conditions are proved automatically, these simplification
  1135 rules are added to the simpset, as in primrec; rules may be named
  1136 individually as well, resulting in a separate list of theorems for
  1137 each equation;
  1138 
  1139 * HOL/While is a new theory that provides a while-combinator. It
  1140 permits the definition of tail-recursive functions without the
  1141 provision of a termination measure. The latter is necessary once the
  1142 invariant proof rule for while is applied.
  1143 
  1144 * HOL: new (overloaded) notation for the set of elements below/above
  1145 some element: {..u}, {..u(}, {l..}, {)l..}. See theory SetInterval.
  1146 
  1147 * HOL: theorems impI, allI, ballI bound as "strip";
  1148 
  1149 * HOL: new tactic induct_thm_tac: thm -> string -> int -> tactic
  1150 induct_tac th "x1 ... xn" expects th to have a conclusion of the form
  1151 P v1 ... vn and abbreviates res_inst_tac [("v1","x1"),...,("vn","xn")] th;
  1152 
  1153 * HOL/Real: "rabs" replaced by overloaded "abs" function;
  1154 
  1155 * HOL: theory Sexp now in HOL/Induct examples (it used to be part of
  1156 main HOL, but was unused);
  1157 
  1158 * HOL: fewer consts declared as global (e.g. have to refer to
  1159 "Lfp.lfp" instead of "lfp" internally; affects ML packages only);
  1160 
  1161 * HOL: tuned AST representation of nested pairs, avoiding bogus output
  1162 in case of overlap with user translations (e.g. judgements over
  1163 tuples); (note that the underlying logical represenation is still
  1164 bogus);
  1165 
  1166 
  1167 *** ZF ***
  1168 
  1169 * ZF: simplification automatically cancels common terms in arithmetic
  1170 expressions over nat and int;
  1171 
  1172 * ZF: new treatment of nat to minimize type-checking: all operators
  1173 coerce their operands to a natural number using the function natify,
  1174 making the algebraic laws unconditional;
  1175 
  1176 * ZF: as above, for int: operators coerce their operands to an integer
  1177 using the function intify;
  1178 
  1179 * ZF: the integer library now contains many of the usual laws for the
  1180 orderings, including $<=, and monotonicity laws for $+ and $*;
  1181 
  1182 * ZF: new example ZF/ex/NatSum to demonstrate integer arithmetic
  1183 simplification;
  1184 
  1185 * FOL and ZF: AddIffs now available, giving theorems of the form P<->Q
  1186 to the simplifier and classical reasoner simultaneously;
  1187 
  1188 
  1189 *** General ***
  1190 
  1191 * Provers: blast_tac now handles actual object-logic rules as
  1192 assumptions; note that auto_tac uses blast_tac internally as well;
  1193 
  1194 * Provers: new functions rulify/rulify_no_asm: thm -> thm for turning
  1195 outer -->/All/Ball into ==>/!!; qed_spec_mp now uses rulify_no_asm;
  1196 
  1197 * Provers: delrules now handles destruct rules as well (no longer need
  1198 explicit make_elim);
  1199 
  1200 * Provers: Blast_tac now warns of and ignores "weak elimination rules" e.g.
  1201   [| inj ?f;          ?f ?x = ?f ?y; ?x = ?y ==> ?W |] ==> ?W
  1202 use instead the strong form,
  1203   [| inj ?f; ~ ?W ==> ?f ?x = ?f ?y; ?x = ?y ==> ?W |] ==> ?W
  1204 in HOL, FOL and ZF the function cla_make_elim will create such rules
  1205 from destruct-rules;
  1206 
  1207 * Provers: Simplifier.easy_setup provides a fast path to basic
  1208 Simplifier setup for new object-logics;
  1209 
  1210 * Pure: AST translation rules no longer require constant head on LHS;
  1211 
  1212 * Pure: improved name spaces: ambiguous output is qualified; support
  1213 for hiding of names;
  1214 
  1215 * system: smart setup of canonical ML_HOME, ISABELLE_INTERFACE, and
  1216 XSYMBOL_HOME; no longer need to do manual configuration in most
  1217 situations;
  1218 
  1219 * system: compression of ML heaps images may now be controlled via -c
  1220 option of isabelle and isatool usedir (currently only observed by
  1221 Poly/ML);
  1222 
  1223 * system: isatool installfonts may handle X-Symbol fonts as well (very
  1224 useful for remote X11);
  1225 
  1226 * system: provide TAGS file for Isabelle sources;
  1227 
  1228 * ML: infix 'OF' is a version of 'MRS' with more appropriate argument
  1229 order;
  1230 
  1231 * ML: renamed flags Syntax.trace_norm_ast to Syntax.trace_ast; global
  1232 timing flag supersedes proof_timing and Toplevel.trace;
  1233 
  1234 * ML: new combinators |>> and |>>> for incremental transformations
  1235 with secondary results (e.g. certain theory extensions):
  1236 
  1237 * ML: PureThy.add_defs gets additional argument to indicate potential
  1238 overloading (usually false);
  1239 
  1240 * ML: PureThy.add_thms/add_axioms/add_defs now return theorems as
  1241 results;
  1242 
  1243 
  1244 
  1245 New in Isabelle99 (October 1999)
  1246 --------------------------------
  1247 
  1248 *** Overview of INCOMPATIBILITIES (see below for more details) ***
  1249 
  1250 * HOL: The THEN and ELSE parts of conditional expressions (if P then x else y)
  1251 are no longer simplified.  (This allows the simplifier to unfold recursive
  1252 functional programs.)  To restore the old behaviour, declare
  1253 
  1254     Delcongs [if_weak_cong];
  1255 
  1256 * HOL: Removed the obsolete syntax "Compl A"; use -A for set
  1257 complement;
  1258 
  1259 * HOL: the predicate "inj" is now defined by translation to "inj_on";
  1260 
  1261 * HOL/datatype: mutual_induct_tac no longer exists --
  1262   use induct_tac "x_1 ... x_n" instead of mutual_induct_tac ["x_1", ..., "x_n"]
  1263 
  1264 * HOL/typedef: fixed type inference for representing set; type
  1265 arguments now have to occur explicitly on the rhs as type constraints;
  1266 
  1267 * ZF: The con_defs part of an inductive definition may no longer refer
  1268 to constants declared in the same theory;
  1269 
  1270 * HOL, ZF: the function mk_cases, generated by the inductive
  1271 definition package, has lost an argument.  To simplify its result, it
  1272 uses the default simpset instead of a supplied list of theorems.
  1273 
  1274 * HOL/List: the constructors of type list are now Nil and Cons;
  1275 
  1276 * Simplifier: the type of the infix ML functions
  1277         setSSolver addSSolver setSolver addSolver
  1278 is now  simpset * solver -> simpset  where `solver' is a new abstract type
  1279 for packaging solvers. A solver is created via
  1280         mk_solver: string -> (thm list -> int -> tactic) -> solver
  1281 where the string argument is only a comment.
  1282 
  1283 
  1284 *** Proof tools ***
  1285 
  1286 * Provers/Arith/fast_lin_arith.ML contains a functor for creating a
  1287 decision procedure for linear arithmetic. Currently it is used for
  1288 types `nat', `int', and `real' in HOL (see below); it can, should and
  1289 will be instantiated for other types and logics as well.
  1290 
  1291 * The simplifier now accepts rewrite rules with flexible heads, eg
  1292      hom ?f ==> ?f(?x+?y) = ?f ?x + ?f ?y
  1293   They are applied like any rule with a non-pattern lhs, i.e. by first-order
  1294   matching.
  1295 
  1296 
  1297 *** General ***
  1298 
  1299 * New Isabelle/Isar subsystem provides an alternative to traditional
  1300 tactical theorem proving; together with the ProofGeneral/isar user
  1301 interface it offers an interactive environment for developing human
  1302 readable proof documents (Isar == Intelligible semi-automated
  1303 reasoning); for further information see isatool doc isar-ref,
  1304 src/HOL/Isar_examples and http://isabelle.in.tum.de/Isar/
  1305 
  1306 * improved and simplified presentation of theories: better HTML markup
  1307 (including colors), graph views in several sizes; isatool usedir now
  1308 provides a proper interface for user theories (via -P option); actual
  1309 document preparation based on (PDF)LaTeX is available as well (for
  1310 new-style theories only); see isatool doc system for more information;
  1311 
  1312 * native support for Proof General, both for classic Isabelle and
  1313 Isabelle/Isar;
  1314 
  1315 * ML function thm_deps visualizes dependencies of theorems and lemmas,
  1316 using the graph browser tool;
  1317 
  1318 * Isabelle manuals now also available as PDF;
  1319 
  1320 * theory loader rewritten from scratch (may not be fully
  1321 bug-compatible); old loadpath variable has been replaced by show_path,
  1322 add_path, del_path, reset_path functions; new operations such as
  1323 update_thy, touch_thy, remove_thy, use/update_thy_only (see also
  1324 isatool doc ref);
  1325 
  1326 * improved isatool install: option -k creates KDE application icon,
  1327 option -p DIR installs standalone binaries;
  1328 
  1329 * added ML_PLATFORM setting (useful for cross-platform installations);
  1330 more robust handling of platform specific ML images for SML/NJ;
  1331 
  1332 * the settings environment is now statically scoped, i.e. it is never
  1333 created again in sub-processes invoked from isabelle, isatool, or
  1334 Isabelle;
  1335 
  1336 * path element specification '~~' refers to '$ISABELLE_HOME';
  1337 
  1338 * in locales, the "assumes" and "defines" parts may be omitted if
  1339 empty;
  1340 
  1341 * new print_mode "xsymbols" for extended symbol support (e.g. genuine
  1342 long arrows);
  1343 
  1344 * new print_mode "HTML";
  1345 
  1346 * new flag show_tags controls display of tags of theorems (which are
  1347 basically just comments that may be attached by some tools);
  1348 
  1349 * Isamode 2.6 requires patch to accomodate change of Isabelle font
  1350 mode and goal output format:
  1351 
  1352 diff -r Isamode-2.6/elisp/isa-load.el Isamode/elisp/isa-load.el
  1353 244c244
  1354 <       (list (isa-getenv "ISABELLE") "-msymbols" logic-name)
  1355 ---
  1356 >       (list (isa-getenv "ISABELLE") "-misabelle_font" "-msymbols" logic-name)
  1357 diff -r Isabelle-2.6/elisp/isa-proofstate.el Isamode/elisp/isa-proofstate.el
  1358 181c181
  1359 < (defconst proofstate-proofstart-regexp "^Level [0-9]+$"
  1360 ---
  1361 > (defconst proofstate-proofstart-regexp "^Level [0-9]+"
  1362 
  1363 * function bind_thms stores lists of theorems (cf. bind_thm);
  1364 
  1365 * new shorthand tactics ftac, eatac, datac, fatac;
  1366 
  1367 * qed (and friends) now accept "" as result name; in that case the
  1368 theorem is not stored, but proper checks and presentation of the
  1369 result still apply;
  1370 
  1371 * theorem database now also indexes constants "Trueprop", "all",
  1372 "==>", "=="; thus thms_containing, findI etc. may retrieve more rules;
  1373 
  1374 
  1375 *** HOL ***
  1376 
  1377 ** HOL arithmetic **
  1378 
  1379 * There are now decision procedures for linear arithmetic over nat and
  1380 int:
  1381 
  1382 1. arith_tac copes with arbitrary formulae involving `=', `<', `<=',
  1383 `+', `-', `Suc', `min', `max' and numerical constants; other subterms
  1384 are treated as atomic; subformulae not involving type `nat' or `int'
  1385 are ignored; quantified subformulae are ignored unless they are
  1386 positive universal or negative existential. The tactic has to be
  1387 invoked by hand and can be a little bit slow. In particular, the
  1388 running time is exponential in the number of occurrences of `min' and
  1389 `max', and `-' on `nat'.
  1390 
  1391 2. fast_arith_tac is a cut-down version of arith_tac: it only takes
  1392 (negated) (in)equalities among the premises and the conclusion into
  1393 account (i.e. no compound formulae) and does not know about `min' and
  1394 `max', and `-' on `nat'. It is fast and is used automatically by the
  1395 simplifier.
  1396 
  1397 NB: At the moment, these decision procedures do not cope with mixed
  1398 nat/int formulae where the two parts interact, such as `m < n ==>
  1399 int(m) < int(n)'.
  1400 
  1401 * HOL/Numeral provides a generic theory of numerals (encoded
  1402 efficiently as bit strings); setup for types nat/int/real is in place;
  1403 INCOMPATIBILITY: since numeral syntax is now polymorphic, rather than
  1404 int, existing theories and proof scripts may require a few additional
  1405 type constraints;
  1406 
  1407 * integer division and remainder can now be performed on constant
  1408 arguments;
  1409 
  1410 * many properties of integer multiplication, division and remainder
  1411 are now available;
  1412 
  1413 * An interface to the Stanford Validity Checker (SVC) is available through the
  1414 tactic svc_tac.  Propositional tautologies and theorems of linear arithmetic
  1415 are proved automatically.  SVC must be installed separately, and its results
  1416 must be TAKEN ON TRUST (Isabelle does not check the proofs, but tags any
  1417 invocation of the underlying oracle).  For SVC see
  1418   http://verify.stanford.edu/SVC
  1419 
  1420 * IsaMakefile: the HOL-Real target now builds an actual image;
  1421 
  1422 
  1423 ** HOL misc **
  1424 
  1425 * HOL/Real/HahnBanach: the Hahn-Banach theorem for real vector spaces
  1426 (in Isabelle/Isar) -- by Gertrud Bauer;
  1427 
  1428 * HOL/BCV: generic model of bytecode verification, i.e. data-flow
  1429 analysis for assembly languages with subtypes;
  1430 
  1431 * HOL/TLA (Lamport's Temporal Logic of Actions): major reorganization
  1432 -- avoids syntactic ambiguities and treats state, transition, and
  1433 temporal levels more uniformly; introduces INCOMPATIBILITIES due to
  1434 changed syntax and (many) tactics;
  1435 
  1436 * HOL/inductive: Now also handles more general introduction rules such
  1437   as "ALL y. (y, x) : r --> y : acc r ==> x : acc r"; monotonicity
  1438   theorems are now maintained within the theory (maintained via the
  1439   "mono" attribute);
  1440 
  1441 * HOL/datatype: Now also handles arbitrarily branching datatypes
  1442   (using function types) such as
  1443 
  1444   datatype 'a tree = Atom 'a | Branch "nat => 'a tree"
  1445 
  1446 * HOL/record: record_simproc (part of the default simpset) takes care
  1447 of selectors applied to updated records; record_split_tac is no longer
  1448 part of the default claset; update_defs may now be removed from the
  1449 simpset in many cases; COMPATIBILITY: old behavior achieved by
  1450 
  1451   claset_ref () := claset() addSWrapper record_split_wrapper;
  1452   Delsimprocs [record_simproc]
  1453 
  1454 * HOL/typedef: fixed type inference for representing set; type
  1455 arguments now have to occur explicitly on the rhs as type constraints;
  1456 
  1457 * HOL/recdef (TFL): 'congs' syntax now expects comma separated list of theorem
  1458 names rather than an ML expression;
  1459 
  1460 * HOL/defer_recdef (TFL): like recdef but the well-founded relation can be
  1461 supplied later.  Program schemes can be defined, such as
  1462     "While B C s = (if B s then While B C (C s) else s)"
  1463 where the well-founded relation can be chosen after B and C have been given.
  1464 
  1465 * HOL/List: the constructors of type list are now Nil and Cons;
  1466 INCOMPATIBILITY: while [] and infix # syntax is still there, of
  1467 course, ML tools referring to List.list.op # etc. have to be adapted;
  1468 
  1469 * HOL_quantifiers flag superseded by "HOL" print mode, which is
  1470 disabled by default; run isabelle with option -m HOL to get back to
  1471 the original Gordon/HOL-style output;
  1472 
  1473 * HOL/Ord.thy: new bounded quantifier syntax (input only): ALL x<y. P,
  1474 ALL x<=y. P, EX x<y. P, EX x<=y. P;
  1475 
  1476 * HOL basic syntax simplified (more orthogonal): all variants of
  1477 All/Ex now support plain / symbolic / HOL notation; plain syntax for
  1478 Eps operator is provided as well: "SOME x. P[x]";
  1479 
  1480 * HOL/Sum.thy: sum_case has been moved to HOL/Datatype;
  1481 
  1482 * HOL/Univ.thy: infix syntax <*>, <+>, <**>, <+> eliminated and made
  1483 thus available for user theories;
  1484 
  1485 * HOLCF/IOA/Sequents: renamed 'Cons' to 'Consq' to avoid clash with
  1486 HOL/List; hardly an INCOMPATIBILITY since '>>' syntax is used all the
  1487 time;
  1488 
  1489 * HOL: new tactic smp_tac: int -> int -> tactic, which applies spec
  1490 several times and then mp;
  1491 
  1492 
  1493 *** LK ***
  1494 
  1495 * the notation <<...>> is now available as a notation for sequences of
  1496 formulas;
  1497 
  1498 * the simplifier is now installed
  1499 
  1500 * the axiom system has been generalized (thanks to Soren Heilmann)
  1501 
  1502 * the classical reasoner now has a default rule database
  1503 
  1504 
  1505 *** ZF ***
  1506 
  1507 * new primrec section allows primitive recursive functions to be given
  1508 directly (as in HOL) over datatypes and the natural numbers;
  1509 
  1510 * new tactics induct_tac and exhaust_tac for induction (or case
  1511 analysis) over datatypes and the natural numbers;
  1512 
  1513 * the datatype declaration of type T now defines the recursor T_rec;
  1514 
  1515 * simplification automatically does freeness reasoning for datatype
  1516 constructors;
  1517 
  1518 * automatic type-inference, with AddTCs command to insert new
  1519 type-checking rules;
  1520 
  1521 * datatype introduction rules are now added as Safe Introduction rules
  1522 to the claset;
  1523 
  1524 * the syntax "if P then x else y" is now available in addition to
  1525 if(P,x,y);
  1526 
  1527 
  1528 *** Internal programming interfaces ***
  1529 
  1530 * tuned simplifier trace output; new flag debug_simp;
  1531 
  1532 * structures Vartab / Termtab (instances of TableFun) offer efficient
  1533 tables indexed by indexname_ord / term_ord (compatible with aconv);
  1534 
  1535 * AxClass.axclass_tac lost the theory argument;
  1536 
  1537 * tuned current_goals_markers semantics: begin / end goal avoids
  1538 printing empty lines;
  1539 
  1540 * removed prs and prs_fn hook, which was broken because it did not
  1541 include \n in its semantics, forcing writeln to add one
  1542 uncoditionally; replaced prs_fn by writeln_fn; consider std_output:
  1543 string -> unit if you really want to output text without newline;
  1544 
  1545 * Symbol.output subject to print mode; INCOMPATIBILITY: defaults to
  1546 plain output, interface builders may have to enable 'isabelle_font'
  1547 mode to get Isabelle font glyphs as before;
  1548 
  1549 * refined token_translation interface; INCOMPATIBILITY: output length
  1550 now of type real instead of int;
  1551 
  1552 * theory loader actions may be traced via new ThyInfo.add_hook
  1553 interface (see src/Pure/Thy/thy_info.ML); example application: keep
  1554 your own database of information attached to *whole* theories -- as
  1555 opposed to intra-theory data slots offered via TheoryDataFun;
  1556 
  1557 * proper handling of dangling sort hypotheses (at last!);
  1558 Thm.strip_shyps and Drule.strip_shyps_warning take care of removing
  1559 extra sort hypotheses that can be witnessed from the type signature;
  1560 the force_strip_shyps flag is gone, any remaining shyps are simply
  1561 left in the theorem (with a warning issued by strip_shyps_warning);
  1562 
  1563 
  1564 
  1565 New in Isabelle98-1 (October 1998)
  1566 ----------------------------------
  1567 
  1568 *** Overview of INCOMPATIBILITIES (see below for more details) ***
  1569 
  1570 * several changes of automated proof tools;
  1571 
  1572 * HOL: major changes to the inductive and datatype packages, including
  1573 some minor incompatibilities of theory syntax;
  1574 
  1575 * HOL: renamed r^-1 to 'converse' from 'inverse'; 'inj_onto' is now
  1576 called `inj_on';
  1577 
  1578 * HOL: removed duplicate thms in Arith:
  1579   less_imp_add_less  should be replaced by  trans_less_add1
  1580   le_imp_add_le      should be replaced by  trans_le_add1
  1581 
  1582 * HOL: unary minus is now overloaded (new type constraints may be
  1583 required);
  1584 
  1585 * HOL and ZF: unary minus for integers is now #- instead of #~.  In
  1586 ZF, expressions such as n#-1 must be changed to n#- 1, since #-1 is
  1587 now taken as an integer constant.
  1588 
  1589 * Pure: ML function 'theory_of' renamed to 'theory';
  1590 
  1591 
  1592 *** Proof tools ***
  1593 
  1594 * Simplifier:
  1595   1. Asm_full_simp_tac is now more aggressive.
  1596      1. It will sometimes reorient premises if that increases their power to
  1597         simplify.
  1598      2. It does no longer proceed strictly from left to right but may also
  1599         rotate premises to achieve further simplification.
  1600      For compatibility reasons there is now Asm_lr_simp_tac which is like the
  1601      old Asm_full_simp_tac in that it does not rotate premises.
  1602   2. The simplifier now knows a little bit about nat-arithmetic.
  1603 
  1604 * Classical reasoner: wrapper mechanism for the classical reasoner now
  1605 allows for selected deletion of wrappers, by introduction of names for
  1606 wrapper functionals.  This implies that addbefore, addSbefore,
  1607 addaltern, and addSaltern now take a pair (name, tactic) as argument,
  1608 and that adding two tactics with the same name overwrites the first
  1609 one (emitting a warning).
  1610   type wrapper = (int -> tactic) -> (int -> tactic)
  1611   setWrapper, setSWrapper, compWrapper and compSWrapper are replaced by
  1612   addWrapper, addSWrapper: claset * (string * wrapper) -> claset
  1613   delWrapper, delSWrapper: claset *  string            -> claset
  1614   getWrapper is renamed to appWrappers, getSWrapper to appSWrappers;
  1615 
  1616 * Classical reasoner: addbefore/addSbefore now have APPEND/ORELSE
  1617 semantics; addbefore now affects only the unsafe part of step_tac
  1618 etc.; this affects addss/auto_tac/force_tac, so EXISTING PROOFS MAY
  1619 FAIL, but proofs should be fixable easily, e.g. by replacing Auto_tac
  1620 by Force_tac;
  1621 
  1622 * Classical reasoner: setwrapper to setWrapper and compwrapper to
  1623 compWrapper; added safe wrapper (and access functions for it);
  1624 
  1625 * HOL/split_all_tac is now much faster and fails if there is nothing
  1626 to split.  Some EXISTING PROOFS MAY REQUIRE ADAPTION because the order
  1627 and the names of the automatically generated variables have changed.
  1628 split_all_tac has moved within claset() from unsafe wrappers to safe
  1629 wrappers, which means that !!-bound variables are split much more
  1630 aggressively, and safe_tac and clarify_tac now split such variables.
  1631 If this splitting is not appropriate, use delSWrapper "split_all_tac".
  1632 Note: the same holds for record_split_tac, which does the job of
  1633 split_all_tac for record fields.
  1634 
  1635 * HOL/Simplifier: Rewrite rules for case distinctions can now be added
  1636 permanently to the default simpset using Addsplits just like
  1637 Addsimps. They can be removed via Delsplits just like
  1638 Delsimps. Lower-case versions are also available.
  1639 
  1640 * HOL/Simplifier: The rule split_if is now part of the default
  1641 simpset. This means that the simplifier will eliminate all occurrences
  1642 of if-then-else in the conclusion of a goal. To prevent this, you can
  1643 either remove split_if completely from the default simpset by
  1644 `Delsplits [split_if]' or remove it in a specific call of the
  1645 simplifier using `... delsplits [split_if]'.  You can also add/delete
  1646 other case splitting rules to/from the default simpset: every datatype
  1647 generates suitable rules `split_t_case' and `split_t_case_asm' (where
  1648 t is the name of the datatype).
  1649 
  1650 * Classical reasoner / Simplifier combination: new force_tac (and
  1651 derivatives Force_tac, force) combines rewriting and classical
  1652 reasoning (and whatever other tools) similarly to auto_tac, but is
  1653 aimed to solve the given subgoal completely.
  1654 
  1655 
  1656 *** General ***
  1657 
  1658 * new top-level commands `Goal' and `Goalw' that improve upon `goal'
  1659 and `goalw': the theory is no longer needed as an explicit argument -
  1660 the current theory context is used; assumptions are no longer returned
  1661 at the ML-level unless one of them starts with ==> or !!; it is
  1662 recommended to convert to these new commands using isatool fixgoal
  1663 (backup your sources first!);
  1664 
  1665 * new top-level commands 'thm' and 'thms' for retrieving theorems from
  1666 the current theory context, and 'theory' to lookup stored theories;
  1667 
  1668 * new theory section 'locale' for declaring constants, assumptions and
  1669 definitions that have local scope;
  1670 
  1671 * new theory section 'nonterminals' for purely syntactic types;
  1672 
  1673 * new theory section 'setup' for generic ML setup functions
  1674 (e.g. package initialization);
  1675 
  1676 * the distribution now includes Isabelle icons: see
  1677 lib/logo/isabelle-{small,tiny}.xpm;
  1678 
  1679 * isatool install - install binaries with absolute references to
  1680 ISABELLE_HOME/bin;
  1681 
  1682 * isatool logo -- create instances of the Isabelle logo (as EPS);
  1683 
  1684 * print mode 'emacs' reserved for Isamode;
  1685 
  1686 * support multiple print (ast) translations per constant name;
  1687 
  1688 * theorems involving oracles are now printed with a suffixed [!];
  1689 
  1690 
  1691 *** HOL ***
  1692 
  1693 * there is now a tutorial on Isabelle/HOL (do 'isatool doc tutorial');
  1694 
  1695 * HOL/inductive package reorganized and improved: now supports mutual
  1696 definitions such as
  1697 
  1698   inductive EVEN ODD
  1699     intrs
  1700       null "0 : EVEN"
  1701       oddI "n : EVEN ==> Suc n : ODD"
  1702       evenI "n : ODD ==> Suc n : EVEN"
  1703 
  1704 new theorem list "elims" contains an elimination rule for each of the
  1705 recursive sets; inductive definitions now handle disjunctive premises
  1706 correctly (also ZF);
  1707 
  1708 INCOMPATIBILITIES: requires Inductive as an ancestor; component
  1709 "mutual_induct" no longer exists - the induction rule is always
  1710 contained in "induct";
  1711 
  1712 
  1713 * HOL/datatype package re-implemented and greatly improved: now
  1714 supports mutually recursive datatypes such as
  1715 
  1716   datatype
  1717     'a aexp = IF_THEN_ELSE ('a bexp) ('a aexp) ('a aexp)
  1718             | SUM ('a aexp) ('a aexp)
  1719             | DIFF ('a aexp) ('a aexp)
  1720             | NUM 'a
  1721   and
  1722     'a bexp = LESS ('a aexp) ('a aexp)
  1723             | AND ('a bexp) ('a bexp)
  1724             | OR ('a bexp) ('a bexp)
  1725 
  1726 as well as indirectly recursive datatypes such as
  1727 
  1728   datatype
  1729     ('a, 'b) term = Var 'a
  1730                   | App 'b ((('a, 'b) term) list)
  1731 
  1732 The new tactic  mutual_induct_tac [<var_1>, ..., <var_n>] i  performs
  1733 induction on mutually / indirectly recursive datatypes.
  1734 
  1735 Primrec equations are now stored in theory and can be accessed via
  1736 <function_name>.simps.
  1737 
  1738 INCOMPATIBILITIES:
  1739 
  1740   - Theories using datatypes must now have theory Datatype as an
  1741     ancestor.
  1742   - The specific <typename>.induct_tac no longer exists - use the
  1743     generic induct_tac instead.
  1744   - natE has been renamed to nat.exhaust - use exhaust_tac
  1745     instead of res_inst_tac ... natE. Note that the variable
  1746     names in nat.exhaust differ from the names in natE, this
  1747     may cause some "fragile" proofs to fail.
  1748   - The theorems split_<typename>_case and split_<typename>_case_asm
  1749     have been renamed to <typename>.split and <typename>.split_asm.
  1750   - Since default sorts of type variables are now handled correctly,
  1751     some datatype definitions may have to be annotated with explicit
  1752     sort constraints.
  1753   - Primrec definitions no longer require function name and type
  1754     of recursive argument.
  1755 
  1756 Consider using isatool fixdatatype to adapt your theories and proof
  1757 scripts to the new package (backup your sources first!).
  1758 
  1759 
  1760 * HOL/record package: considerably improved implementation; now
  1761 includes concrete syntax for record types, terms, updates; theorems
  1762 for surjective pairing and splitting !!-bound record variables; proof
  1763 support is as follows:
  1764 
  1765   1) standard conversions (selectors or updates applied to record
  1766 constructor terms) are part of the standard simpset;
  1767 
  1768   2) inject equations of the form ((x, y) = (x', y')) == x=x' & y=y' are
  1769 made part of standard simpset and claset via addIffs;
  1770 
  1771   3) a tactic for record field splitting (record_split_tac) is part of
  1772 the standard claset (addSWrapper);
  1773 
  1774 To get a better idea about these rules you may retrieve them via
  1775 something like 'thms "foo.simps"' or 'thms "foo.iffs"', where "foo" is
  1776 the name of your record type.
  1777 
  1778 The split tactic 3) conceptually simplifies by the following rule:
  1779 
  1780   "(!!x. PROP ?P x) == (!!a b. PROP ?P (a, b))"
  1781 
  1782 Thus any record variable that is bound by meta-all will automatically
  1783 blow up into some record constructor term, consequently the
  1784 simplifications of 1), 2) apply.  Thus force_tac, auto_tac etc. shall
  1785 solve record problems automatically.
  1786 
  1787 
  1788 * reorganized the main HOL image: HOL/Integ and String loaded by
  1789 default; theory Main includes everything;
  1790 
  1791 * automatic simplification of integer sums and comparisons, using cancellation;
  1792 
  1793 * added option_map_eq_Some and not_Some_eq to the default simpset and claset;
  1794 
  1795 * added disj_not1 = "(~P | Q) = (P --> Q)" to the default simpset;
  1796 
  1797 * many new identities for unions, intersections, set difference, etc.;
  1798 
  1799 * expand_if, expand_split, expand_sum_case and expand_nat_case are now
  1800 called split_if, split_split, split_sum_case and split_nat_case (to go
  1801 with add/delsplits);
  1802 
  1803 * HOL/Prod introduces simplification procedure unit_eq_proc rewriting
  1804 (?x::unit) = (); this is made part of the default simpset, which COULD
  1805 MAKE EXISTING PROOFS FAIL under rare circumstances (consider
  1806 'Delsimprocs [unit_eq_proc];' as last resort); also note that
  1807 unit_abs_eta_conv is added in order to counter the effect of
  1808 unit_eq_proc on (%u::unit. f u), replacing it by f rather than by
  1809 %u.f();
  1810 
  1811 * HOL/Fun INCOMPATIBILITY: `inj_onto' is now called `inj_on' (which
  1812 makes more sense);
  1813 
  1814 * HOL/Set INCOMPATIBILITY: rule `equals0D' is now a well-formed destruct rule;
  1815   It and 'sym RS equals0D' are now in the default  claset, giving automatic
  1816   disjointness reasoning but breaking a few old proofs.
  1817 
  1818 * HOL/Relation INCOMPATIBILITY: renamed the relational operator r^-1
  1819 to 'converse' from 'inverse' (for compatibility with ZF and some
  1820 literature);
  1821 
  1822 * HOL/recdef can now declare non-recursive functions, with {} supplied as
  1823 the well-founded relation;
  1824 
  1825 * HOL/Set INCOMPATIBILITY: the complement of set A is now written -A instead of
  1826     Compl A.  The "Compl" syntax remains available as input syntax for this
  1827     release ONLY.
  1828 
  1829 * HOL/Update: new theory of function updates:
  1830     f(a:=b) == %x. if x=a then b else f x
  1831 may also be iterated as in f(a:=b,c:=d,...);
  1832 
  1833 * HOL/Vimage: new theory for inverse image of a function, syntax f-``B;
  1834 
  1835 * HOL/List:
  1836   - new function list_update written xs[i:=v] that updates the i-th
  1837     list position. May also be iterated as in xs[i:=a,j:=b,...].
  1838   - new function `upt' written [i..j(] which generates the list
  1839     [i,i+1,...,j-1], i.e. the upper bound is excluded. To include the upper
  1840     bound write [i..j], which is a shorthand for [i..j+1(].
  1841   - new lexicographic orderings and corresponding wellfoundedness theorems.
  1842 
  1843 * HOL/Arith:
  1844   - removed 'pred' (predecessor) function;
  1845   - generalized some theorems about n-1;
  1846   - many new laws about "div" and "mod";
  1847   - new laws about greatest common divisors (see theory ex/Primes);
  1848 
  1849 * HOL/Relation: renamed the relational operator r^-1 "converse"
  1850 instead of "inverse";
  1851 
  1852 * HOL/Induct/Multiset: a theory of multisets, including the wellfoundedness
  1853   of the multiset ordering;
  1854 
  1855 * directory HOL/Real: a construction of the reals using Dedekind cuts
  1856   (not included by default);
  1857 
  1858 * directory HOL/UNITY: Chandy and Misra's UNITY formalism;
  1859 
  1860 * directory HOL/Hoare: a new version of Hoare logic which permits many-sorted
  1861   programs, i.e. different program variables may have different types.
  1862 
  1863 * calling (stac rew i) now fails if "rew" has no effect on the goal
  1864   [previously, this check worked only if the rewrite rule was unconditional]
  1865   Now rew can involve either definitions or equalities (either == or =).
  1866 
  1867 
  1868 *** ZF ***
  1869 
  1870 * theory Main includes everything; INCOMPATIBILITY: theory ZF.thy contains
  1871   only the theorems proved on ZF.ML;
  1872 
  1873 * ZF INCOMPATIBILITY: rule `equals0D' is now a well-formed destruct rule;
  1874   It and 'sym RS equals0D' are now in the default  claset, giving automatic
  1875   disjointness reasoning but breaking a few old proofs.
  1876 
  1877 * ZF/Update: new theory of function updates
  1878     with default rewrite rule  f(x:=y) ` z = if(z=x, y, f`z)
  1879   may also be iterated as in f(a:=b,c:=d,...);
  1880 
  1881 * in  let x=t in u(x), neither t nor u(x) has to be an FOL term.
  1882 
  1883 * calling (stac rew i) now fails if "rew" has no effect on the goal
  1884   [previously, this check worked only if the rewrite rule was unconditional]
  1885   Now rew can involve either definitions or equalities (either == or =).
  1886 
  1887 * case_tac provided for compatibility with HOL
  1888     (like the old excluded_middle_tac, but with subgoals swapped)
  1889 
  1890 
  1891 *** Internal programming interfaces ***
  1892 
  1893 * Pure: several new basic modules made available for general use, see
  1894 also src/Pure/README;
  1895 
  1896 * improved the theory data mechanism to support encapsulation (data
  1897 kind name replaced by private Object.kind, acting as authorization
  1898 key); new type-safe user interface via functor TheoryDataFun; generic
  1899 print_data function becomes basically useless;
  1900 
  1901 * removed global_names compatibility flag -- all theory declarations
  1902 are qualified by default;
  1903 
  1904 * module Pure/Syntax now offers quote / antiquote translation
  1905 functions (useful for Hoare logic etc. with implicit dependencies);
  1906 see HOL/ex/Antiquote for an example use;
  1907 
  1908 * Simplifier now offers conversions (asm_)(full_)rewrite: simpset ->
  1909 cterm -> thm;
  1910 
  1911 * new tactical CHANGED_GOAL for checking that a tactic modifies a
  1912 subgoal;
  1913 
  1914 * Display.print_goals function moved to Locale.print_goals;
  1915 
  1916 * standard print function for goals supports current_goals_markers
  1917 variable for marking begin of proof, end of proof, start of goal; the
  1918 default is ("", "", ""); setting current_goals_markers := ("<proof>",
  1919 "</proof>", "<goal>") causes SGML like tagged proof state printing,
  1920 for example;
  1921 
  1922 
  1923 
  1924 New in Isabelle98 (January 1998)
  1925 --------------------------------
  1926 
  1927 *** Overview of INCOMPATIBILITIES (see below for more details) ***
  1928 
  1929 * changed lexical syntax of terms / types: dots made part of long
  1930 identifiers, e.g. "%x.x" no longer possible, should be "%x. x";
  1931 
  1932 * simpset (and claset) reference variable replaced by functions
  1933 simpset / simpset_ref;
  1934 
  1935 * no longer supports theory aliases (via merge) and non-trivial
  1936 implicit merge of thms' signatures;
  1937 
  1938 * most internal names of constants changed due to qualified names;
  1939 
  1940 * changed Pure/Sequence interface (see Pure/seq.ML);
  1941 
  1942 
  1943 *** General Changes ***
  1944 
  1945 * hierachically structured name spaces (for consts, types, axms, thms
  1946 etc.); new lexical class 'longid' (e.g. Foo.bar.x) may render much of
  1947 old input syntactically incorrect (e.g. "%x.x"); COMPATIBILITY:
  1948 isatool fixdots ensures space after dots (e.g. "%x. x"); set
  1949 long_names for fully qualified output names; NOTE: ML programs
  1950 (special tactics, packages etc.) referring to internal names may have
  1951 to be adapted to cope with fully qualified names; in case of severe
  1952 backward campatibility problems try setting 'global_names' at compile
  1953 time to have enrything declared within a flat name space; one may also
  1954 fine tune name declarations in theories via the 'global' and 'local'
  1955 section;
  1956 
  1957 * reimplemented the implicit simpset and claset using the new anytype
  1958 data filed in signatures; references simpset:simpset ref etc. are
  1959 replaced by functions simpset:unit->simpset and
  1960 simpset_ref:unit->simpset ref; COMPATIBILITY: use isatool fixclasimp
  1961 to patch your ML files accordingly;
  1962 
  1963 * HTML output now includes theory graph data for display with Java
  1964 applet or isatool browser; data generated automatically via isatool
  1965 usedir (see -i option, ISABELLE_USEDIR_OPTIONS);
  1966 
  1967 * defs may now be conditional; improved rewrite_goals_tac to handle
  1968 conditional equations;
  1969 
  1970 * defs now admits additional type arguments, using TYPE('a) syntax;
  1971 
  1972 * theory aliases via merge (e.g. M=A+B+C) no longer supported, always
  1973 creates a new theory node; implicit merge of thms' signatures is
  1974 restricted to 'trivial' ones; COMPATIBILITY: one may have to use
  1975 transfer:theory->thm->thm in (rare) cases;
  1976 
  1977 * improved handling of draft signatures / theories; draft thms (and
  1978 ctyps, cterms) are automatically promoted to real ones;
  1979 
  1980 * slightly changed interfaces for oracles: admit many per theory, named
  1981 (e.g. oracle foo = mlfun), additional name argument for invoke_oracle;
  1982 
  1983 * print_goals: optional output of const types (set show_consts and
  1984 show_types);
  1985 
  1986 * improved output of warnings (###) and errors (***);
  1987 
  1988 * subgoal_tac displays a warning if the new subgoal has type variables;
  1989 
  1990 * removed old README and Makefiles;
  1991 
  1992 * replaced print_goals_ref hook by print_current_goals_fn and result_error_fn;
  1993 
  1994 * removed obsolete init_pps and init_database;
  1995 
  1996 * deleted the obsolete tactical STATE, which was declared by
  1997     fun STATE tacfun st = tacfun st st;
  1998 
  1999 * cd and use now support path variables, e.g. $ISABELLE_HOME, or ~
  2000 (which abbreviates $HOME);
  2001 
  2002 * changed Pure/Sequence interface (see Pure/seq.ML); COMPATIBILITY:
  2003 use isatool fixseq to adapt your ML programs (this works for fully
  2004 qualified references to the Sequence structure only!);
  2005 
  2006 * use_thy no longer requires writable current directory; it always
  2007 reloads .ML *and* .thy file, if either one is out of date;
  2008 
  2009 
  2010 *** Classical Reasoner ***
  2011 
  2012 * Clarify_tac, clarify_tac, clarify_step_tac, Clarify_step_tac: new
  2013 tactics that use classical reasoning to simplify a subgoal without
  2014 splitting it into several subgoals;
  2015 
  2016 * Safe_tac: like safe_tac but uses the default claset;
  2017 
  2018 
  2019 *** Simplifier ***
  2020 
  2021 * added simplification meta rules:
  2022     (asm_)(full_)simplify: simpset -> thm -> thm;
  2023 
  2024 * simplifier.ML no longer part of Pure -- has to be loaded by object
  2025 logics (again);
  2026 
  2027 * added prems argument to simplification procedures;
  2028 
  2029 * HOL, FOL, ZF: added infix function `addsplits':
  2030   instead of `<simpset> setloop (split_tac <thms>)'
  2031   you can simply write `<simpset> addsplits <thms>'
  2032 
  2033 
  2034 *** Syntax ***
  2035 
  2036 * TYPE('a) syntax for type reflection terms;
  2037 
  2038 * no longer handles consts with name "" -- declare as 'syntax' instead;
  2039 
  2040 * pretty printer: changed order of mixfix annotation preference (again!);
  2041 
  2042 * Pure: fixed idt/idts vs. pttrn/pttrns syntactic categories;
  2043 
  2044 
  2045 *** HOL ***
  2046 
  2047 * HOL: there is a new splitter `split_asm_tac' that can be used e.g.
  2048   with `addloop' of the simplifier to faciliate case splitting in premises.
  2049 
  2050 * HOL/TLA: Stephan Merz's formalization of Lamport's Temporal Logic of Actions;
  2051 
  2052 * HOL/Auth: new protocol proofs including some for the Internet
  2053   protocol TLS;
  2054 
  2055 * HOL/Map: new theory of `maps' a la VDM;
  2056 
  2057 * HOL/simplifier: simplification procedures nat_cancel_sums for
  2058 cancelling out common nat summands from =, <, <= (in)equalities, or
  2059 differences; simplification procedures nat_cancel_factor for
  2060 cancelling common factor from =, <, <= (in)equalities over natural
  2061 sums; nat_cancel contains both kinds of procedures, it is installed by
  2062 default in Arith.thy -- this COULD MAKE EXISTING PROOFS FAIL;
  2063 
  2064 * HOL/simplifier: terms of the form
  2065   `? x. P1(x) & ... & Pn(x) & x=t & Q1(x) & ... Qn(x)'  (or t=x)
  2066   are rewritten to
  2067   `P1(t) & ... & Pn(t) & Q1(t) & ... Qn(t)',
  2068   and those of the form
  2069   `! x. P1(x) & ... & Pn(x) & x=t & Q1(x) & ... Qn(x) --> R(x)'  (or t=x)
  2070   are rewritten to
  2071   `P1(t) & ... & Pn(t) & Q1(t) & ... Qn(t) --> R(t)',
  2072 
  2073 * HOL/datatype
  2074   Each datatype `t' now comes with a theorem `split_t_case' of the form
  2075 
  2076   P(t_case f1 ... fn x) =
  2077      ( (!y1 ... ym1. x = C1 y1 ... ym1 --> P(f1 y1 ... ym1)) &
  2078         ...
  2079        (!y1 ... ymn. x = Cn y1 ... ymn --> P(f1 y1 ... ymn))
  2080      )
  2081 
  2082   and a theorem `split_t_case_asm' of the form
  2083 
  2084   P(t_case f1 ... fn x) =
  2085     ~( (? y1 ... ym1. x = C1 y1 ... ym1 & ~P(f1 y1 ... ym1)) |
  2086         ...
  2087        (? y1 ... ymn. x = Cn y1 ... ymn & ~P(f1 y1 ... ymn))
  2088      )
  2089   which can be added to a simpset via `addsplits'. The existing theorems
  2090   expand_list_case and expand_option_case have been renamed to
  2091   split_list_case and split_option_case.
  2092 
  2093 * HOL/Arithmetic:
  2094   - `pred n' is automatically converted to `n-1'.
  2095     Users are strongly encouraged not to use `pred' any longer,
  2096     because it will disappear altogether at some point.
  2097   - Users are strongly encouraged to write "0 < n" rather than
  2098     "n ~= 0". Theorems and proof tools have been modified towards this
  2099     `standard'.
  2100 
  2101 * HOL/Lists:
  2102   the function "set_of_list" has been renamed "set" (and its theorems too);
  2103   the function "nth" now takes its arguments in the reverse order and
  2104   has acquired the infix notation "!" as in "xs!n".
  2105 
  2106 * HOL/Set: UNIV is now a constant and is no longer translated to Compl{};
  2107 
  2108 * HOL/Set: The operator (UN x.B x) now abbreviates (UN x:UNIV. B x) and its
  2109   specialist theorems (like UN1_I) are gone.  Similarly for (INT x.B x);
  2110 
  2111 * HOL/record: extensible records with schematic structural subtyping
  2112 (single inheritance); EXPERIMENTAL version demonstrating the encoding,
  2113 still lacks various theorems and concrete record syntax;
  2114 
  2115 
  2116 *** HOLCF ***
  2117 
  2118 * removed "axioms" and "generated by" sections;
  2119 
  2120 * replaced "ops" section by extended "consts" section, which is capable of
  2121   handling the continuous function space "->" directly;
  2122 
  2123 * domain package:
  2124   . proves theorems immediately and stores them in the theory,
  2125   . creates hierachical name space,
  2126   . now uses normal mixfix annotations (instead of cinfix...),
  2127   . minor changes to some names and values (for consistency),
  2128   . e.g. cases -> casedist, dists_eq -> dist_eqs, [take_lemma] -> take_lemmas,
  2129   . separator between mutual domain defs: changed "," to "and",
  2130   . improved handling of sort constraints;  now they have to
  2131     appear on the left-hand side of the equations only;
  2132 
  2133 * fixed LAM <x,y,zs>.b syntax;
  2134 
  2135 * added extended adm_tac to simplifier in HOLCF -- can now discharge
  2136 adm (%x. P (t x)), where P is chainfinite and t continuous;
  2137 
  2138 
  2139 *** FOL and ZF ***
  2140 
  2141 * FOL: there is a new splitter `split_asm_tac' that can be used e.g.
  2142   with `addloop' of the simplifier to faciliate case splitting in premises.
  2143 
  2144 * qed_spec_mp, qed_goal_spec_mp, qed_goalw_spec_mp are available, as
  2145 in HOL, they strip ALL and --> from proved theorems;
  2146 
  2147 
  2148 
  2149 New in Isabelle94-8 (May 1997)
  2150 ------------------------------
  2151 
  2152 *** General Changes ***
  2153 
  2154 * new utilities to build / run / maintain Isabelle etc. (in parts
  2155 still somewhat experimental); old Makefiles etc. still functional;
  2156 
  2157 * new 'Isabelle System Manual';
  2158 
  2159 * INSTALL text, together with ./configure and ./build scripts;
  2160 
  2161 * reimplemented type inference for greater efficiency, better error
  2162 messages and clean internal interface;
  2163 
  2164 * prlim command for dealing with lots of subgoals (an easier way of
  2165 setting goals_limit);
  2166 
  2167 
  2168 *** Syntax ***
  2169 
  2170 * supports alternative (named) syntax tables (parser and pretty
  2171 printer); internal interface is provided by add_modesyntax(_i);
  2172 
  2173 * Pure, FOL, ZF, HOL, HOLCF now support symbolic input and output; to
  2174 be used in conjunction with the Isabelle symbol font; uses the
  2175 "symbols" syntax table;
  2176 
  2177 * added token_translation interface (may translate name tokens in
  2178 arbitrary ways, dependent on their type (free, bound, tfree, ...) and
  2179 the current print_mode); IMPORTANT: user print translation functions
  2180 are responsible for marking newly introduced bounds
  2181 (Syntax.mark_boundT);
  2182 
  2183 * token translations for modes "xterm" and "xterm_color" that display
  2184 names in bold, underline etc. or colors (which requires a color
  2185 version of xterm);
  2186 
  2187 * infixes may now be declared with names independent of their syntax;
  2188 
  2189 * added typed_print_translation (like print_translation, but may
  2190 access type of constant);
  2191 
  2192 
  2193 *** Classical Reasoner ***
  2194 
  2195 Blast_tac: a new tactic!  It is often more powerful than fast_tac, but has
  2196 some limitations.  Blast_tac...
  2197   + ignores addss, addbefore, addafter; this restriction is intrinsic
  2198   + ignores elimination rules that don't have the correct format
  2199         (the conclusion MUST be a formula variable)
  2200   + ignores types, which can make HOL proofs fail
  2201   + rules must not require higher-order unification, e.g. apply_type in ZF
  2202     [message "Function Var's argument not a bound variable" relates to this]
  2203   + its proof strategy is more general but can actually be slower
  2204 
  2205 * substitution with equality assumptions no longer permutes other
  2206 assumptions;
  2207 
  2208 * minor changes in semantics of addafter (now called addaltern); renamed
  2209 setwrapper to setWrapper and compwrapper to compWrapper; added safe wrapper
  2210 (and access functions for it);
  2211 
  2212 * improved combination of classical reasoner and simplifier:
  2213   + functions for handling clasimpsets
  2214   + improvement of addss: now the simplifier is called _after_ the
  2215     safe steps.
  2216   + safe variant of addss called addSss: uses safe simplifications
  2217     _during_ the safe steps. It is more complete as it allows multiple
  2218     instantiations of unknowns (e.g. with slow_tac).
  2219 
  2220 *** Simplifier ***
  2221 
  2222 * added interface for simplification procedures (functions that
  2223 produce *proven* rewrite rules on the fly, depending on current
  2224 redex);
  2225 
  2226 * ordering on terms as parameter (used for ordered rewriting);
  2227 
  2228 * new functions delcongs, deleqcongs, and Delcongs. richer rep_ss;
  2229 
  2230 * the solver is now split into a safe and an unsafe part.
  2231 This should be invisible for the normal user, except that the
  2232 functions setsolver and addsolver have been renamed to setSolver and
  2233 addSolver; added safe_asm_full_simp_tac;
  2234 
  2235 
  2236 *** HOL ***
  2237 
  2238 * a generic induction tactic `induct_tac' which works for all datatypes and
  2239 also for type `nat';
  2240 
  2241 * a generic case distinction tactic `exhaust_tac' which works for all
  2242 datatypes and also for type `nat';
  2243 
  2244 * each datatype comes with a function `size';
  2245 
  2246 * patterns in case expressions allow tuple patterns as arguments to
  2247 constructors, for example `case x of [] => ... | (x,y,z)#ps => ...';
  2248 
  2249 * primrec now also works with type nat;
  2250 
  2251 * recdef: a new declaration form, allows general recursive functions to be
  2252 defined in theory files.  See HOL/ex/Fib, HOL/ex/Primes, HOL/Subst/Unify.
  2253 
  2254 * the constant for negation has been renamed from "not" to "Not" to
  2255 harmonize with FOL, ZF, LK, etc.;
  2256 
  2257 * HOL/ex/LFilter theory of a corecursive "filter" functional for
  2258 infinite lists;
  2259 
  2260 * HOL/Modelcheck demonstrates invocation of model checker oracle;
  2261 
  2262 * HOL/ex/Ring.thy declares cring_simp, which solves equational
  2263 problems in commutative rings, using axiomatic type classes for + and *;
  2264 
  2265 * more examples in HOL/MiniML and HOL/Auth;
  2266 
  2267 * more default rewrite rules for quantifiers, union/intersection;
  2268 
  2269 * a new constant `arbitrary == @x.False';
  2270 
  2271 * HOLCF/IOA replaces old HOL/IOA;
  2272 
  2273 * HOLCF changes: derived all rules and arities
  2274   + axiomatic type classes instead of classes
  2275   + typedef instead of faking type definitions
  2276   + eliminated the internal constants less_fun, less_cfun, UU_fun, UU_cfun etc.
  2277   + new axclasses cpo, chfin, flat with flat < chfin < pcpo < cpo < po
  2278   + eliminated the types void, one, tr
  2279   + use unit lift and bool lift (with translations) instead of one and tr
  2280   + eliminated blift from Lift3.thy (use Def instead of blift)
  2281   all eliminated rules are derived as theorems --> no visible changes ;
  2282 
  2283 
  2284 *** ZF ***
  2285 
  2286 * ZF now has Fast_tac, Simp_tac and Auto_tac.  Union_iff is a now a default
  2287 rewrite rule; this may affect some proofs.  eq_cs is gone but can be put back
  2288 as ZF_cs addSIs [equalityI];
  2289 
  2290 
  2291 
  2292 New in Isabelle94-7 (November 96)
  2293 ---------------------------------
  2294 
  2295 * allowing negative levels (as offsets) in prlev and choplev;
  2296 
  2297 * super-linear speedup for large simplifications;
  2298 
  2299 * FOL, ZF and HOL now use miniscoping: rewriting pushes
  2300 quantifications in as far as possible (COULD MAKE EXISTING PROOFS
  2301 FAIL); can suppress it using the command Delsimps (ex_simps @
  2302 all_simps); De Morgan laws are also now included, by default;
  2303 
  2304 * improved printing of ==>  :  ~:
  2305 
  2306 * new object-logic "Sequents" adds linear logic, while replacing LK
  2307 and Modal (thanks to Sara Kalvala);
  2308 
  2309 * HOL/Auth: correctness proofs for authentication protocols;
  2310 
  2311 * HOL: new auto_tac combines rewriting and classical reasoning (many
  2312 examples on HOL/Auth);
  2313 
  2314 * HOL: new command AddIffs for declaring theorems of the form P=Q to
  2315 the rewriter and classical reasoner simultaneously;
  2316 
  2317 * function uresult no longer returns theorems in "standard" format;
  2318 regain previous version by: val uresult = standard o uresult;
  2319 
  2320 
  2321 
  2322 New in Isabelle94-6
  2323 -------------------
  2324 
  2325 * oracles -- these establish an interface between Isabelle and trusted
  2326 external reasoners, which may deliver results as theorems;
  2327 
  2328 * proof objects (in particular record all uses of oracles);
  2329 
  2330 * Simp_tac, Fast_tac, etc. that refer to implicit simpset / claset;
  2331 
  2332 * "constdefs" section in theory files;
  2333 
  2334 * "primrec" section (HOL) no longer requires names;
  2335 
  2336 * internal type "tactic" now simply "thm -> thm Sequence.seq";
  2337 
  2338 
  2339 
  2340 New in Isabelle94-5
  2341 -------------------
  2342 
  2343 * reduced space requirements;
  2344 
  2345 * automatic HTML generation from theories;
  2346 
  2347 * theory files no longer require "..." (quotes) around most types;
  2348 
  2349 * new examples, including two proofs of the Church-Rosser theorem;
  2350 
  2351 * non-curried (1994) version of HOL is no longer distributed;
  2352 
  2353 
  2354 
  2355 New in Isabelle94-4
  2356 -------------------
  2357 
  2358 * greatly reduced space requirements;
  2359 
  2360 * theory files (.thy) no longer require \...\ escapes at line breaks;
  2361 
  2362 * searchable theorem database (see the section "Retrieving theorems" on
  2363 page 8 of the Reference Manual);
  2364 
  2365 * new examples, including Grabczewski's monumental case study of the
  2366 Axiom of Choice;
  2367 
  2368 * The previous version of HOL renamed to Old_HOL;
  2369 
  2370 * The new version of HOL (previously called CHOL) uses a curried syntax
  2371 for functions.  Application looks like f a b instead of f(a,b);
  2372 
  2373 * Mutually recursive inductive definitions finally work in HOL;
  2374 
  2375 * In ZF, pattern-matching on tuples is now available in all abstractions and
  2376 translates to the operator "split";
  2377 
  2378 
  2379 
  2380 New in Isabelle94-3
  2381 -------------------
  2382 
  2383 * new infix operator, addss, allowing the classical reasoner to
  2384 perform simplification at each step of its search.  Example:
  2385         fast_tac (cs addss ss)
  2386 
  2387 * a new logic, CHOL, the same as HOL, but with a curried syntax
  2388 for functions.  Application looks like f a b instead of f(a,b).  Also pairs
  2389 look like (a,b) instead of <a,b>;
  2390 
  2391 * PLEASE NOTE: CHOL will eventually replace HOL!
  2392 
  2393 * In CHOL, pattern-matching on tuples is now available in all abstractions.
  2394 It translates to the operator "split".  A new theory of integers is available;
  2395 
  2396 * In ZF, integer numerals now denote two's-complement binary integers.
  2397 Arithmetic operations can be performed by rewriting.  See ZF/ex/Bin.ML;
  2398 
  2399 * Many new examples: I/O automata, Church-Rosser theorem, equivalents
  2400 of the Axiom of Choice;
  2401 
  2402 
  2403 
  2404 New in Isabelle94-2
  2405 -------------------
  2406 
  2407 * Significantly faster resolution;
  2408 
  2409 * the different sections in a .thy file can now be mixed and repeated
  2410 freely;
  2411 
  2412 * Database of theorems for FOL, HOL and ZF.  New
  2413 commands including qed, qed_goal and bind_thm store theorems in the database.
  2414 
  2415 * Simple database queries: return a named theorem (get_thm) or all theorems of
  2416 a given theory (thms_of), or find out what theory a theorem was proved in
  2417 (theory_of_thm);
  2418 
  2419 * Bugs fixed in the inductive definition and datatype packages;
  2420 
  2421 * The classical reasoner provides deepen_tac and depth_tac, making FOL_dup_cs
  2422 and HOL_dup_cs obsolete;
  2423 
  2424 * Syntactic ambiguities caused by the new treatment of syntax in Isabelle94-1
  2425 have been removed;
  2426 
  2427 * Simpler definition of function space in ZF;
  2428 
  2429 * new results about cardinal and ordinal arithmetic in ZF;
  2430 
  2431 * 'subtype' facility in HOL for introducing new types as subsets of existing
  2432 types;
  2433 
  2434 
  2435 $Id$