2 Isabelle NEWS -- history user-relevant changes
3 ==============================================
5 New in Isabelle2001 (?? 2001)
6 -----------------------------
8 *** Document preparation ***
10 * support bold style (for single symbols only), input syntax is like
11 this: "\<^bold>\<alpha>" or "\<^bold>A";
13 * \<bullet> is now output as bold \cdot by default, which looks much
14 better in printed text;
16 * added default LaTeX bindings for \<tturnstile> and \<TTurnstile>;
17 note that these symbols are currently unavailable in Proof General /
23 * improved proof by cases and induction:
24 - 'case' command admits impromptu naming of parameters (such as
26 - 'induct' method divinates rule instantiation from the inductive
27 claim; no longer requires excessive ?P bindings for proper
28 instantiation of cases;
29 - 'induct' method properly enumerates all possibilities of set/type
30 rules; as a consequence facts may be also passed through *type*
31 rules without further ado;
32 - removed obsolete "(simplified)" and "(stripped)" options of methods;
33 - added 'print_induct_rules' (covered by help item in Proof General > 3.3);
34 - moved induct/cases attributes to Pure, methods to Provers;
35 - generic method setup instantiated for FOL and HOL;
37 * Pure: renamed "antecedent" case to "rule_context";
39 * Pure: added 'corollary' command;
41 * Pure: fixed 'token_translation' command;
43 * HOL: 'typedef' now allows alternative names for Rep/Abs morphisms;
45 * HOL: 'recdef' now fails on unfinished automated proofs, use
46 "(permissive)" option to recover old behavior;
48 * HOL: 'inductive' now longer features separate (collective)
49 attributes for 'intros';
51 * HOL: canonical cases/induct rules for n-tuples (n = 3..7);
56 * HOL: moved over to sane numeral syntax; the new policy is as
59 - 0 and 1 are polymorphic constants, which are defined on any
60 numeric type (nat, int, real etc.);
62 - 2, 3, 4, ... and -1, -2, -3, ... are polymorphic numerals, based
63 binary representation internally;
65 - type nat has special constructor Suc, and generally prefers Suc 0
66 over 1::nat and Suc (Suc 0) over 2::nat;
68 This change may cause significant INCOMPATIBILITIES; here are some
69 hints on converting existing sources:
71 - due to the new "num" token, "-0" and "-1" etc. are now atomic
72 entities, so expressions involving "-" (unary or binary minus) need
73 to be spaced properly;
75 - existing occurrences of "1" may need to be constraint "1::nat" or
76 even replaced by Suc 0; similar for old "2";
78 - replace "#nnn" by "nnn", and "#-nnn" by "-nnn";
80 - remove all special provisions on numerals in proofs;
82 * HOL: concrete setsum syntax "\<Sum>i:A. b" == "setsum (%i. b) A"
83 (beware of argument permutation!);
85 * HOL: linorder_less_split superseded by linorder_cases;
87 * HOL: added "The" definite description operator; move Hilbert's "Eps"
88 to peripheral theory "Hilbert_Choice";
90 * HOL: made split_all_tac safe; EXISTING PROOFS MAY FAIL OR LOOP, so
91 in this (rare) case use:
93 delSWrapper "split_all_tac"
94 addSbefore ("unsafe_split_all_tac", unsafe_split_all_tac)
96 * HOL: added safe wrapper "split_conv_tac" to claset; EXISTING PROOFS
99 * HOL: introduced f^n = f o ... o f; warning: due to the limits of
100 Isabelle's type classes, ^ on functions and relations has too general
101 a domain, namely ('a * 'b) set and 'a => 'b; this means that it may be
102 necessary to attach explicit type constraints;
104 * HOL: syntax translations now work properly with numerals and records
107 * HOL/GroupTheory: group theory examples including Sylow's theorem, by
110 * HOL: eliminated global items
112 const "()" -> "Product_Type.Unity"
113 type "unit" -> "Product_Type.unit"
118 * ZF: the integer library now covers quotients and remainders, with
119 many laws relating division to addition, multiplication, etc.;
124 * kernel: meta-level proof terms (by Stefan Berghofer), see also ref
127 * classical reasoner: renamed addaltern to addafter, addSaltern to
130 * syntax: new token syntax "num" for plain numerals (without "#" of
131 "xnum"); potential INCOMPATIBILITY, since -0, -1 etc. are now separate
132 tokens, so expressions involving minus need to be spaced properly;
134 * syntax: support non-oriented infixes;
136 * syntax: print modes "type_brackets" and "no_type_brackets" control
137 output of nested => (types); the default behavior is "type_brackets";
139 * syntax: builtin parse translation for "_constify" turns valued
140 tokens into AST constants;
142 * system: support Poly/ML 4.1.1 (now able to manage large heaps);
144 * system: Proof General keywords specification is now part of the
145 Isabelle distribution (see etc/isar-keywords.el);
147 * system: smart selection of Isabelle process versus Isabelle
148 interface, accommodates case-insensitive file systems (e.g. HFS+); may
149 run both "isabelle" and "Isabelle" even if file names are badly
150 damaged (executable inspects the case of the first letter of its own
151 name); added separate "isabelle-process" and "isabelle-interface";
155 New in Isabelle99-2 (February 2001)
156 -----------------------------------
158 *** Overview of INCOMPATIBILITIES ***
160 * HOL: please note that theories in the Library and elsewhere often use the
161 new-style (Isar) format; to refer to their theorems in an ML script you must
162 bind them to ML identifers by e.g. val thm_name = thm "thm_name";
164 * HOL: inductive package no longer splits induction rule aggressively,
165 but only as far as specified by the introductions given; the old
166 format may be recovered via ML function complete_split_rule or attribute
167 'split_rule (complete)';
169 * HOL: induct renamed to lfp_induct, lfp_Tarski to lfp_unfold,
170 gfp_Tarski to gfp_unfold;
172 * HOL: contrapos, contrapos2 renamed to contrapos_nn, contrapos_pp;
174 * HOL: infix "dvd" now has priority 50 rather than 70 (because it is a
175 relation); infix "^^" has been renamed "``"; infix "``" has been
176 renamed "`"; "univalent" has been renamed "single_valued";
178 * HOL/Real: "rinv" and "hrinv" replaced by overloaded "inverse"
181 * HOLCF: infix "`" has been renamed "$"; the symbol syntax is \<cdot>;
183 * Isar: 'obtain' no longer declares "that" fact as simp/intro;
185 * Isar/HOL: method 'induct' now handles non-atomic goals; as a
186 consequence, it is no longer monotonic wrt. the local goal context
187 (which is now passed through the inductive cases);
189 * Document preparation: renamed standard symbols \<ll> to \<lless> and
190 \<gg> to \<ggreater>;
193 *** Document preparation ***
195 * \isabellestyle{NAME} selects version of Isabelle output (currently
196 available: are "it" for near math-mode best-style output, "sl" for
197 slanted text style, and "tt" for plain type-writer; if no
198 \isabellestyle command is given, output is according to slanted
201 * support sub/super scripts (for single symbols only), input syntax is
202 like this: "A\<^sup>*" or "A\<^sup>\<star>";
204 * some more standard symbols; see Appendix A of the system manual for
205 the complete list of symbols defined in isabellesym.sty;
207 * improved isabelle style files; more abstract symbol implementation
208 (should now use \isamath{...} and \isatext{...} in custom symbol
211 * antiquotation @{goals} and @{subgoals} for output of *dynamic* goals
212 state; Note that presentation of goal states does not conform to
213 actual human-readable proof documents. Please do not include goal
214 states into document output unless you really know what you are doing!
216 * proper indentation of antiquoted output with proportional LaTeX
219 * no_document ML operator temporarily disables LaTeX document
222 * isatool unsymbolize tunes sources for plain ASCII communication;
227 * Pure: Isar now suffers initial goal statements to contain unbound
228 schematic variables (this does not conform to actual readable proof
229 documents, due to unpredictable outcome and non-compositional proof
230 checking); users who know what they are doing may use schematic goals
231 for Prolog-style synthesis of proven results;
233 * Pure: assumption method (an implicit finishing) now handles actual
236 * Pure: improved 'obtain' --- moved to Pure, insert "that" into
237 initial goal, declare "that" only as Pure intro (only for single
238 steps); the "that" rule assumption may now be involved in implicit
239 finishing, thus ".." becomes a feasible for trivial obtains;
241 * Pure: default proof step now includes 'intro_classes'; thus trivial
242 instance proofs may be performed by "..";
244 * Pure: ?thesis / ?this / "..." now work for pure meta-level
247 * Pure: more robust selection of calculational rules;
249 * Pure: the builtin notion of 'finished' goal now includes the ==-refl
250 rule (as well as the assumption rule);
252 * Pure: 'thm_deps' command visualizes dependencies of theorems and
253 lemmas, using the graph browser tool;
255 * Pure: predict failure of "show" in interactive mode;
257 * Pure: 'thms_containing' now takes actual terms as arguments;
259 * HOL: improved method 'induct' --- now handles non-atomic goals
260 (potential INCOMPATIBILITY); tuned error handling;
262 * HOL: cases and induct rules now provide explicit hints about the
263 number of facts to be consumed (0 for "type" and 1 for "set" rules);
264 any remaining facts are inserted into the goal verbatim;
266 * HOL: local contexts (aka cases) may now contain term bindings as
267 well; the 'cases' and 'induct' methods new provide a ?case binding for
268 the result to be shown in each case;
270 * HOL: added 'recdef_tc' command;
272 * isatool convert assists in eliminating legacy ML scripts;
277 * HOL/Library: a collection of generic theories to be used together
278 with main HOL; the theory loader path already includes this directory
279 by default; the following existing theories have been moved here:
280 HOL/Induct/Multiset, HOL/Induct/Acc (as Accessible_Part), HOL/While
281 (as While_Combinator), HOL/Lex/Prefix (as List_Prefix);
283 * HOL/Unix: "Some aspects of Unix file-system security", a typical
284 modelling and verification task performed in Isabelle/HOL +
285 Isabelle/Isar + Isabelle document preparation (by Markus Wenzel).
287 * HOL/Algebra: special summation operator SUM no longer exists, it has
288 been replaced by setsum; infix 'assoc' now has priority 50 (like
289 'dvd'); axiom 'one_not_zero' has been moved from axclass 'ring' to
290 'domain', this makes the theory consistent with mathematical
293 * HOL basics: added overloaded operations "inverse" and "divide"
294 (infix "/"), syntax for generic "abs" operation, generic summation
297 * HOL/typedef: simplified package, provide more useful rules (see also
300 * HOL/datatype: induction rule for arbitrarily branching datatypes is
301 now expressed as a proper nested rule (old-style tactic scripts may
302 require atomize_strip_tac to cope with non-atomic premises);
304 * HOL: renamed theory "Prod" to "Product_Type", renamed "split" rule
305 to "split_conv" (old name still available for compatibility);
307 * HOL: improved concrete syntax for strings (e.g. allows translation
308 rules with string literals);
310 * HOL-Hyperreal: a new target, extending HOL-Real with the hyperreals
311 and Fleuriot's mechanization of analysis;
313 * HOL/Real, HOL/Hyperreal: improved arithmetic simplification;
318 * CTT: x-symbol support for Pi, Sigma, -->, : (membership); note that
319 "lam" is displayed as TWO lambda-symbols
321 * CTT: theory Main now available, containing everything (that is, Bool
327 * Pure: the Simplifier has been implemented properly as a derived rule
328 outside of the actual kernel (at last!); the overall performance
329 penalty in practical applications is about 50%, while reliability of
330 the Isabelle inference kernel has been greatly improved;
332 * print modes "brackets" and "no_brackets" control output of nested =>
333 (types) and ==> (props); the default behaviour is "brackets";
335 * Provers: fast_tac (and friends) now handle actual object-logic rules
336 as assumptions as well;
338 * system: support Poly/ML 4.0;
340 * system: isatool install handles KDE version 1 or 2;
344 New in Isabelle99-1 (October 2000)
345 ----------------------------------
347 *** Overview of INCOMPATIBILITIES ***
349 * HOL: simplification of natural numbers is much changed; to partly
350 recover the old behaviour (e.g. to prevent n+n rewriting to #2*n)
351 issue the following ML commands:
353 Delsimprocs Nat_Numeral_Simprocs.cancel_numerals;
354 Delsimprocs [Nat_Numeral_Simprocs.combine_numerals];
356 * HOL: simplification no longer dives into case-expressions; this is
357 controlled by "t.weak_case_cong" for each datatype t;
359 * HOL: nat_less_induct renamed to less_induct;
361 * HOL: systematic renaming of the SOME (Eps) rules, may use isatool
362 fixsome to patch .thy and .ML sources automatically;
364 select_equality -> some_equality
365 select_eq_Ex -> some_eq_ex
366 selectI2EX -> someI2_ex
369 select1_equality -> some1_equality
370 Eps_sym_eq -> some_sym_eq_trivial
371 Eps_eq -> some_eq_trivial
373 * HOL: exhaust_tac on datatypes superceded by new generic case_tac;
375 * HOL: removed obsolete theorem binding expand_if (refer to split_if
378 * HOL: the recursion equations generated by 'recdef' are now called
379 f.simps instead of f.rules;
381 * HOL: qed_spec_mp now also handles bounded ALL as well;
383 * HOL: 0 is now overloaded, so the type constraint ":: nat" may
386 * HOL: the constant for "f``x" is now "image" rather than "op ``";
388 * HOL: the constant for "f-``x" is now "vimage" rather than "op -``";
390 * HOL: the disjoint sum is now "<+>" instead of "Plus"; the cartesian
391 product is now "<*>" instead of "Times"; the lexicographic product is
392 now "<*lex*>" instead of "**";
394 * HOL: theory Sexp is now in HOL/Induct examples (it used to be part
395 of main HOL, but was unused); better use HOL's datatype package;
397 * HOL: removed "symbols" syntax for constant "override" of theory Map;
398 the old syntax may be recovered as follows:
401 override :: "('a ~=> 'b) => ('a ~=> 'b) => ('a ~=> 'b)"
402 (infixl "\\<oplus>" 100)
404 * HOL/Real: "rabs" replaced by overloaded "abs" function;
406 * HOL/ML: even fewer consts are declared as global (see theories Ord,
407 Lfp, Gfp, WF); this only affects ML packages that refer to const names
410 * HOL and ZF: syntax for quotienting wrt an equivalence relation
411 changed from A/r to A//r;
413 * ZF: new treatment of arithmetic (nat & int) may break some old
416 * Isar: renamed some attributes (RS -> THEN, simplify -> simplified,
417 rulify -> rule_format, elimify -> elim_format, ...);
419 * Isar/Provers: intro/elim/dest attributes changed; renamed
420 intro/intro!/intro!! flags to intro!/intro/intro? (in most cases, one
421 should have to change intro!! to intro? only); replaced "delrule" by
424 * Isar/HOL: renamed "intrs" to "intros" in inductive definitions;
426 * Provers: strengthened force_tac by using new first_best_tac;
428 * LaTeX document preparation: several changes of isabelle.sty (see
432 *** Document preparation ***
434 * formal comments (text blocks etc.) in new-style theories may now
435 contain antiquotations of thm/prop/term/typ/text to be presented
436 according to latex print mode; concrete syntax is like this:
437 @{term[show_types] "f(x) = a + x"};
439 * isatool mkdir provides easy setup of Isabelle session directories,
440 including proper document sources;
442 * generated LaTeX sources are now deleted after successful run
443 (isatool document -c); may retain a copy somewhere else via -D option
446 * isatool usedir -D now lets isatool latex -o sty update the Isabelle
447 style files, achieving self-contained LaTeX sources and simplifying
450 * old-style theories now produce (crude) LaTeX output as well;
452 * browser info session directories are now self-contained (may be put
453 on WWW server seperately); improved graphs of nested sessions; removed
454 graph for 'all sessions';
456 * several improvements in isabelle style files; \isabellestyle{it}
457 produces fake math mode output; \isamarkupheader is now \section by
458 default; see lib/texinputs/isabelle.sty etc.;
463 * Isar/Pure: local results and corresponding term bindings are now
464 subject to Hindley-Milner polymorphism (similar to ML); this
465 accommodates incremental type-inference very nicely;
467 * Isar/Pure: new derived language element 'obtain' supports
468 generalized existence reasoning;
470 * Isar/Pure: new calculational elements 'moreover' and 'ultimately'
471 support accumulation of results, without applying any rules yet;
472 useful to collect intermediate results without explicit name
473 references, and for use with transitivity rules with more than 2
476 * Isar/Pure: scalable support for case-analysis type proofs: new
477 'case' language element refers to local contexts symbolically, as
478 produced by certain proof methods; internally, case names are attached
479 to theorems as "tags";
481 * Isar/Pure: theory command 'hide' removes declarations from
482 class/type/const name spaces;
484 * Isar/Pure: theory command 'defs' supports option "(overloaded)" to
485 indicate potential overloading;
487 * Isar/Pure: changed syntax of local blocks from {{ }} to { };
489 * Isar/Pure: syntax of sorts made 'inner', i.e. have to write
490 "{a,b,c}" instead of {a,b,c};
492 * Isar/Pure now provides its own version of intro/elim/dest
493 attributes; useful for building new logics, but beware of confusion
494 with the version in Provers/classical;
496 * Isar/Pure: the local context of (non-atomic) goals is provided via
497 case name 'antecedent';
499 * Isar/Pure: removed obsolete 'transfer' attribute (transfer of thms
500 to the current context is now done automatically);
502 * Isar/Pure: theory command 'method_setup' provides a simple interface
503 for definining proof methods in ML;
505 * Isar/Provers: intro/elim/dest attributes changed; renamed
506 intro/intro!/intro!! flags to intro!/intro/intro? (INCOMPATIBILITY, in
507 most cases, one should have to change intro!! to intro? only);
508 replaced "delrule" by "rule del";
510 * Isar/Provers: new 'hypsubst' method, plain 'subst' method and
511 'symmetric' attribute (the latter supercedes [RS sym]);
513 * Isar/Provers: splitter support (via 'split' attribute and 'simp'
514 method modifier); 'simp' method: 'only:' modifier removes loopers as
515 well (including splits);
517 * Isar/Provers: Simplifier and Classical methods now support all kind
518 of modifiers used in the past, including 'cong', 'iff', etc.
520 * Isar/Provers: added 'fastsimp' and 'clarsimp' methods (combination
521 of Simplifier and Classical reasoner);
523 * Isar/HOL: new proof method 'cases' and improved version of 'induct'
524 now support named cases; major packages (inductive, datatype, primrec,
525 recdef) support case names and properly name parameters;
527 * Isar/HOL: new transitivity rules for substitution in inequalities --
528 monotonicity conditions are extracted to be proven at end of
531 * Isar/HOL: removed 'case_split' thm binding, should use 'cases' proof
534 * Isar/HOL: removed old expand_if = split_if; theorems if_splits =
535 split_if split_if_asm; datatype package provides theorems foo.splits =
536 foo.split foo.split_asm for each datatype;
538 * Isar/HOL: tuned inductive package, rename "intrs" to "intros"
539 (potential INCOMPATIBILITY), emulation of mk_cases feature for proof
540 scripts: new 'inductive_cases' command and 'ind_cases' method; (Note:
541 use "(cases (simplified))" method in proper proof texts);
543 * Isar/HOL: added global 'arith_split' attribute for 'arith' method;
545 * Isar: names of theorems etc. may be natural numbers as well;
547 * Isar: 'pr' command: optional arguments for goals_limit and
548 ProofContext.prems_limit; no longer prints theory contexts, but only
551 * Isar: diagnostic commands 'pr', 'thm', 'prop', 'term', 'typ' admit
552 additional print modes to be specified; e.g. "pr(latex)" will print
553 proof state according to the Isabelle LaTeX style;
555 * Isar: improved support for emulating tactic scripts, including proof
556 methods 'rule_tac' etc., 'cut_tac', 'thin_tac', 'subgoal_tac',
557 'rename_tac', 'rotate_tac', 'tactic', and 'case_tac' / 'induct_tac'
560 * Isar: simplified (more robust) goal selection of proof methods: 1st
561 goal, all goals, or explicit goal specifier (tactic emulation); thus
562 'proof method scripts' have to be in depth-first order;
564 * Isar: tuned 'let' syntax: replaced 'as' keyword by 'and';
566 * Isar: removed 'help' command, which hasn't been too helpful anyway;
567 should instead use individual commands for printing items
568 (print_commands, print_methods etc.);
570 * Isar: added 'nothing' --- the empty list of theorems;
575 * HOL/MicroJava: formalization of a fragment of Java, together with a
576 corresponding virtual machine and a specification of its bytecode
577 verifier and a lightweight bytecode verifier, including proofs of
578 type-safety; by Gerwin Klein, Tobias Nipkow, David von Oheimb, and
579 Cornelia Pusch (see also the homepage of project Bali at
580 http://isabelle.in.tum.de/Bali/);
582 * HOL/Algebra: new theory of rings and univariate polynomials, by
585 * HOL/NumberTheory: fundamental Theorem of Arithmetic, Chinese
586 Remainder Theorem, Fermat/Euler Theorem, Wilson's Theorem, by Thomas M
589 * HOL/Lattice: fundamental concepts of lattice theory and order
590 structures, including duals, properties of bounds versus algebraic
591 laws, lattice operations versus set-theoretic ones, the Knaster-Tarski
592 Theorem for complete lattices etc.; may also serve as a demonstration
593 for abstract algebraic reasoning using axiomatic type classes, and
594 mathematics-style proof in Isabelle/Isar; by Markus Wenzel;
596 * HOL/Prolog: a (bare-bones) implementation of Lambda-Prolog, by David
599 * HOL/IMPP: extension of IMP with local variables and mutually
600 recursive procedures, by David von Oheimb;
602 * HOL/Lambda: converted into new-style theory and document;
604 * HOL/ex/Multiquote: example of multiple nested quotations and
605 anti-quotations -- basically a generalized version of de-Bruijn
606 representation; very useful in avoiding lifting of operations;
608 * HOL/record: added general record equality rule to simpset; fixed
609 select-update simplification procedure to handle extended records as
610 well; admit "r" as field name;
612 * HOL: 0 is now overloaded over the new sort "zero", allowing its use with
613 other numeric types and also as the identity of groups, rings, etc.;
615 * HOL: new axclass plus_ac0 for addition with the AC-laws and 0 as identity.
616 Types nat and int belong to this axclass;
618 * HOL: greatly improved simplification involving numerals of type nat, int, real:
619 (i + #8 + j) = Suc k simplifies to #7 + (i + j) = k
620 i*j + k + j*#3*i simplifies to #4*(i*j) + k
621 two terms #m*u and #n*u are replaced by #(m+n)*u
622 (where #m, #n and u can implicitly be 1; this is simproc combine_numerals)
623 and the term/formula #m*u+x ~~ #n*u+y simplifies simplifies to #(m-n)+x ~~ y
624 or x ~~ #(n-m)+y, where ~~ is one of = < <= or - (simproc cancel_numerals);
626 * HOL: meson_tac is available (previously in ex/meson.ML); it is a
627 powerful prover for predicate logic but knows nothing of clasets; see
628 ex/mesontest.ML and ex/mesontest2.ML for example applications;
630 * HOL: new version of "case_tac" subsumes both boolean case split and
631 "exhaust_tac" on datatypes; INCOMPATIBILITY: exhaust_tac no longer
632 exists, may define val exhaust_tac = case_tac for ad-hoc portability;
634 * HOL: simplification no longer dives into case-expressions: only the
635 selector expression is simplified, but not the remaining arms; to
636 enable full simplification of case-expressions for datatype t, you may
637 remove t.weak_case_cong from the simpset, either globally (Delcongs
638 [thm"t.weak_case_cong"];) or locally (delcongs [...]).
640 * HOL/recdef: the recursion equations generated by 'recdef' for
641 function 'f' are now called f.simps instead of f.rules; if all
642 termination conditions are proved automatically, these simplification
643 rules are added to the simpset, as in primrec; rules may be named
644 individually as well, resulting in a separate list of theorems for
647 * HOL/While is a new theory that provides a while-combinator. It
648 permits the definition of tail-recursive functions without the
649 provision of a termination measure. The latter is necessary once the
650 invariant proof rule for while is applied.
652 * HOL: new (overloaded) notation for the set of elements below/above
653 some element: {..u}, {..u(}, {l..}, {)l..}. See theory SetInterval.
655 * HOL: theorems impI, allI, ballI bound as "strip";
657 * HOL: new tactic induct_thm_tac: thm -> string -> int -> tactic
658 induct_tac th "x1 ... xn" expects th to have a conclusion of the form
659 P v1 ... vn and abbreviates res_inst_tac [("v1","x1"),...,("vn","xn")] th;
661 * HOL/Real: "rabs" replaced by overloaded "abs" function;
663 * HOL: theory Sexp now in HOL/Induct examples (it used to be part of
664 main HOL, but was unused);
666 * HOL: fewer consts declared as global (e.g. have to refer to
667 "Lfp.lfp" instead of "lfp" internally; affects ML packages only);
669 * HOL: tuned AST representation of nested pairs, avoiding bogus output
670 in case of overlap with user translations (e.g. judgements over
671 tuples); (note that the underlying logical represenation is still
677 * ZF: simplification automatically cancels common terms in arithmetic
678 expressions over nat and int;
680 * ZF: new treatment of nat to minimize type-checking: all operators
681 coerce their operands to a natural number using the function natify,
682 making the algebraic laws unconditional;
684 * ZF: as above, for int: operators coerce their operands to an integer
685 using the function intify;
687 * ZF: the integer library now contains many of the usual laws for the
688 orderings, including $<=, and monotonicity laws for $+ and $*;
690 * ZF: new example ZF/ex/NatSum to demonstrate integer arithmetic
693 * FOL and ZF: AddIffs now available, giving theorems of the form P<->Q
694 to the simplifier and classical reasoner simultaneously;
699 * Provers: blast_tac now handles actual object-logic rules as
700 assumptions; note that auto_tac uses blast_tac internally as well;
702 * Provers: new functions rulify/rulify_no_asm: thm -> thm for turning
703 outer -->/All/Ball into ==>/!!; qed_spec_mp now uses rulify_no_asm;
705 * Provers: delrules now handles destruct rules as well (no longer need
708 * Provers: Blast_tac now warns of and ignores "weak elimination rules" e.g.
709 [| inj ?f; ?f ?x = ?f ?y; ?x = ?y ==> ?W |] ==> ?W
710 use instead the strong form,
711 [| inj ?f; ~ ?W ==> ?f ?x = ?f ?y; ?x = ?y ==> ?W |] ==> ?W
712 in HOL, FOL and ZF the function cla_make_elim will create such rules
715 * Provers: Simplifier.easy_setup provides a fast path to basic
716 Simplifier setup for new object-logics;
718 * Pure: AST translation rules no longer require constant head on LHS;
720 * Pure: improved name spaces: ambiguous output is qualified; support
723 * system: smart setup of canonical ML_HOME, ISABELLE_INTERFACE, and
724 XSYMBOL_HOME; no longer need to do manual configuration in most
727 * system: compression of ML heaps images may now be controlled via -c
728 option of isabelle and isatool usedir (currently only observed by
731 * system: isatool installfonts may handle X-Symbol fonts as well (very
732 useful for remote X11);
734 * system: provide TAGS file for Isabelle sources;
736 * ML: infix 'OF' is a version of 'MRS' with more appropriate argument
739 * ML: renamed flags Syntax.trace_norm_ast to Syntax.trace_ast; global
740 timing flag supersedes proof_timing and Toplevel.trace;
742 * ML: new combinators |>> and |>>> for incremental transformations
743 with secondary results (e.g. certain theory extensions):
745 * ML: PureThy.add_defs gets additional argument to indicate potential
746 overloading (usually false);
748 * ML: PureThy.add_thms/add_axioms/add_defs now return theorems as
753 New in Isabelle99 (October 1999)
754 --------------------------------
756 *** Overview of INCOMPATIBILITIES (see below for more details) ***
758 * HOL: The THEN and ELSE parts of conditional expressions (if P then x else y)
759 are no longer simplified. (This allows the simplifier to unfold recursive
760 functional programs.) To restore the old behaviour, declare
762 Delcongs [if_weak_cong];
764 * HOL: Removed the obsolete syntax "Compl A"; use -A for set
767 * HOL: the predicate "inj" is now defined by translation to "inj_on";
769 * HOL/datatype: mutual_induct_tac no longer exists --
770 use induct_tac "x_1 ... x_n" instead of mutual_induct_tac ["x_1", ..., "x_n"]
772 * HOL/typedef: fixed type inference for representing set; type
773 arguments now have to occur explicitly on the rhs as type constraints;
775 * ZF: The con_defs part of an inductive definition may no longer refer
776 to constants declared in the same theory;
778 * HOL, ZF: the function mk_cases, generated by the inductive
779 definition package, has lost an argument. To simplify its result, it
780 uses the default simpset instead of a supplied list of theorems.
782 * HOL/List: the constructors of type list are now Nil and Cons;
784 * Simplifier: the type of the infix ML functions
785 setSSolver addSSolver setSolver addSolver
786 is now simpset * solver -> simpset where `solver' is a new abstract type
787 for packaging solvers. A solver is created via
788 mk_solver: string -> (thm list -> int -> tactic) -> solver
789 where the string argument is only a comment.
794 * Provers/Arith/fast_lin_arith.ML contains a functor for creating a
795 decision procedure for linear arithmetic. Currently it is used for
796 types `nat', `int', and `real' in HOL (see below); it can, should and
797 will be instantiated for other types and logics as well.
799 * The simplifier now accepts rewrite rules with flexible heads, eg
800 hom ?f ==> ?f(?x+?y) = ?f ?x + ?f ?y
801 They are applied like any rule with a non-pattern lhs, i.e. by first-order
807 * New Isabelle/Isar subsystem provides an alternative to traditional
808 tactical theorem proving; together with the ProofGeneral/isar user
809 interface it offers an interactive environment for developing human
810 readable proof documents (Isar == Intelligible semi-automated
811 reasoning); for further information see isatool doc isar-ref,
812 src/HOL/Isar_examples and http://isabelle.in.tum.de/Isar/
814 * improved and simplified presentation of theories: better HTML markup
815 (including colors), graph views in several sizes; isatool usedir now
816 provides a proper interface for user theories (via -P option); actual
817 document preparation based on (PDF)LaTeX is available as well (for
818 new-style theories only); see isatool doc system for more information;
820 * native support for Proof General, both for classic Isabelle and
823 * ML function thm_deps visualizes dependencies of theorems and lemmas,
824 using the graph browser tool;
826 * Isabelle manuals now also available as PDF;
828 * theory loader rewritten from scratch (may not be fully
829 bug-compatible); old loadpath variable has been replaced by show_path,
830 add_path, del_path, reset_path functions; new operations such as
831 update_thy, touch_thy, remove_thy, use/update_thy_only (see also
834 * improved isatool install: option -k creates KDE application icon,
835 option -p DIR installs standalone binaries;
837 * added ML_PLATFORM setting (useful for cross-platform installations);
838 more robust handling of platform specific ML images for SML/NJ;
840 * the settings environment is now statically scoped, i.e. it is never
841 created again in sub-processes invoked from isabelle, isatool, or
844 * path element specification '~~' refers to '$ISABELLE_HOME';
846 * in locales, the "assumes" and "defines" parts may be omitted if
849 * new print_mode "xsymbols" for extended symbol support (e.g. genuine
852 * new print_mode "HTML";
854 * new flag show_tags controls display of tags of theorems (which are
855 basically just comments that may be attached by some tools);
857 * Isamode 2.6 requires patch to accomodate change of Isabelle font
858 mode and goal output format:
860 diff -r Isamode-2.6/elisp/isa-load.el Isamode/elisp/isa-load.el
862 < (list (isa-getenv "ISABELLE") "-msymbols" logic-name)
864 > (list (isa-getenv "ISABELLE") "-misabelle_font" "-msymbols" logic-name)
865 diff -r Isabelle-2.6/elisp/isa-proofstate.el Isamode/elisp/isa-proofstate.el
867 < (defconst proofstate-proofstart-regexp "^Level [0-9]+$"
869 > (defconst proofstate-proofstart-regexp "^Level [0-9]+"
871 * function bind_thms stores lists of theorems (cf. bind_thm);
873 * new shorthand tactics ftac, eatac, datac, fatac;
875 * qed (and friends) now accept "" as result name; in that case the
876 theorem is not stored, but proper checks and presentation of the
879 * theorem database now also indexes constants "Trueprop", "all",
880 "==>", "=="; thus thms_containing, findI etc. may retrieve more rules;
887 * There are now decision procedures for linear arithmetic over nat and
890 1. arith_tac copes with arbitrary formulae involving `=', `<', `<=',
891 `+', `-', `Suc', `min', `max' and numerical constants; other subterms
892 are treated as atomic; subformulae not involving type `nat' or `int'
893 are ignored; quantified subformulae are ignored unless they are
894 positive universal or negative existential. The tactic has to be
895 invoked by hand and can be a little bit slow. In particular, the
896 running time is exponential in the number of occurrences of `min' and
897 `max', and `-' on `nat'.
899 2. fast_arith_tac is a cut-down version of arith_tac: it only takes
900 (negated) (in)equalities among the premises and the conclusion into
901 account (i.e. no compound formulae) and does not know about `min' and
902 `max', and `-' on `nat'. It is fast and is used automatically by the
905 NB: At the moment, these decision procedures do not cope with mixed
906 nat/int formulae where the two parts interact, such as `m < n ==>
909 * HOL/Numeral provides a generic theory of numerals (encoded
910 efficiently as bit strings); setup for types nat/int/real is in place;
911 INCOMPATIBILITY: since numeral syntax is now polymorphic, rather than
912 int, existing theories and proof scripts may require a few additional
915 * integer division and remainder can now be performed on constant
918 * many properties of integer multiplication, division and remainder
921 * An interface to the Stanford Validity Checker (SVC) is available through the
922 tactic svc_tac. Propositional tautologies and theorems of linear arithmetic
923 are proved automatically. SVC must be installed separately, and its results
924 must be TAKEN ON TRUST (Isabelle does not check the proofs, but tags any
925 invocation of the underlying oracle). For SVC see
926 http://verify.stanford.edu/SVC
928 * IsaMakefile: the HOL-Real target now builds an actual image;
933 * HOL/Real/HahnBanach: the Hahn-Banach theorem for real vector spaces
934 (in Isabelle/Isar) -- by Gertrud Bauer;
936 * HOL/BCV: generic model of bytecode verification, i.e. data-flow
937 analysis for assembly languages with subtypes;
939 * HOL/TLA (Lamport's Temporal Logic of Actions): major reorganization
940 -- avoids syntactic ambiguities and treats state, transition, and
941 temporal levels more uniformly; introduces INCOMPATIBILITIES due to
942 changed syntax and (many) tactics;
944 * HOL/inductive: Now also handles more general introduction rules such
945 as "ALL y. (y, x) : r --> y : acc r ==> x : acc r"; monotonicity
946 theorems are now maintained within the theory (maintained via the
949 * HOL/datatype: Now also handles arbitrarily branching datatypes
950 (using function types) such as
952 datatype 'a tree = Atom 'a | Branch "nat => 'a tree"
954 * HOL/record: record_simproc (part of the default simpset) takes care
955 of selectors applied to updated records; record_split_tac is no longer
956 part of the default claset; update_defs may now be removed from the
957 simpset in many cases; COMPATIBILITY: old behavior achieved by
959 claset_ref () := claset() addSWrapper record_split_wrapper;
960 Delsimprocs [record_simproc]
962 * HOL/typedef: fixed type inference for representing set; type
963 arguments now have to occur explicitly on the rhs as type constraints;
965 * HOL/recdef (TFL): 'congs' syntax now expects comma separated list of theorem
966 names rather than an ML expression;
968 * HOL/defer_recdef (TFL): like recdef but the well-founded relation can be
969 supplied later. Program schemes can be defined, such as
970 "While B C s = (if B s then While B C (C s) else s)"
971 where the well-founded relation can be chosen after B and C have been given.
973 * HOL/List: the constructors of type list are now Nil and Cons;
974 INCOMPATIBILITY: while [] and infix # syntax is still there, of
975 course, ML tools referring to List.list.op # etc. have to be adapted;
977 * HOL_quantifiers flag superseded by "HOL" print mode, which is
978 disabled by default; run isabelle with option -m HOL to get back to
979 the original Gordon/HOL-style output;
981 * HOL/Ord.thy: new bounded quantifier syntax (input only): ALL x<y. P,
982 ALL x<=y. P, EX x<y. P, EX x<=y. P;
984 * HOL basic syntax simplified (more orthogonal): all variants of
985 All/Ex now support plain / symbolic / HOL notation; plain syntax for
986 Eps operator is provided as well: "SOME x. P[x]";
988 * HOL/Sum.thy: sum_case has been moved to HOL/Datatype;
990 * HOL/Univ.thy: infix syntax <*>, <+>, <**>, <+> eliminated and made
991 thus available for user theories;
993 * HOLCF/IOA/Sequents: renamed 'Cons' to 'Consq' to avoid clash with
994 HOL/List; hardly an INCOMPATIBILITY since '>>' syntax is used all the
997 * HOL: new tactic smp_tac: int -> int -> tactic, which applies spec
998 several times and then mp;
1003 * the notation <<...>> is now available as a notation for sequences of
1006 * the simplifier is now installed
1008 * the axiom system has been generalized (thanks to Soren Heilmann)
1010 * the classical reasoner now has a default rule database
1015 * new primrec section allows primitive recursive functions to be given
1016 directly (as in HOL) over datatypes and the natural numbers;
1018 * new tactics induct_tac and exhaust_tac for induction (or case
1019 analysis) over datatypes and the natural numbers;
1021 * the datatype declaration of type T now defines the recursor T_rec;
1023 * simplification automatically does freeness reasoning for datatype
1026 * automatic type-inference, with AddTCs command to insert new
1027 type-checking rules;
1029 * datatype introduction rules are now added as Safe Introduction rules
1032 * the syntax "if P then x else y" is now available in addition to
1036 *** Internal programming interfaces ***
1038 * tuned simplifier trace output; new flag debug_simp;
1040 * structures Vartab / Termtab (instances of TableFun) offer efficient
1041 tables indexed by indexname_ord / term_ord (compatible with aconv);
1043 * AxClass.axclass_tac lost the theory argument;
1045 * tuned current_goals_markers semantics: begin / end goal avoids
1046 printing empty lines;
1048 * removed prs and prs_fn hook, which was broken because it did not
1049 include \n in its semantics, forcing writeln to add one
1050 uncoditionally; replaced prs_fn by writeln_fn; consider std_output:
1051 string -> unit if you really want to output text without newline;
1053 * Symbol.output subject to print mode; INCOMPATIBILITY: defaults to
1054 plain output, interface builders may have to enable 'isabelle_font'
1055 mode to get Isabelle font glyphs as before;
1057 * refined token_translation interface; INCOMPATIBILITY: output length
1058 now of type real instead of int;
1060 * theory loader actions may be traced via new ThyInfo.add_hook
1061 interface (see src/Pure/Thy/thy_info.ML); example application: keep
1062 your own database of information attached to *whole* theories -- as
1063 opposed to intra-theory data slots offered via TheoryDataFun;
1065 * proper handling of dangling sort hypotheses (at last!);
1066 Thm.strip_shyps and Drule.strip_shyps_warning take care of removing
1067 extra sort hypotheses that can be witnessed from the type signature;
1068 the force_strip_shyps flag is gone, any remaining shyps are simply
1069 left in the theorem (with a warning issued by strip_shyps_warning);
1073 New in Isabelle98-1 (October 1998)
1074 ----------------------------------
1076 *** Overview of INCOMPATIBILITIES (see below for more details) ***
1078 * several changes of automated proof tools;
1080 * HOL: major changes to the inductive and datatype packages, including
1081 some minor incompatibilities of theory syntax;
1083 * HOL: renamed r^-1 to 'converse' from 'inverse'; 'inj_onto' is now
1086 * HOL: removed duplicate thms in Arith:
1087 less_imp_add_less should be replaced by trans_less_add1
1088 le_imp_add_le should be replaced by trans_le_add1
1090 * HOL: unary minus is now overloaded (new type constraints may be
1093 * HOL and ZF: unary minus for integers is now #- instead of #~. In
1094 ZF, expressions such as n#-1 must be changed to n#- 1, since #-1 is
1095 now taken as an integer constant.
1097 * Pure: ML function 'theory_of' renamed to 'theory';
1103 1. Asm_full_simp_tac is now more aggressive.
1104 1. It will sometimes reorient premises if that increases their power to
1106 2. It does no longer proceed strictly from left to right but may also
1107 rotate premises to achieve further simplification.
1108 For compatibility reasons there is now Asm_lr_simp_tac which is like the
1109 old Asm_full_simp_tac in that it does not rotate premises.
1110 2. The simplifier now knows a little bit about nat-arithmetic.
1112 * Classical reasoner: wrapper mechanism for the classical reasoner now
1113 allows for selected deletion of wrappers, by introduction of names for
1114 wrapper functionals. This implies that addbefore, addSbefore,
1115 addaltern, and addSaltern now take a pair (name, tactic) as argument,
1116 and that adding two tactics with the same name overwrites the first
1117 one (emitting a warning).
1118 type wrapper = (int -> tactic) -> (int -> tactic)
1119 setWrapper, setSWrapper, compWrapper and compSWrapper are replaced by
1120 addWrapper, addSWrapper: claset * (string * wrapper) -> claset
1121 delWrapper, delSWrapper: claset * string -> claset
1122 getWrapper is renamed to appWrappers, getSWrapper to appSWrappers;
1124 * Classical reasoner: addbefore/addSbefore now have APPEND/ORELSE
1125 semantics; addbefore now affects only the unsafe part of step_tac
1126 etc.; this affects addss/auto_tac/force_tac, so EXISTING PROOFS MAY
1127 FAIL, but proofs should be fixable easily, e.g. by replacing Auto_tac
1130 * Classical reasoner: setwrapper to setWrapper and compwrapper to
1131 compWrapper; added safe wrapper (and access functions for it);
1133 * HOL/split_all_tac is now much faster and fails if there is nothing
1134 to split. Some EXISTING PROOFS MAY REQUIRE ADAPTION because the order
1135 and the names of the automatically generated variables have changed.
1136 split_all_tac has moved within claset() from unsafe wrappers to safe
1137 wrappers, which means that !!-bound variables are split much more
1138 aggressively, and safe_tac and clarify_tac now split such variables.
1139 If this splitting is not appropriate, use delSWrapper "split_all_tac".
1140 Note: the same holds for record_split_tac, which does the job of
1141 split_all_tac for record fields.
1143 * HOL/Simplifier: Rewrite rules for case distinctions can now be added
1144 permanently to the default simpset using Addsplits just like
1145 Addsimps. They can be removed via Delsplits just like
1146 Delsimps. Lower-case versions are also available.
1148 * HOL/Simplifier: The rule split_if is now part of the default
1149 simpset. This means that the simplifier will eliminate all occurrences
1150 of if-then-else in the conclusion of a goal. To prevent this, you can
1151 either remove split_if completely from the default simpset by
1152 `Delsplits [split_if]' or remove it in a specific call of the
1153 simplifier using `... delsplits [split_if]'. You can also add/delete
1154 other case splitting rules to/from the default simpset: every datatype
1155 generates suitable rules `split_t_case' and `split_t_case_asm' (where
1156 t is the name of the datatype).
1158 * Classical reasoner / Simplifier combination: new force_tac (and
1159 derivatives Force_tac, force) combines rewriting and classical
1160 reasoning (and whatever other tools) similarly to auto_tac, but is
1161 aimed to solve the given subgoal completely.
1166 * new top-level commands `Goal' and `Goalw' that improve upon `goal'
1167 and `goalw': the theory is no longer needed as an explicit argument -
1168 the current theory context is used; assumptions are no longer returned
1169 at the ML-level unless one of them starts with ==> or !!; it is
1170 recommended to convert to these new commands using isatool fixgoal
1171 (backup your sources first!);
1173 * new top-level commands 'thm' and 'thms' for retrieving theorems from
1174 the current theory context, and 'theory' to lookup stored theories;
1176 * new theory section 'locale' for declaring constants, assumptions and
1177 definitions that have local scope;
1179 * new theory section 'nonterminals' for purely syntactic types;
1181 * new theory section 'setup' for generic ML setup functions
1182 (e.g. package initialization);
1184 * the distribution now includes Isabelle icons: see
1185 lib/logo/isabelle-{small,tiny}.xpm;
1187 * isatool install - install binaries with absolute references to
1190 * isatool logo -- create instances of the Isabelle logo (as EPS);
1192 * print mode 'emacs' reserved for Isamode;
1194 * support multiple print (ast) translations per constant name;
1196 * theorems involving oracles are now printed with a suffixed [!];
1201 * there is now a tutorial on Isabelle/HOL (do 'isatool doc tutorial');
1203 * HOL/inductive package reorganized and improved: now supports mutual
1209 oddI "n : EVEN ==> Suc n : ODD"
1210 evenI "n : ODD ==> Suc n : EVEN"
1212 new theorem list "elims" contains an elimination rule for each of the
1213 recursive sets; inductive definitions now handle disjunctive premises
1214 correctly (also ZF);
1216 INCOMPATIBILITIES: requires Inductive as an ancestor; component
1217 "mutual_induct" no longer exists - the induction rule is always
1218 contained in "induct";
1221 * HOL/datatype package re-implemented and greatly improved: now
1222 supports mutually recursive datatypes such as
1225 'a aexp = IF_THEN_ELSE ('a bexp) ('a aexp) ('a aexp)
1226 | SUM ('a aexp) ('a aexp)
1227 | DIFF ('a aexp) ('a aexp)
1230 'a bexp = LESS ('a aexp) ('a aexp)
1231 | AND ('a bexp) ('a bexp)
1232 | OR ('a bexp) ('a bexp)
1234 as well as indirectly recursive datatypes such as
1237 ('a, 'b) term = Var 'a
1238 | App 'b ((('a, 'b) term) list)
1240 The new tactic mutual_induct_tac [<var_1>, ..., <var_n>] i performs
1241 induction on mutually / indirectly recursive datatypes.
1243 Primrec equations are now stored in theory and can be accessed via
1244 <function_name>.simps.
1248 - Theories using datatypes must now have theory Datatype as an
1250 - The specific <typename>.induct_tac no longer exists - use the
1251 generic induct_tac instead.
1252 - natE has been renamed to nat.exhaust - use exhaust_tac
1253 instead of res_inst_tac ... natE. Note that the variable
1254 names in nat.exhaust differ from the names in natE, this
1255 may cause some "fragile" proofs to fail.
1256 - The theorems split_<typename>_case and split_<typename>_case_asm
1257 have been renamed to <typename>.split and <typename>.split_asm.
1258 - Since default sorts of type variables are now handled correctly,
1259 some datatype definitions may have to be annotated with explicit
1261 - Primrec definitions no longer require function name and type
1262 of recursive argument.
1264 Consider using isatool fixdatatype to adapt your theories and proof
1265 scripts to the new package (backup your sources first!).
1268 * HOL/record package: considerably improved implementation; now
1269 includes concrete syntax for record types, terms, updates; theorems
1270 for surjective pairing and splitting !!-bound record variables; proof
1271 support is as follows:
1273 1) standard conversions (selectors or updates applied to record
1274 constructor terms) are part of the standard simpset;
1276 2) inject equations of the form ((x, y) = (x', y')) == x=x' & y=y' are
1277 made part of standard simpset and claset via addIffs;
1279 3) a tactic for record field splitting (record_split_tac) is part of
1280 the standard claset (addSWrapper);
1282 To get a better idea about these rules you may retrieve them via
1283 something like 'thms "foo.simps"' or 'thms "foo.iffs"', where "foo" is
1284 the name of your record type.
1286 The split tactic 3) conceptually simplifies by the following rule:
1288 "(!!x. PROP ?P x) == (!!a b. PROP ?P (a, b))"
1290 Thus any record variable that is bound by meta-all will automatically
1291 blow up into some record constructor term, consequently the
1292 simplifications of 1), 2) apply. Thus force_tac, auto_tac etc. shall
1293 solve record problems automatically.
1296 * reorganized the main HOL image: HOL/Integ and String loaded by
1297 default; theory Main includes everything;
1299 * automatic simplification of integer sums and comparisons, using cancellation;
1301 * added option_map_eq_Some and not_Some_eq to the default simpset and claset;
1303 * added disj_not1 = "(~P | Q) = (P --> Q)" to the default simpset;
1305 * many new identities for unions, intersections, set difference, etc.;
1307 * expand_if, expand_split, expand_sum_case and expand_nat_case are now
1308 called split_if, split_split, split_sum_case and split_nat_case (to go
1309 with add/delsplits);
1311 * HOL/Prod introduces simplification procedure unit_eq_proc rewriting
1312 (?x::unit) = (); this is made part of the default simpset, which COULD
1313 MAKE EXISTING PROOFS FAIL under rare circumstances (consider
1314 'Delsimprocs [unit_eq_proc];' as last resort); also note that
1315 unit_abs_eta_conv is added in order to counter the effect of
1316 unit_eq_proc on (%u::unit. f u), replacing it by f rather than by
1319 * HOL/Fun INCOMPATIBILITY: `inj_onto' is now called `inj_on' (which
1322 * HOL/Set INCOMPATIBILITY: rule `equals0D' is now a well-formed destruct rule;
1323 It and 'sym RS equals0D' are now in the default claset, giving automatic
1324 disjointness reasoning but breaking a few old proofs.
1326 * HOL/Relation INCOMPATIBILITY: renamed the relational operator r^-1
1327 to 'converse' from 'inverse' (for compatibility with ZF and some
1330 * HOL/recdef can now declare non-recursive functions, with {} supplied as
1331 the well-founded relation;
1333 * HOL/Set INCOMPATIBILITY: the complement of set A is now written -A instead of
1334 Compl A. The "Compl" syntax remains available as input syntax for this
1337 * HOL/Update: new theory of function updates:
1338 f(a:=b) == %x. if x=a then b else f x
1339 may also be iterated as in f(a:=b,c:=d,...);
1341 * HOL/Vimage: new theory for inverse image of a function, syntax f-``B;
1344 - new function list_update written xs[i:=v] that updates the i-th
1345 list position. May also be iterated as in xs[i:=a,j:=b,...].
1346 - new function `upt' written [i..j(] which generates the list
1347 [i,i+1,...,j-1], i.e. the upper bound is excluded. To include the upper
1348 bound write [i..j], which is a shorthand for [i..j+1(].
1349 - new lexicographic orderings and corresponding wellfoundedness theorems.
1352 - removed 'pred' (predecessor) function;
1353 - generalized some theorems about n-1;
1354 - many new laws about "div" and "mod";
1355 - new laws about greatest common divisors (see theory ex/Primes);
1357 * HOL/Relation: renamed the relational operator r^-1 "converse"
1358 instead of "inverse";
1360 * HOL/Induct/Multiset: a theory of multisets, including the wellfoundedness
1361 of the multiset ordering;
1363 * directory HOL/Real: a construction of the reals using Dedekind cuts
1364 (not included by default);
1366 * directory HOL/UNITY: Chandy and Misra's UNITY formalism;
1368 * directory HOL/Hoare: a new version of Hoare logic which permits many-sorted
1369 programs, i.e. different program variables may have different types.
1371 * calling (stac rew i) now fails if "rew" has no effect on the goal
1372 [previously, this check worked only if the rewrite rule was unconditional]
1373 Now rew can involve either definitions or equalities (either == or =).
1378 * theory Main includes everything; INCOMPATIBILITY: theory ZF.thy contains
1379 only the theorems proved on ZF.ML;
1381 * ZF INCOMPATIBILITY: rule `equals0D' is now a well-formed destruct rule;
1382 It and 'sym RS equals0D' are now in the default claset, giving automatic
1383 disjointness reasoning but breaking a few old proofs.
1385 * ZF/Update: new theory of function updates
1386 with default rewrite rule f(x:=y) ` z = if(z=x, y, f`z)
1387 may also be iterated as in f(a:=b,c:=d,...);
1389 * in let x=t in u(x), neither t nor u(x) has to be an FOL term.
1391 * calling (stac rew i) now fails if "rew" has no effect on the goal
1392 [previously, this check worked only if the rewrite rule was unconditional]
1393 Now rew can involve either definitions or equalities (either == or =).
1395 * case_tac provided for compatibility with HOL
1396 (like the old excluded_middle_tac, but with subgoals swapped)
1399 *** Internal programming interfaces ***
1401 * Pure: several new basic modules made available for general use, see
1402 also src/Pure/README;
1404 * improved the theory data mechanism to support encapsulation (data
1405 kind name replaced by private Object.kind, acting as authorization
1406 key); new type-safe user interface via functor TheoryDataFun; generic
1407 print_data function becomes basically useless;
1409 * removed global_names compatibility flag -- all theory declarations
1410 are qualified by default;
1412 * module Pure/Syntax now offers quote / antiquote translation
1413 functions (useful for Hoare logic etc. with implicit dependencies);
1414 see HOL/ex/Antiquote for an example use;
1416 * Simplifier now offers conversions (asm_)(full_)rewrite: simpset ->
1419 * new tactical CHANGED_GOAL for checking that a tactic modifies a
1422 * Display.print_goals function moved to Locale.print_goals;
1424 * standard print function for goals supports current_goals_markers
1425 variable for marking begin of proof, end of proof, start of goal; the
1426 default is ("", "", ""); setting current_goals_markers := ("<proof>",
1427 "</proof>", "<goal>") causes SGML like tagged proof state printing,
1432 New in Isabelle98 (January 1998)
1433 --------------------------------
1435 *** Overview of INCOMPATIBILITIES (see below for more details) ***
1437 * changed lexical syntax of terms / types: dots made part of long
1438 identifiers, e.g. "%x.x" no longer possible, should be "%x. x";
1440 * simpset (and claset) reference variable replaced by functions
1441 simpset / simpset_ref;
1443 * no longer supports theory aliases (via merge) and non-trivial
1444 implicit merge of thms' signatures;
1446 * most internal names of constants changed due to qualified names;
1448 * changed Pure/Sequence interface (see Pure/seq.ML);
1451 *** General Changes ***
1453 * hierachically structured name spaces (for consts, types, axms, thms
1454 etc.); new lexical class 'longid' (e.g. Foo.bar.x) may render much of
1455 old input syntactically incorrect (e.g. "%x.x"); COMPATIBILITY:
1456 isatool fixdots ensures space after dots (e.g. "%x. x"); set
1457 long_names for fully qualified output names; NOTE: ML programs
1458 (special tactics, packages etc.) referring to internal names may have
1459 to be adapted to cope with fully qualified names; in case of severe
1460 backward campatibility problems try setting 'global_names' at compile
1461 time to have enrything declared within a flat name space; one may also
1462 fine tune name declarations in theories via the 'global' and 'local'
1465 * reimplemented the implicit simpset and claset using the new anytype
1466 data filed in signatures; references simpset:simpset ref etc. are
1467 replaced by functions simpset:unit->simpset and
1468 simpset_ref:unit->simpset ref; COMPATIBILITY: use isatool fixclasimp
1469 to patch your ML files accordingly;
1471 * HTML output now includes theory graph data for display with Java
1472 applet or isatool browser; data generated automatically via isatool
1473 usedir (see -i option, ISABELLE_USEDIR_OPTIONS);
1475 * defs may now be conditional; improved rewrite_goals_tac to handle
1476 conditional equations;
1478 * defs now admits additional type arguments, using TYPE('a) syntax;
1480 * theory aliases via merge (e.g. M=A+B+C) no longer supported, always
1481 creates a new theory node; implicit merge of thms' signatures is
1482 restricted to 'trivial' ones; COMPATIBILITY: one may have to use
1483 transfer:theory->thm->thm in (rare) cases;
1485 * improved handling of draft signatures / theories; draft thms (and
1486 ctyps, cterms) are automatically promoted to real ones;
1488 * slightly changed interfaces for oracles: admit many per theory, named
1489 (e.g. oracle foo = mlfun), additional name argument for invoke_oracle;
1491 * print_goals: optional output of const types (set show_consts and
1494 * improved output of warnings (###) and errors (***);
1496 * subgoal_tac displays a warning if the new subgoal has type variables;
1498 * removed old README and Makefiles;
1500 * replaced print_goals_ref hook by print_current_goals_fn and result_error_fn;
1502 * removed obsolete init_pps and init_database;
1504 * deleted the obsolete tactical STATE, which was declared by
1505 fun STATE tacfun st = tacfun st st;
1507 * cd and use now support path variables, e.g. $ISABELLE_HOME, or ~
1508 (which abbreviates $HOME);
1510 * changed Pure/Sequence interface (see Pure/seq.ML); COMPATIBILITY:
1511 use isatool fixseq to adapt your ML programs (this works for fully
1512 qualified references to the Sequence structure only!);
1514 * use_thy no longer requires writable current directory; it always
1515 reloads .ML *and* .thy file, if either one is out of date;
1518 *** Classical Reasoner ***
1520 * Clarify_tac, clarify_tac, clarify_step_tac, Clarify_step_tac: new
1521 tactics that use classical reasoning to simplify a subgoal without
1522 splitting it into several subgoals;
1524 * Safe_tac: like safe_tac but uses the default claset;
1529 * added simplification meta rules:
1530 (asm_)(full_)simplify: simpset -> thm -> thm;
1532 * simplifier.ML no longer part of Pure -- has to be loaded by object
1535 * added prems argument to simplification procedures;
1537 * HOL, FOL, ZF: added infix function `addsplits':
1538 instead of `<simpset> setloop (split_tac <thms>)'
1539 you can simply write `<simpset> addsplits <thms>'
1544 * TYPE('a) syntax for type reflection terms;
1546 * no longer handles consts with name "" -- declare as 'syntax' instead;
1548 * pretty printer: changed order of mixfix annotation preference (again!);
1550 * Pure: fixed idt/idts vs. pttrn/pttrns syntactic categories;
1555 * HOL: there is a new splitter `split_asm_tac' that can be used e.g.
1556 with `addloop' of the simplifier to faciliate case splitting in premises.
1558 * HOL/TLA: Stephan Merz's formalization of Lamport's Temporal Logic of Actions;
1560 * HOL/Auth: new protocol proofs including some for the Internet
1563 * HOL/Map: new theory of `maps' a la VDM;
1565 * HOL/simplifier: simplification procedures nat_cancel_sums for
1566 cancelling out common nat summands from =, <, <= (in)equalities, or
1567 differences; simplification procedures nat_cancel_factor for
1568 cancelling common factor from =, <, <= (in)equalities over natural
1569 sums; nat_cancel contains both kinds of procedures, it is installed by
1570 default in Arith.thy -- this COULD MAKE EXISTING PROOFS FAIL;
1572 * HOL/simplifier: terms of the form
1573 `? x. P1(x) & ... & Pn(x) & x=t & Q1(x) & ... Qn(x)' (or t=x)
1575 `P1(t) & ... & Pn(t) & Q1(t) & ... Qn(t)',
1576 and those of the form
1577 `! x. P1(x) & ... & Pn(x) & x=t & Q1(x) & ... Qn(x) --> R(x)' (or t=x)
1579 `P1(t) & ... & Pn(t) & Q1(t) & ... Qn(t) --> R(t)',
1582 Each datatype `t' now comes with a theorem `split_t_case' of the form
1584 P(t_case f1 ... fn x) =
1585 ( (!y1 ... ym1. x = C1 y1 ... ym1 --> P(f1 y1 ... ym1)) &
1587 (!y1 ... ymn. x = Cn y1 ... ymn --> P(f1 y1 ... ymn))
1590 and a theorem `split_t_case_asm' of the form
1592 P(t_case f1 ... fn x) =
1593 ~( (? y1 ... ym1. x = C1 y1 ... ym1 & ~P(f1 y1 ... ym1)) |
1595 (? y1 ... ymn. x = Cn y1 ... ymn & ~P(f1 y1 ... ymn))
1597 which can be added to a simpset via `addsplits'. The existing theorems
1598 expand_list_case and expand_option_case have been renamed to
1599 split_list_case and split_option_case.
1602 - `pred n' is automatically converted to `n-1'.
1603 Users are strongly encouraged not to use `pred' any longer,
1604 because it will disappear altogether at some point.
1605 - Users are strongly encouraged to write "0 < n" rather than
1606 "n ~= 0". Theorems and proof tools have been modified towards this
1610 the function "set_of_list" has been renamed "set" (and its theorems too);
1611 the function "nth" now takes its arguments in the reverse order and
1612 has acquired the infix notation "!" as in "xs!n".
1614 * HOL/Set: UNIV is now a constant and is no longer translated to Compl{};
1616 * HOL/Set: The operator (UN x.B x) now abbreviates (UN x:UNIV. B x) and its
1617 specialist theorems (like UN1_I) are gone. Similarly for (INT x.B x);
1619 * HOL/record: extensible records with schematic structural subtyping
1620 (single inheritance); EXPERIMENTAL version demonstrating the encoding,
1621 still lacks various theorems and concrete record syntax;
1626 * removed "axioms" and "generated by" sections;
1628 * replaced "ops" section by extended "consts" section, which is capable of
1629 handling the continuous function space "->" directly;
1632 . proves theorems immediately and stores them in the theory,
1633 . creates hierachical name space,
1634 . now uses normal mixfix annotations (instead of cinfix...),
1635 . minor changes to some names and values (for consistency),
1636 . e.g. cases -> casedist, dists_eq -> dist_eqs, [take_lemma] -> take_lemmas,
1637 . separator between mutual domain defs: changed "," to "and",
1638 . improved handling of sort constraints; now they have to
1639 appear on the left-hand side of the equations only;
1641 * fixed LAM <x,y,zs>.b syntax;
1643 * added extended adm_tac to simplifier in HOLCF -- can now discharge
1644 adm (%x. P (t x)), where P is chainfinite and t continuous;
1649 * FOL: there is a new splitter `split_asm_tac' that can be used e.g.
1650 with `addloop' of the simplifier to faciliate case splitting in premises.
1652 * qed_spec_mp, qed_goal_spec_mp, qed_goalw_spec_mp are available, as
1653 in HOL, they strip ALL and --> from proved theorems;
1657 New in Isabelle94-8 (May 1997)
1658 ------------------------------
1660 *** General Changes ***
1662 * new utilities to build / run / maintain Isabelle etc. (in parts
1663 still somewhat experimental); old Makefiles etc. still functional;
1665 * new 'Isabelle System Manual';
1667 * INSTALL text, together with ./configure and ./build scripts;
1669 * reimplemented type inference for greater efficiency, better error
1670 messages and clean internal interface;
1672 * prlim command for dealing with lots of subgoals (an easier way of
1673 setting goals_limit);
1678 * supports alternative (named) syntax tables (parser and pretty
1679 printer); internal interface is provided by add_modesyntax(_i);
1681 * Pure, FOL, ZF, HOL, HOLCF now support symbolic input and output; to
1682 be used in conjunction with the Isabelle symbol font; uses the
1683 "symbols" syntax table;
1685 * added token_translation interface (may translate name tokens in
1686 arbitrary ways, dependent on their type (free, bound, tfree, ...) and
1687 the current print_mode); IMPORTANT: user print translation functions
1688 are responsible for marking newly introduced bounds
1689 (Syntax.mark_boundT);
1691 * token translations for modes "xterm" and "xterm_color" that display
1692 names in bold, underline etc. or colors (which requires a color
1695 * infixes may now be declared with names independent of their syntax;
1697 * added typed_print_translation (like print_translation, but may
1698 access type of constant);
1701 *** Classical Reasoner ***
1703 Blast_tac: a new tactic! It is often more powerful than fast_tac, but has
1704 some limitations. Blast_tac...
1705 + ignores addss, addbefore, addafter; this restriction is intrinsic
1706 + ignores elimination rules that don't have the correct format
1707 (the conclusion MUST be a formula variable)
1708 + ignores types, which can make HOL proofs fail
1709 + rules must not require higher-order unification, e.g. apply_type in ZF
1710 [message "Function Var's argument not a bound variable" relates to this]
1711 + its proof strategy is more general but can actually be slower
1713 * substitution with equality assumptions no longer permutes other
1716 * minor changes in semantics of addafter (now called addaltern); renamed
1717 setwrapper to setWrapper and compwrapper to compWrapper; added safe wrapper
1718 (and access functions for it);
1720 * improved combination of classical reasoner and simplifier:
1721 + functions for handling clasimpsets
1722 + improvement of addss: now the simplifier is called _after_ the
1724 + safe variant of addss called addSss: uses safe simplifications
1725 _during_ the safe steps. It is more complete as it allows multiple
1726 instantiations of unknowns (e.g. with slow_tac).
1730 * added interface for simplification procedures (functions that
1731 produce *proven* rewrite rules on the fly, depending on current
1734 * ordering on terms as parameter (used for ordered rewriting);
1736 * new functions delcongs, deleqcongs, and Delcongs. richer rep_ss;
1738 * the solver is now split into a safe and an unsafe part.
1739 This should be invisible for the normal user, except that the
1740 functions setsolver and addsolver have been renamed to setSolver and
1741 addSolver; added safe_asm_full_simp_tac;
1746 * a generic induction tactic `induct_tac' which works for all datatypes and
1747 also for type `nat';
1749 * a generic case distinction tactic `exhaust_tac' which works for all
1750 datatypes and also for type `nat';
1752 * each datatype comes with a function `size';
1754 * patterns in case expressions allow tuple patterns as arguments to
1755 constructors, for example `case x of [] => ... | (x,y,z)#ps => ...';
1757 * primrec now also works with type nat;
1759 * recdef: a new declaration form, allows general recursive functions to be
1760 defined in theory files. See HOL/ex/Fib, HOL/ex/Primes, HOL/Subst/Unify.
1762 * the constant for negation has been renamed from "not" to "Not" to
1763 harmonize with FOL, ZF, LK, etc.;
1765 * HOL/ex/LFilter theory of a corecursive "filter" functional for
1768 * HOL/Modelcheck demonstrates invocation of model checker oracle;
1770 * HOL/ex/Ring.thy declares cring_simp, which solves equational
1771 problems in commutative rings, using axiomatic type classes for + and *;
1773 * more examples in HOL/MiniML and HOL/Auth;
1775 * more default rewrite rules for quantifiers, union/intersection;
1777 * a new constant `arbitrary == @x.False';
1779 * HOLCF/IOA replaces old HOL/IOA;
1781 * HOLCF changes: derived all rules and arities
1782 + axiomatic type classes instead of classes
1783 + typedef instead of faking type definitions
1784 + eliminated the internal constants less_fun, less_cfun, UU_fun, UU_cfun etc.
1785 + new axclasses cpo, chfin, flat with flat < chfin < pcpo < cpo < po
1786 + eliminated the types void, one, tr
1787 + use unit lift and bool lift (with translations) instead of one and tr
1788 + eliminated blift from Lift3.thy (use Def instead of blift)
1789 all eliminated rules are derived as theorems --> no visible changes ;
1794 * ZF now has Fast_tac, Simp_tac and Auto_tac. Union_iff is a now a default
1795 rewrite rule; this may affect some proofs. eq_cs is gone but can be put back
1796 as ZF_cs addSIs [equalityI];
1800 New in Isabelle94-7 (November 96)
1801 ---------------------------------
1803 * allowing negative levels (as offsets) in prlev and choplev;
1805 * super-linear speedup for large simplifications;
1807 * FOL, ZF and HOL now use miniscoping: rewriting pushes
1808 quantifications in as far as possible (COULD MAKE EXISTING PROOFS
1809 FAIL); can suppress it using the command Delsimps (ex_simps @
1810 all_simps); De Morgan laws are also now included, by default;
1812 * improved printing of ==> : ~:
1814 * new object-logic "Sequents" adds linear logic, while replacing LK
1815 and Modal (thanks to Sara Kalvala);
1817 * HOL/Auth: correctness proofs for authentication protocols;
1819 * HOL: new auto_tac combines rewriting and classical reasoning (many
1820 examples on HOL/Auth);
1822 * HOL: new command AddIffs for declaring theorems of the form P=Q to
1823 the rewriter and classical reasoner simultaneously;
1825 * function uresult no longer returns theorems in "standard" format;
1826 regain previous version by: val uresult = standard o uresult;
1833 * oracles -- these establish an interface between Isabelle and trusted
1834 external reasoners, which may deliver results as theorems;
1836 * proof objects (in particular record all uses of oracles);
1838 * Simp_tac, Fast_tac, etc. that refer to implicit simpset / claset;
1840 * "constdefs" section in theory files;
1842 * "primrec" section (HOL) no longer requires names;
1844 * internal type "tactic" now simply "thm -> thm Sequence.seq";
1851 * reduced space requirements;
1853 * automatic HTML generation from theories;
1855 * theory files no longer require "..." (quotes) around most types;
1857 * new examples, including two proofs of the Church-Rosser theorem;
1859 * non-curried (1994) version of HOL is no longer distributed;
1866 * greatly reduced space requirements;
1868 * theory files (.thy) no longer require \...\ escapes at line breaks;
1870 * searchable theorem database (see the section "Retrieving theorems" on
1871 page 8 of the Reference Manual);
1873 * new examples, including Grabczewski's monumental case study of the
1876 * The previous version of HOL renamed to Old_HOL;
1878 * The new version of HOL (previously called CHOL) uses a curried syntax
1879 for functions. Application looks like f a b instead of f(a,b);
1881 * Mutually recursive inductive definitions finally work in HOL;
1883 * In ZF, pattern-matching on tuples is now available in all abstractions and
1884 translates to the operator "split";
1891 * new infix operator, addss, allowing the classical reasoner to
1892 perform simplification at each step of its search. Example:
1893 fast_tac (cs addss ss)
1895 * a new logic, CHOL, the same as HOL, but with a curried syntax
1896 for functions. Application looks like f a b instead of f(a,b). Also pairs
1897 look like (a,b) instead of <a,b>;
1899 * PLEASE NOTE: CHOL will eventually replace HOL!
1901 * In CHOL, pattern-matching on tuples is now available in all abstractions.
1902 It translates to the operator "split". A new theory of integers is available;
1904 * In ZF, integer numerals now denote two's-complement binary integers.
1905 Arithmetic operations can be performed by rewriting. See ZF/ex/Bin.ML;
1907 * Many new examples: I/O automata, Church-Rosser theorem, equivalents
1908 of the Axiom of Choice;
1915 * Significantly faster resolution;
1917 * the different sections in a .thy file can now be mixed and repeated
1920 * Database of theorems for FOL, HOL and ZF. New
1921 commands including qed, qed_goal and bind_thm store theorems in the database.
1923 * Simple database queries: return a named theorem (get_thm) or all theorems of
1924 a given theory (thms_of), or find out what theory a theorem was proved in
1927 * Bugs fixed in the inductive definition and datatype packages;
1929 * The classical reasoner provides deepen_tac and depth_tac, making FOL_dup_cs
1930 and HOL_dup_cs obsolete;
1932 * Syntactic ambiguities caused by the new treatment of syntax in Isabelle94-1
1935 * Simpler definition of function space in ZF;
1937 * new results about cardinal and ordinal arithmetic in ZF;
1939 * 'subtype' facility in HOL for introducing new types as subsets of existing