summary |
shortlog |
changelog |
graph |
tags |
branches |
files |
changeset |
file |
revisions |
annotate |
diff |
raw

NEWS

author | paulson |

Tue Jan 19 11:18:11 1999 +0100 (1999-01-19) | |

changeset 6141 | a6922171b396 |

parent 6131 | 2d9e609abcdb |

child 6155 | e387d188d0ca |

permissions | -rw-r--r-- |

removal of the (thm list) argument of mk_cases

1 Isabelle NEWS -- history user-relevant changes

2 ==============================================

4 New in this Isabelle version

5 ----------------------------

7 *** Overview of INCOMPATIBILITIES (see below for more details) ***

9 * HOL: Removed the obsolete syntax "Compl A"; use -A for set complement

11 * ZF: The con_defs part of an inductive definition may no longer refer to

12 constants declared in the same theory;

14 * HOL, ZF: the function mk_cases, generated by the inductive definition

15 package, has lost an argument. To simplify its result, it uses the default

16 simpset instead of a supplied list of theorems.

19 *** Proof tools ***

21 * Provers/Arith/fast_lin_arith.ML contains a functor for creating a decision

22 procedure for linear arithmetic. Currently it is used for types `nat' and

23 `int' in HOL (see below) but can, should and will be instantiated for other

24 types and logics as well.

27 *** General ***

29 * in locales, the "assumes" and "defines" parts may be omitted if empty;

31 * new print_mode "xsymbols" for extended symbol support

32 (e.g. genuiely long arrows)

34 *** HOL ***

36 * There are now decision procedures for linear arithmetic over nat and int:

38 1. arith_tac copes with arbitrary formulae involving `=', `<', `<=', `+',

39 `-', `Suc' and numerical constants; other subterms are treated as atomic;

40 subformulae not involving type `nat' or `int' are ignored; quantified

41 subformulae are ignored unless they are positive universal or negative

42 existential. The tactic has to be invoked by hand and can be a little bit

43 slow. In particular, the running time is exponential in the number of

44 occurrences of `-' on `nat'.

46 2. fast_arith_tac is a cut-down version of arith_tac: it only takes (negated)

47 (in)equalities among the premises and the conclusion into account (i.e. no

48 compound formulae) and does not know about `-' on `nat'. It is fast and is

49 used automatically by the simplifier.

51 NB: At the moment, these decision procedures do not cope with mixed nat/int

52 formulae where the two parts interact, such as `m < n ==> int(m) < int(n)'.

54 *** Internal programming interfaces ***

56 * tuned current_goals_markers semantics: begin / end goal avoids

57 printing empty lines;

59 * removed prs and prs_fn hook, which was broken because it did not

60 include \n in its semantics, forcing writeln to add one

61 uncoditionally; replaced prs_fn by writeln_fn; consider std_output:

62 string -> unit if you really want to output text without newline;

65 *** ZF ***

67 * new primrec section allows primitive recursive functions to be given

68 directly (as in HOL) over datatypes and the natural numbers;

70 * new tactics induct_tac and exhaust_tac for induction (or case analysis)

71 over datatypes and the natural numbers;

73 * the datatype declaration of type T now defines the recursor T_rec;

75 * simplification automatically does freeness reasoning for datatype

76 constructors;

78 * The syntax "if P then x else y" is now available in addition to if(P,x,y).

81 New in Isabelle98-1 (October 1998)

82 ----------------------------------

84 *** Overview of INCOMPATIBILITIES (see below for more details) ***

86 * several changes of automated proof tools;

88 * HOL: major changes to the inductive and datatype packages, including

89 some minor incompatibilities of theory syntax;

91 * HOL: renamed r^-1 to 'converse' from 'inverse'; 'inj_onto' is now

92 called `inj_on';

94 * HOL: removed duplicate thms in Arith:

95 less_imp_add_less should be replaced by trans_less_add1

96 le_imp_add_le should be replaced by trans_le_add1

98 * HOL: unary minus is now overloaded (new type constraints may be

99 required);

101 * HOL and ZF: unary minus for integers is now #- instead of #~. In

102 ZF, expressions such as n#-1 must be changed to n#- 1, since #-1 is

103 now taken as an integer constant.

105 * Pure: ML function 'theory_of' renamed to 'theory';

108 *** Proof tools ***

110 * Simplifier:

111 1. Asm_full_simp_tac is now more aggressive.

112 1. It will sometimes reorient premises if that increases their power to

113 simplify.

114 2. It does no longer proceed strictly from left to right but may also

115 rotate premises to achieve further simplification.

116 For compatibility reasons there is now Asm_lr_simp_tac which is like the

117 old Asm_full_simp_tac in that it does not rotate premises.

118 2. The simplifier now knows a little bit about nat-arithmetic.

120 * Classical reasoner: wrapper mechanism for the classical reasoner now

121 allows for selected deletion of wrappers, by introduction of names for

122 wrapper functionals. This implies that addbefore, addSbefore,

123 addaltern, and addSaltern now take a pair (name, tactic) as argument,

124 and that adding two tactics with the same name overwrites the first

125 one (emitting a warning).

126 type wrapper = (int -> tactic) -> (int -> tactic)

127 setWrapper, setSWrapper, compWrapper and compSWrapper are replaced by

128 addWrapper, addSWrapper: claset * (string * wrapper) -> claset

129 delWrapper, delSWrapper: claset * string -> claset

130 getWrapper is renamed to appWrappers, getSWrapper to appSWrappers;

132 * Classical reasoner: addbefore/addSbefore now have APPEND/ORELSE

133 semantics; addbefore now affects only the unsafe part of step_tac

134 etc.; this affects addss/auto_tac/force_tac, so EXISTING PROOFS MAY

135 FAIL, but proofs should be fixable easily, e.g. by replacing Auto_tac

136 by Force_tac;

138 * Classical reasoner: setwrapper to setWrapper and compwrapper to

139 compWrapper; added safe wrapper (and access functions for it);

141 * HOL/split_all_tac is now much faster and fails if there is nothing

142 to split. Some EXISTING PROOFS MAY REQUIRE ADAPTION because the order

143 and the names of the automatically generated variables have changed.

144 split_all_tac has moved within claset() from unsafe wrappers to safe

145 wrappers, which means that !!-bound variables are split much more

146 aggressively, and safe_tac and clarify_tac now split such variables.

147 If this splitting is not appropriate, use delSWrapper "split_all_tac".

148 Note: the same holds for record_split_tac, which does the job of

149 split_all_tac for record fields.

151 * HOL/Simplifier: Rewrite rules for case distinctions can now be added

152 permanently to the default simpset using Addsplits just like

153 Addsimps. They can be removed via Delsplits just like

154 Delsimps. Lower-case versions are also available.

156 * HOL/Simplifier: The rule split_if is now part of the default

157 simpset. This means that the simplifier will eliminate all occurrences

158 of if-then-else in the conclusion of a goal. To prevent this, you can

159 either remove split_if completely from the default simpset by

160 `Delsplits [split_if]' or remove it in a specific call of the

161 simplifier using `... delsplits [split_if]'. You can also add/delete

162 other case splitting rules to/from the default simpset: every datatype

163 generates suitable rules `split_t_case' and `split_t_case_asm' (where

164 t is the name of the datatype).

166 * Classical reasoner / Simplifier combination: new force_tac (and

167 derivatives Force_tac, force) combines rewriting and classical

168 reasoning (and whatever other tools) similarly to auto_tac, but is

169 aimed to solve the given subgoal completely.

172 *** General ***

174 * new top-level commands `Goal' and `Goalw' that improve upon `goal'

175 and `goalw': the theory is no longer needed as an explicit argument -

176 the current theory context is used; assumptions are no longer returned

177 at the ML-level unless one of them starts with ==> or !!; it is

178 recommended to convert to these new commands using isatool fixgoal

179 (backup your sources first!);

181 * new top-level commands 'thm' and 'thms' for retrieving theorems from

182 the current theory context, and 'theory' to lookup stored theories;

184 * new theory section 'locale' for declaring constants, assumptions and

185 definitions that have local scope;

187 * new theory section 'nonterminals' for purely syntactic types;

189 * new theory section 'setup' for generic ML setup functions

190 (e.g. package initialization);

192 * the distribution now includes Isabelle icons: see

193 lib/logo/isabelle-{small,tiny}.xpm;

195 * isatool install - install binaries with absolute references to

196 ISABELLE_HOME/bin;

198 * isatool logo -- create instances of the Isabelle logo (as EPS);

200 * print mode 'emacs' reserved for Isamode;

202 * support multiple print (ast) translations per constant name;

205 *** HOL ***

207 * there is now a tutorial on Isabelle/HOL (do 'isatool doc tutorial');

209 * HOL/inductive package reorganized and improved: now supports mutual

210 definitions such as

212 inductive EVEN ODD

213 intrs

214 null "0 : EVEN"

215 oddI "n : EVEN ==> Suc n : ODD"

216 evenI "n : ODD ==> Suc n : EVEN"

218 new theorem list "elims" contains an elimination rule for each of the

219 recursive sets; inductive definitions now handle disjunctive premises

220 correctly (also ZF);

222 INCOMPATIBILITIES: requires Inductive as an ancestor; component

223 "mutual_induct" no longer exists - the induction rule is always

224 contained in "induct";

227 * HOL/datatype package re-implemented and greatly improved: now

228 supports mutually recursive datatypes such as

230 datatype

231 'a aexp = IF_THEN_ELSE ('a bexp) ('a aexp) ('a aexp)

232 | SUM ('a aexp) ('a aexp)

233 | DIFF ('a aexp) ('a aexp)

234 | NUM 'a

235 and

236 'a bexp = LESS ('a aexp) ('a aexp)

237 | AND ('a bexp) ('a bexp)

238 | OR ('a bexp) ('a bexp)

240 as well as indirectly recursive datatypes such as

242 datatype

243 ('a, 'b) term = Var 'a

244 | App 'b ((('a, 'b) term) list)

246 The new tactic mutual_induct_tac [<var_1>, ..., <var_n>] i performs

247 induction on mutually / indirectly recursive datatypes.

249 Primrec equations are now stored in theory and can be accessed via

250 <function_name>.simps.

252 INCOMPATIBILITIES:

254 - Theories using datatypes must now have theory Datatype as an

255 ancestor.

256 - The specific <typename>.induct_tac no longer exists - use the

257 generic induct_tac instead.

258 - natE has been renamed to nat.exhaust - use exhaust_tac

259 instead of res_inst_tac ... natE. Note that the variable

260 names in nat.exhaust differ from the names in natE, this

261 may cause some "fragile" proofs to fail.

262 - The theorems split_<typename>_case and split_<typename>_case_asm

263 have been renamed to <typename>.split and <typename>.split_asm.

264 - Since default sorts of type variables are now handled correctly,

265 some datatype definitions may have to be annotated with explicit

266 sort constraints.

267 - Primrec definitions no longer require function name and type

268 of recursive argument.

270 Consider using isatool fixdatatype to adapt your theories and proof

271 scripts to the new package (backup your sources first!).

274 * HOL/record package: considerably improved implementation; now

275 includes concrete syntax for record types, terms, updates; theorems

276 for surjective pairing and splitting !!-bound record variables; proof

277 support is as follows:

279 1) standard conversions (selectors or updates applied to record

280 constructor terms) are part of the standard simpset;

282 2) inject equations of the form ((x, y) = (x', y')) == x=x' & y=y' are

283 made part of standard simpset and claset via addIffs;

285 3) a tactic for record field splitting (record_split_tac) is part of

286 the standard claset (addSWrapper);

288 To get a better idea about these rules you may retrieve them via

289 something like 'thms "foo.simps"' or 'thms "foo.iffs"', where "foo" is

290 the name of your record type.

292 The split tactic 3) conceptually simplifies by the following rule:

294 "(!!x. PROP ?P x) == (!!a b. PROP ?P (a, b))"

296 Thus any record variable that is bound by meta-all will automatically

297 blow up into some record constructor term, consequently the

298 simplifications of 1), 2) apply. Thus force_tac, auto_tac etc. shall

299 solve record problems automatically.

302 * reorganized the main HOL image: HOL/Integ and String loaded by

303 default; theory Main includes everything;

305 * automatic simplification of integer sums and comparisons, using cancellation;

307 * added option_map_eq_Some and not_Some_eq to the default simpset and claset;

309 * added disj_not1 = "(~P | Q) = (P --> Q)" to the default simpset;

311 * many new identities for unions, intersections, set difference, etc.;

313 * expand_if, expand_split, expand_sum_case and expand_nat_case are now

314 called split_if, split_split, split_sum_case and split_nat_case (to go

315 with add/delsplits);

317 * HOL/Prod introduces simplification procedure unit_eq_proc rewriting

318 (?x::unit) = (); this is made part of the default simpset, which COULD

319 MAKE EXISTING PROOFS FAIL under rare circumstances (consider

320 'Delsimprocs [unit_eq_proc];' as last resort); also note that

321 unit_abs_eta_conv is added in order to counter the effect of

322 unit_eq_proc on (%u::unit. f u), replacing it by f rather than by

323 %u.f();

325 * HOL/Fun INCOMPATIBILITY: `inj_onto' is now called `inj_on' (which

326 makes more sense);

328 * HOL/Set INCOMPATIBILITY: rule `equals0D' is now a well-formed destruct rule;

329 It and 'sym RS equals0D' are now in the default claset, giving automatic

330 disjointness reasoning but breaking a few old proofs.

332 * HOL/Relation INCOMPATIBILITY: renamed the relational operator r^-1

333 to 'converse' from 'inverse' (for compatibility with ZF and some

334 literature);

336 * HOL/recdef can now declare non-recursive functions, with {} supplied as

337 the well-founded relation;

339 * HOL/Set INCOMPATIBILITY: the complement of set A is now written -A instead of

340 Compl A. The "Compl" syntax remains available as input syntax for this

341 release ONLY.

343 * HOL/Update: new theory of function updates:

344 f(a:=b) == %x. if x=a then b else f x

345 may also be iterated as in f(a:=b,c:=d,...);

347 * HOL/Vimage: new theory for inverse image of a function, syntax f-``B;

349 * HOL/List:

350 - new function list_update written xs[i:=v] that updates the i-th

351 list position. May also be iterated as in xs[i:=a,j:=b,...].

352 - new function `upt' written [i..j(] which generates the list

353 [i,i+1,...,j-1], i.e. the upper bound is excluded. To include the upper

354 bound write [i..j], which is a shorthand for [i..j+1(].

355 - new lexicographic orderings and corresponding wellfoundedness theorems.

357 * HOL/Arith:

358 - removed 'pred' (predecessor) function;

359 - generalized some theorems about n-1;

360 - many new laws about "div" and "mod";

361 - new laws about greatest common divisors (see theory ex/Primes);

363 * HOL/Relation: renamed the relational operator r^-1 "converse"

364 instead of "inverse";

366 * HOL/Induct/Multiset: a theory of multisets, including the wellfoundedness

367 of the multiset ordering;

369 * directory HOL/Real: a construction of the reals using Dedekind cuts

370 (not included by default);

372 * directory HOL/UNITY: Chandy and Misra's UNITY formalism;

374 * directory HOL/Hoare: a new version of Hoare logic which permits many-sorted

375 programs, i.e. different program variables may have different types.

377 * calling (stac rew i) now fails if "rew" has no effect on the goal

378 [previously, this check worked only if the rewrite rule was unconditional]

379 Now rew can involve either definitions or equalities (either == or =).

382 *** ZF ***

384 * theory Main includes everything; INCOMPATIBILITY: theory ZF.thy contains

385 only the theorems proved on ZF.ML;

387 * ZF INCOMPATIBILITY: rule `equals0D' is now a well-formed destruct rule;

388 It and 'sym RS equals0D' are now in the default claset, giving automatic

389 disjointness reasoning but breaking a few old proofs.

391 * ZF/Update: new theory of function updates

392 with default rewrite rule f(x:=y) ` z = if(z=x, y, f`z)

393 may also be iterated as in f(a:=b,c:=d,...);

395 * in let x=t in u(x), neither t nor u(x) has to be an FOL term.

397 * calling (stac rew i) now fails if "rew" has no effect on the goal

398 [previously, this check worked only if the rewrite rule was unconditional]

399 Now rew can involve either definitions or equalities (either == or =).

401 * case_tac provided for compatibility with HOL

402 (like the old excluded_middle_tac, but with subgoals swapped)

405 *** Internal programming interfaces ***

407 * Pure: several new basic modules made available for general use, see

408 also src/Pure/README;

410 * improved the theory data mechanism to support encapsulation (data

411 kind name replaced by private Object.kind, acting as authorization

412 key); new type-safe user interface via functor TheoryDataFun; generic

413 print_data function becomes basically useless;

415 * removed global_names compatibility flag -- all theory declarations

416 are qualified by default;

418 * module Pure/Syntax now offers quote / antiquote translation

419 functions (useful for Hoare logic etc. with implicit dependencies);

420 see HOL/ex/Antiquote for an example use;

422 * Simplifier now offers conversions (asm_)(full_)rewrite: simpset ->

423 cterm -> thm;

425 * new tactical CHANGED_GOAL for checking that a tactic modifies a

426 subgoal;

428 * Display.print_goals function moved to Locale.print_goals;

430 * standard print function for goals supports current_goals_markers

431 variable for marking begin of proof, end of proof, start of goal; the

432 default is ("", "", ""); setting current_goals_markers := ("<proof>",

433 "</proof>", "<goal>") causes SGML like tagged proof state printing,

434 for example;

438 New in Isabelle98 (January 1998)

439 --------------------------------

441 *** Overview of INCOMPATIBILITIES (see below for more details) ***

443 * changed lexical syntax of terms / types: dots made part of long

444 identifiers, e.g. "%x.x" no longer possible, should be "%x. x";

446 * simpset (and claset) reference variable replaced by functions

447 simpset / simpset_ref;

449 * no longer supports theory aliases (via merge) and non-trivial

450 implicit merge of thms' signatures;

452 * most internal names of constants changed due to qualified names;

454 * changed Pure/Sequence interface (see Pure/seq.ML);

457 *** General Changes ***

459 * hierachically structured name spaces (for consts, types, axms, thms

460 etc.); new lexical class 'longid' (e.g. Foo.bar.x) may render much of

461 old input syntactically incorrect (e.g. "%x.x"); COMPATIBILITY:

462 isatool fixdots ensures space after dots (e.g. "%x. x"); set

463 long_names for fully qualified output names; NOTE: ML programs

464 (special tactics, packages etc.) referring to internal names may have

465 to be adapted to cope with fully qualified names; in case of severe

466 backward campatibility problems try setting 'global_names' at compile

467 time to have enrything declared within a flat name space; one may also

468 fine tune name declarations in theories via the 'global' and 'local'

469 section;

471 * reimplemented the implicit simpset and claset using the new anytype

472 data filed in signatures; references simpset:simpset ref etc. are

473 replaced by functions simpset:unit->simpset and

474 simpset_ref:unit->simpset ref; COMPATIBILITY: use isatool fixclasimp

475 to patch your ML files accordingly;

477 * HTML output now includes theory graph data for display with Java

478 applet or isatool browser; data generated automatically via isatool

479 usedir (see -i option, ISABELLE_USEDIR_OPTIONS);

481 * defs may now be conditional; improved rewrite_goals_tac to handle

482 conditional equations;

484 * defs now admits additional type arguments, using TYPE('a) syntax;

486 * theory aliases via merge (e.g. M=A+B+C) no longer supported, always

487 creates a new theory node; implicit merge of thms' signatures is

488 restricted to 'trivial' ones; COMPATIBILITY: one may have to use

489 transfer:theory->thm->thm in (rare) cases;

491 * improved handling of draft signatures / theories; draft thms (and

492 ctyps, cterms) are automatically promoted to real ones;

494 * slightly changed interfaces for oracles: admit many per theory, named

495 (e.g. oracle foo = mlfun), additional name argument for invoke_oracle;

497 * print_goals: optional output of const types (set show_consts and

498 show_types);

500 * improved output of warnings (###) and errors (***);

502 * subgoal_tac displays a warning if the new subgoal has type variables;

504 * removed old README and Makefiles;

506 * replaced print_goals_ref hook by print_current_goals_fn and result_error_fn;

508 * removed obsolete init_pps and init_database;

510 * deleted the obsolete tactical STATE, which was declared by

511 fun STATE tacfun st = tacfun st st;

513 * cd and use now support path variables, e.g. $ISABELLE_HOME, or ~

514 (which abbreviates $HOME);

516 * changed Pure/Sequence interface (see Pure/seq.ML); COMPATIBILITY:

517 use isatool fixseq to adapt your ML programs (this works for fully

518 qualified references to the Sequence structure only!);

520 * use_thy no longer requires writable current directory; it always

521 reloads .ML *and* .thy file, if either one is out of date;

524 *** Classical Reasoner ***

526 * Clarify_tac, clarify_tac, clarify_step_tac, Clarify_step_tac: new

527 tactics that use classical reasoning to simplify a subgoal without

528 splitting it into several subgoals;

530 * Safe_tac: like safe_tac but uses the default claset;

533 *** Simplifier ***

535 * added simplification meta rules:

536 (asm_)(full_)simplify: simpset -> thm -> thm;

538 * simplifier.ML no longer part of Pure -- has to be loaded by object

539 logics (again);

541 * added prems argument to simplification procedures;

543 * HOL, FOL, ZF: added infix function `addsplits':

544 instead of `<simpset> setloop (split_tac <thms>)'

545 you can simply write `<simpset> addsplits <thms>'

548 *** Syntax ***

550 * TYPE('a) syntax for type reflection terms;

552 * no longer handles consts with name "" -- declare as 'syntax' instead;

554 * pretty printer: changed order of mixfix annotation preference (again!);

556 * Pure: fixed idt/idts vs. pttrn/pttrns syntactic categories;

559 *** HOL ***

561 * HOL: there is a new splitter `split_asm_tac' that can be used e.g.

562 with `addloop' of the simplifier to faciliate case splitting in premises.

564 * HOL/TLA: Stephan Merz's formalization of Lamport's Temporal Logic of Actions;

566 * HOL/Auth: new protocol proofs including some for the Internet

567 protocol TLS;

569 * HOL/Map: new theory of `maps' a la VDM;

571 * HOL/simplifier: simplification procedures nat_cancel_sums for

572 cancelling out common nat summands from =, <, <= (in)equalities, or

573 differences; simplification procedures nat_cancel_factor for

574 cancelling common factor from =, <, <= (in)equalities over natural

575 sums; nat_cancel contains both kinds of procedures, it is installed by

576 default in Arith.thy -- this COULD MAKE EXISTING PROOFS FAIL;

578 * HOL/simplifier: terms of the form

579 `? x. P1(x) & ... & Pn(x) & x=t & Q1(x) & ... Qn(x)' (or t=x)

580 are rewritten to

581 `P1(t) & ... & Pn(t) & Q1(t) & ... Qn(t)',

582 and those of the form

583 `! x. P1(x) & ... & Pn(x) & x=t & Q1(x) & ... Qn(x) --> R(x)' (or t=x)

584 are rewritten to

585 `P1(t) & ... & Pn(t) & Q1(t) & ... Qn(t) --> R(t)',

587 * HOL/datatype

588 Each datatype `t' now comes with a theorem `split_t_case' of the form

590 P(t_case f1 ... fn x) =

591 ( (!y1 ... ym1. x = C1 y1 ... ym1 --> P(f1 y1 ... ym1)) &

592 ...

593 (!y1 ... ymn. x = Cn y1 ... ymn --> P(f1 y1 ... ymn))

594 )

596 and a theorem `split_t_case_asm' of the form

598 P(t_case f1 ... fn x) =

599 ~( (? y1 ... ym1. x = C1 y1 ... ym1 & ~P(f1 y1 ... ym1)) |

600 ...

601 (? y1 ... ymn. x = Cn y1 ... ymn & ~P(f1 y1 ... ymn))

602 )

603 which can be added to a simpset via `addsplits'. The existing theorems

604 expand_list_case and expand_option_case have been renamed to

605 split_list_case and split_option_case.

607 * HOL/Arithmetic:

608 - `pred n' is automatically converted to `n-1'.

609 Users are strongly encouraged not to use `pred' any longer,

610 because it will disappear altogether at some point.

611 - Users are strongly encouraged to write "0 < n" rather than

612 "n ~= 0". Theorems and proof tools have been modified towards this

613 `standard'.

615 * HOL/Lists:

616 the function "set_of_list" has been renamed "set" (and its theorems too);

617 the function "nth" now takes its arguments in the reverse order and

618 has acquired the infix notation "!" as in "xs!n".

620 * HOL/Set: UNIV is now a constant and is no longer translated to Compl{};

622 * HOL/Set: The operator (UN x.B x) now abbreviates (UN x:UNIV. B x) and its

623 specialist theorems (like UN1_I) are gone. Similarly for (INT x.B x);

625 * HOL/record: extensible records with schematic structural subtyping

626 (single inheritance); EXPERIMENTAL version demonstrating the encoding,

627 still lacks various theorems and concrete record syntax;

630 *** HOLCF ***

632 * removed "axioms" and "generated by" sections;

634 * replaced "ops" section by extended "consts" section, which is capable of

635 handling the continuous function space "->" directly;

637 * domain package:

638 . proves theorems immediately and stores them in the theory,

639 . creates hierachical name space,

640 . now uses normal mixfix annotations (instead of cinfix...),

641 . minor changes to some names and values (for consistency),

642 . e.g. cases -> casedist, dists_eq -> dist_eqs, [take_lemma] -> take_lemmas,

643 . separator between mutual domain defs: changed "," to "and",

644 . improved handling of sort constraints; now they have to

645 appear on the left-hand side of the equations only;

647 * fixed LAM <x,y,zs>.b syntax;

649 * added extended adm_tac to simplifier in HOLCF -- can now discharge

650 adm (%x. P (t x)), where P is chainfinite and t continuous;

653 *** FOL and ZF ***

655 * FOL: there is a new splitter `split_asm_tac' that can be used e.g.

656 with `addloop' of the simplifier to faciliate case splitting in premises.

658 * qed_spec_mp, qed_goal_spec_mp, qed_goalw_spec_mp are available, as

659 in HOL, they strip ALL and --> from proved theorems;

663 New in Isabelle94-8 (May 1997)

664 ------------------------------

666 *** General Changes ***

668 * new utilities to build / run / maintain Isabelle etc. (in parts

669 still somewhat experimental); old Makefiles etc. still functional;

671 * new 'Isabelle System Manual';

673 * INSTALL text, together with ./configure and ./build scripts;

675 * reimplemented type inference for greater efficiency, better error

676 messages and clean internal interface;

678 * prlim command for dealing with lots of subgoals (an easier way of

679 setting goals_limit);

682 *** Syntax ***

684 * supports alternative (named) syntax tables (parser and pretty

685 printer); internal interface is provided by add_modesyntax(_i);

687 * Pure, FOL, ZF, HOL, HOLCF now support symbolic input and output; to

688 be used in conjunction with the Isabelle symbol font; uses the

689 "symbols" syntax table;

691 * added token_translation interface (may translate name tokens in

692 arbitrary ways, dependent on their type (free, bound, tfree, ...) and

693 the current print_mode); IMPORTANT: user print translation functions

694 are responsible for marking newly introduced bounds

695 (Syntax.mark_boundT);

697 * token translations for modes "xterm" and "xterm_color" that display

698 names in bold, underline etc. or colors (which requires a color

699 version of xterm);

701 * infixes may now be declared with names independent of their syntax;

703 * added typed_print_translation (like print_translation, but may

704 access type of constant);

707 *** Classical Reasoner ***

709 Blast_tac: a new tactic! It is often more powerful than fast_tac, but has

710 some limitations. Blast_tac...

711 + ignores addss, addbefore, addafter; this restriction is intrinsic

712 + ignores elimination rules that don't have the correct format

713 (the conclusion MUST be a formula variable)

714 + ignores types, which can make HOL proofs fail

715 + rules must not require higher-order unification, e.g. apply_type in ZF

716 [message "Function Var's argument not a bound variable" relates to this]

717 + its proof strategy is more general but can actually be slower

719 * substitution with equality assumptions no longer permutes other

720 assumptions;

722 * minor changes in semantics of addafter (now called addaltern); renamed

723 setwrapper to setWrapper and compwrapper to compWrapper; added safe wrapper

724 (and access functions for it);

726 * improved combination of classical reasoner and simplifier:

727 + functions for handling clasimpsets

728 + improvement of addss: now the simplifier is called _after_ the

729 safe steps.

730 + safe variant of addss called addSss: uses safe simplifications

731 _during_ the safe steps. It is more complete as it allows multiple

732 instantiations of unknowns (e.g. with slow_tac).

734 *** Simplifier ***

736 * added interface for simplification procedures (functions that

737 produce *proven* rewrite rules on the fly, depending on current

738 redex);

740 * ordering on terms as parameter (used for ordered rewriting);

742 * new functions delcongs, deleqcongs, and Delcongs. richer rep_ss;

744 * the solver is now split into a safe and an unsafe part.

745 This should be invisible for the normal user, except that the

746 functions setsolver and addsolver have been renamed to setSolver and

747 addSolver; added safe_asm_full_simp_tac;

750 *** HOL ***

752 * a generic induction tactic `induct_tac' which works for all datatypes and

753 also for type `nat';

755 * a generic case distinction tactic `exhaust_tac' which works for all

756 datatypes and also for type `nat';

758 * each datatype comes with a function `size';

760 * patterns in case expressions allow tuple patterns as arguments to

761 constructors, for example `case x of [] => ... | (x,y,z)#ps => ...';

763 * primrec now also works with type nat;

765 * recdef: a new declaration form, allows general recursive functions to be

766 defined in theory files. See HOL/ex/Fib, HOL/ex/Primes, HOL/Subst/Unify.

768 * the constant for negation has been renamed from "not" to "Not" to

769 harmonize with FOL, ZF, LK, etc.;

771 * HOL/ex/LFilter theory of a corecursive "filter" functional for

772 infinite lists;

774 * HOL/Modelcheck demonstrates invocation of model checker oracle;

776 * HOL/ex/Ring.thy declares cring_simp, which solves equational

777 problems in commutative rings, using axiomatic type classes for + and *;

779 * more examples in HOL/MiniML and HOL/Auth;

781 * more default rewrite rules for quantifiers, union/intersection;

783 * a new constant `arbitrary == @x.False';

785 * HOLCF/IOA replaces old HOL/IOA;

787 * HOLCF changes: derived all rules and arities

788 + axiomatic type classes instead of classes

789 + typedef instead of faking type definitions

790 + eliminated the internal constants less_fun, less_cfun, UU_fun, UU_cfun etc.

791 + new axclasses cpo, chfin, flat with flat < chfin < pcpo < cpo < po

792 + eliminated the types void, one, tr

793 + use unit lift and bool lift (with translations) instead of one and tr

794 + eliminated blift from Lift3.thy (use Def instead of blift)

795 all eliminated rules are derived as theorems --> no visible changes ;

798 *** ZF ***

800 * ZF now has Fast_tac, Simp_tac and Auto_tac. Union_iff is a now a default

801 rewrite rule; this may affect some proofs. eq_cs is gone but can be put back

802 as ZF_cs addSIs [equalityI];

806 New in Isabelle94-7 (November 96)

807 ---------------------------------

809 * allowing negative levels (as offsets) in prlev and choplev;

811 * super-linear speedup for large simplifications;

813 * FOL, ZF and HOL now use miniscoping: rewriting pushes

814 quantifications in as far as possible (COULD MAKE EXISTING PROOFS

815 FAIL); can suppress it using the command Delsimps (ex_simps @

816 all_simps); De Morgan laws are also now included, by default;

818 * improved printing of ==> : ~:

820 * new object-logic "Sequents" adds linear logic, while replacing LK

821 and Modal (thanks to Sara Kalvala);

823 * HOL/Auth: correctness proofs for authentication protocols;

825 * HOL: new auto_tac combines rewriting and classical reasoning (many

826 examples on HOL/Auth);

828 * HOL: new command AddIffs for declaring theorems of the form P=Q to

829 the rewriter and classical reasoner simultaneously;

831 * function uresult no longer returns theorems in "standard" format;

832 regain previous version by: val uresult = standard o uresult;

836 New in Isabelle94-6

837 -------------------

839 * oracles -- these establish an interface between Isabelle and trusted

840 external reasoners, which may deliver results as theorems;

842 * proof objects (in particular record all uses of oracles);

844 * Simp_tac, Fast_tac, etc. that refer to implicit simpset / claset;

846 * "constdefs" section in theory files;

848 * "primrec" section (HOL) no longer requires names;

850 * internal type "tactic" now simply "thm -> thm Sequence.seq";

854 New in Isabelle94-5

855 -------------------

857 * reduced space requirements;

859 * automatic HTML generation from theories;

861 * theory files no longer require "..." (quotes) around most types;

863 * new examples, including two proofs of the Church-Rosser theorem;

865 * non-curried (1994) version of HOL is no longer distributed;

869 New in Isabelle94-4

870 -------------------

872 * greatly reduced space requirements;

874 * theory files (.thy) no longer require \...\ escapes at line breaks;

876 * searchable theorem database (see the section "Retrieving theorems" on

877 page 8 of the Reference Manual);

879 * new examples, including Grabczewski's monumental case study of the

880 Axiom of Choice;

882 * The previous version of HOL renamed to Old_HOL;

884 * The new version of HOL (previously called CHOL) uses a curried syntax

885 for functions. Application looks like f a b instead of f(a,b);

887 * Mutually recursive inductive definitions finally work in HOL;

889 * In ZF, pattern-matching on tuples is now available in all abstractions and

890 translates to the operator "split";

894 New in Isabelle94-3

895 -------------------

897 * new infix operator, addss, allowing the classical reasoner to

898 perform simplification at each step of its search. Example:

899 fast_tac (cs addss ss)

901 * a new logic, CHOL, the same as HOL, but with a curried syntax

902 for functions. Application looks like f a b instead of f(a,b). Also pairs

903 look like (a,b) instead of <a,b>;

905 * PLEASE NOTE: CHOL will eventually replace HOL!

907 * In CHOL, pattern-matching on tuples is now available in all abstractions.

908 It translates to the operator "split". A new theory of integers is available;

910 * In ZF, integer numerals now denote two's-complement binary integers.

911 Arithmetic operations can be performed by rewriting. See ZF/ex/Bin.ML;

913 * Many new examples: I/O automata, Church-Rosser theorem, equivalents

914 of the Axiom of Choice;

918 New in Isabelle94-2

919 -------------------

921 * Significantly faster resolution;

923 * the different sections in a .thy file can now be mixed and repeated

924 freely;

926 * Database of theorems for FOL, HOL and ZF. New

927 commands including qed, qed_goal and bind_thm store theorems in the database.

929 * Simple database queries: return a named theorem (get_thm) or all theorems of

930 a given theory (thms_of), or find out what theory a theorem was proved in

931 (theory_of_thm);

933 * Bugs fixed in the inductive definition and datatype packages;

935 * The classical reasoner provides deepen_tac and depth_tac, making FOL_dup_cs

936 and HOL_dup_cs obsolete;

938 * Syntactic ambiguities caused by the new treatment of syntax in Isabelle94-1

939 have been removed;

941 * Simpler definition of function space in ZF;

943 * new results about cardinal and ordinal arithmetic in ZF;

945 * 'subtype' facility in HOL for introducing new types as subsets of existing

946 types;

949 $Id$