NEWS
author wenzelm
Sat Dec 09 18:05:36 2006 +0100 (2006-12-09)
changeset 21717 410ca6910f6f
parent 21647 fccafa917a68
child 21735 0c65e072f4be
permissions -rw-r--r--
added antiquotation abbrev;
     1 Isabelle NEWS -- history user-relevant changes
     2 ==============================================
     3 
     4 New in this Isabelle version
     5 ----------------------------
     6 
     7 *** General ***
     8 
     9 * Theory syntax: the header format ``theory A = B + C:'' has been
    10 discontinued in favour of ``theory A imports B C begin''.  Use isatool
    11 fixheaders to convert existing theory files.  INCOMPATIBILITY.
    12 
    13 * Theory syntax: the old non-Isar theory file format has been
    14 discontinued altogether.  Note that ML proof scripts may still be used
    15 with Isar theories; migration is usually quite simple with the ML
    16 function use_legacy_bindings.  INCOMPATIBILITY.
    17 
    18 * Theory syntax: some popular names (e.g. "class", "if", "fun") are
    19 now keywords.  INCOMPATIBILITY, use double quotes.
    20 
    21 * Legacy goal package: reduced interface to the bare minimum required
    22 to keep existing proof scripts running.  Most other user-level
    23 functions are now part of the OldGoals structure, which is *not* open
    24 by default (consider isatool expandshort before open OldGoals).
    25 Removed top_sg, prin, printyp, pprint_term/typ altogether, because
    26 these tend to cause confusion about the actual goal (!) context being
    27 used here, which is not necessarily the same as the_context().
    28 
    29 * Command 'find_theorems': support "*" wildcard in "name:" criterion.
    30 
    31 * The ``prems limit'' option (cf. ProofContext.prems_limit) is now -1
    32 by default, which means that "prems" (and also "fixed variables") are
    33 suppressed from proof state output.  Note that the ProofGeneral
    34 settings mechanism allows to change and save options persistently, but
    35 older versions of Isabelle will fail to start up if a negative prems
    36 limit is imposed.
    37 
    38 * Local theory targets may be specified by non-nested blocks of
    39 ``context/locale/class ... begin'' followed by ``end''.  The body may
    40 contain definitions, theorems etc., including any derived mechanism
    41 that has been implemented on top of these primitives.  This concept
    42 generalizes the existing ``theorem (in ...)'' towards more versatility
    43 and scalability.
    44 
    45 
    46 *** Document preparation ***
    47 
    48 * Added antiquotation @{theory name} which prints the given name,
    49 after checking that it refers to a valid ancestor theory in the
    50 current context.
    51 
    52 * Added antiquotations @{ML_type text} and @{ML_struct text} which
    53 check the given source text as ML type/structure, printing verbatim.
    54 
    55 * Added antiquotation @{abbrev "c args"} which prints the abbreviation
    56 "c args == rhs" given in the current context.  (Any number of
    57 arguments on the LHS may be given.)
    58 
    59 
    60 
    61 *** Pure ***
    62 
    63 * class_package.ML offers a combination of axclasses and locales to
    64 achieve Haskell-like type classes in Isabelle. See
    65 HOL/ex/Classpackage.thy for examples.
    66 
    67 * Yet another code generator framework allows to generate executable
    68 code for ML and Haskell (including "class"es). A short usage sketch:
    69 
    70     internal compilation:
    71         code_gen <list of constants (term syntax)> (SML #)
    72     writing SML code to a file:
    73         code_gen <list of constants (term syntax)> (SML <filename>)
    74     writing Haskell code to a bunch of files:
    75         code_gen <list of constants (term syntax)> (Haskell <filename>)
    76 
    77 Reasonable default setup of framework in HOL/Main.
    78 
    79 See HOL/ex/Code*.thy for examples.
    80 
    81 Theorem attributs for selecting and transforming function equations theorems:
    82 
    83     [code fun]:       select a theorem as function equation for a specific constant
    84     [code nofun]:     deselect a theorem as function equation for a specific constant
    85     [code inline]:    select an equation theorem for unfolding (inlining) in place
    86     [code noinline]:  deselect an equation theorem for unfolding (inlining) in place
    87 
    88 User-defined serializations (target in {SML, Haskell}):
    89 
    90     code_const <and-list of constants (term syntax)>
    91       {(target) <and-list of const target syntax>}+
    92 
    93     code_type <and-list of type constructors>
    94       {(target) <and-list of type target syntax>}+
    95 
    96     code_instance <and-list of instances>
    97       {(target)}+
    98         where instance ::= <type constructor> :: <class>
    99 
   100     code_class <and_list of classes>
   101       {(target) <and-list of class target syntax>}+
   102         where class target syntax ::= <class name> {where {<classop> == <target syntax>}+}?
   103 
   104 For code_instance and code_class, target SML is silently ignored.
   105 
   106 See HOL theories and HOL/ex/Code*.thy for usage examples.  Doc/Isar/Advanced/Codegen/
   107 provides a tutorial.
   108 
   109 * Command 'no_translations' removes translation rules from theory
   110 syntax.
   111 
   112 * Overloaded definitions are now actually checked for acyclic
   113 dependencies.  The overloading scheme is slightly more general than
   114 that of Haskell98, although Isabelle does not demand an exact
   115 correspondence to type class and instance declarations.
   116 INCOMPATIBILITY, use ``defs (unchecked overloaded)'' to admit more
   117 exotic versions of overloading -- at the discretion of the user!
   118 
   119 Polymorphic constants are represented via type arguments, i.e. the
   120 instantiation that matches an instance against the most general
   121 declaration given in the signature.  For example, with the declaration
   122 c :: 'a => 'a => 'a, an instance c :: nat => nat => nat is represented
   123 as c(nat).  Overloading is essentially simultaneous structural
   124 recursion over such type arguments.  Incomplete specification patterns
   125 impose global constraints on all occurrences, e.g. c('a * 'a) on the
   126 LHS means that more general c('a * 'b) will be disallowed on any RHS.
   127 Command 'print_theory' outputs the normalized system of recursive
   128 equations, see section "definitions".
   129 
   130 * Isar: improper proof element 'guess' is like 'obtain', but derives
   131 the obtained context from the course of reasoning!  For example:
   132 
   133   assume "EX x y. A x & B y"   -- "any previous fact"
   134   then guess x and y by clarify
   135 
   136 This technique is potentially adventurous, depending on the facts and
   137 proof tools being involved here.
   138 
   139 * Isar: known facts from the proof context may be specified as literal
   140 propositions, using ASCII back-quote syntax.  This works wherever
   141 named facts used to be allowed so far, in proof commands, proof
   142 methods, attributes etc.  Literal facts are retrieved from the context
   143 according to unification of type and term parameters.  For example,
   144 provided that "A" and "A ==> B" and "!!x. P x ==> Q x" are known
   145 theorems in the current context, then these are valid literal facts:
   146 `A` and `A ==> B` and `!!x. P x ==> Q x" as well as `P a ==> Q a` etc.
   147 
   148 There is also a proof method "fact" which does the same composition
   149 for explicit goal states, e.g. the following proof texts coincide with
   150 certain special cases of literal facts:
   151 
   152   have "A" by fact                 ==  note `A`
   153   have "A ==> B" by fact           ==  note `A ==> B`
   154   have "!!x. P x ==> Q x" by fact  ==  note `!!x. P x ==> Q x`
   155   have "P a ==> Q a" by fact       ==  note `P a ==> Q a`
   156 
   157 * Isar: ":" (colon) is no longer a symbolic identifier character in
   158 outer syntax.  Thus symbolic identifiers may be used without
   159 additional white space in declarations like this: ``assume *: A''.
   160 
   161 * Isar: 'print_facts' prints all local facts of the current context,
   162 both named and unnamed ones.
   163 
   164 * Isar: 'def' now admits simultaneous definitions, e.g.:
   165 
   166   def x == "t" and y == "u"
   167 
   168 * Isar: added command 'unfolding', which is structurally similar to
   169 'using', but affects both the goal state and facts by unfolding given
   170 rewrite rules.  Thus many occurrences of the 'unfold' method or
   171 'unfolded' attribute may be replaced by first-class proof text.
   172 
   173 * Isar: methods 'unfold' / 'fold', attributes 'unfolded' / 'folded',
   174 and command 'unfolding' now all support object-level equalities
   175 (potentially conditional).  The underlying notion of rewrite rule is
   176 analogous to the 'rule_format' attribute, but *not* that of the
   177 Simplifier (which is usually more generous).
   178 
   179 * Isar: the goal restriction operator [N] (default N = 1) evaluates a
   180 method expression within a sandbox consisting of the first N
   181 sub-goals, which need to exist.  For example, ``simp_all [3]''
   182 simplifies the first three sub-goals, while (rule foo, simp_all)[]
   183 simplifies all new goals that emerge from applying rule foo to the
   184 originally first one.
   185 
   186 * Isar: schematic goals are no longer restricted to higher-order
   187 patterns; e.g. ``lemma "?P(?x)" by (rule TrueI)'' now works as
   188 expected.
   189 
   190 * Isar: the conclusion of a long theorem statement is now either
   191 'shows' (a simultaneous conjunction, as before), or 'obtains'
   192 (essentially a disjunction of cases with local parameters and
   193 assumptions).  The latter allows to express general elimination rules
   194 adequately; in this notation common elimination rules look like this:
   195 
   196   lemma exE:    -- "EX x. P x ==> (!!x. P x ==> thesis) ==> thesis"
   197     assumes "EX x. P x"
   198     obtains x where "P x"
   199 
   200   lemma conjE:  -- "A & B ==> (A ==> B ==> thesis) ==> thesis"
   201     assumes "A & B"
   202     obtains A and B
   203 
   204   lemma disjE:  -- "A | B ==> (A ==> thesis) ==> (B ==> thesis) ==> thesis"
   205     assumes "A | B"
   206     obtains
   207       A
   208     | B
   209 
   210 The subsequent classical rules even refer to the formal "thesis"
   211 explicitly:
   212 
   213   lemma classical:     -- "(~ thesis ==> thesis) ==> thesis"
   214     obtains "~ thesis"
   215 
   216   lemma Peirce's_Law:  -- "((thesis ==> something) ==> thesis) ==> thesis"
   217     obtains "thesis ==> something"
   218 
   219 The actual proof of an 'obtains' statement is analogous to that of the
   220 Isar proof element 'obtain', only that there may be several cases.
   221 Optional case names may be specified in parentheses; these will be
   222 available both in the present proof and as annotations in the
   223 resulting rule, for later use with the 'cases' method (cf. attribute
   224 case_names).
   225 
   226 * Isar: the assumptions of a long theorem statement are available as
   227 "assms" fact in the proof context.  This is more appropriate than the
   228 (historical) "prems", which refers to all assumptions of the current
   229 context, including those from the target locale, proof body etc.
   230 
   231 * Isar: 'print_statement' prints theorems from the current theory or
   232 proof context in long statement form, according to the syntax of a
   233 top-level lemma.
   234 
   235 * Isar: 'obtain' takes an optional case name for the local context
   236 introduction rule (default "that").
   237 
   238 * Isar: removed obsolete 'concl is' patterns.  INCOMPATIBILITY, use
   239 explicit (is "_ ==> ?foo") in the rare cases where this still happens
   240 to occur.
   241 
   242 * Pure: syntax "CONST name" produces a fully internalized constant
   243 according to the current context.  This is particularly useful for
   244 syntax translations that should refer to internal constant
   245 representations independently of name spaces.
   246 
   247 * Pure: syntax constant for foo (binder "FOO ") is called "foo_binder"
   248 instead of "FOO ". This allows multiple binder declarations to coexist
   249 in the same context.  INCOMPATIBILITY.
   250 
   251 * Isar/locales: 'notation' provides a robust interface to the 'syntax'
   252 primitive that also works in a locale context (both for constants and
   253 fixed variables).  Type declaration and internal syntactic
   254 representation of given constants retrieved from the context.
   255 
   256 * Isar/locales: new derived specification elements 'axiomatization',
   257 'definition', 'abbreviation', which support type-inference, admit
   258 object-level specifications (equality, equivalence).  See also the
   259 isar-ref manual.  Examples:
   260 
   261   axiomatization
   262     eq  (infix "===" 50) where
   263     eq_refl: "x === x" and eq_subst: "x === y ==> P x ==> P y"
   264 
   265   definition "f x y = x + y + 1"
   266   definition g where "g x = f x x"
   267 
   268   abbreviation
   269     neq  (infix "=!=" 50) where
   270     "x =!= y == ~ (x === y)"
   271 
   272 These specifications may be also used in a locale context.  Then the
   273 constants being introduced depend on certain fixed parameters, and the
   274 constant name is qualified by the locale base name.  An internal
   275 abbreviation takes care for convenient input and output, making the
   276 parameters implicit and using the original short name.  See also
   277 HOL/ex/Abstract_NAT.thy for an example of deriving polymorphic
   278 entities from a monomorphic theory.
   279 
   280 Presently, abbreviations are only available 'in' a target locale, but
   281 not inherited by general import expressions.  Also note that
   282 'abbreviation' may be used as a type-safe replacement for 'syntax' +
   283 'translations' in common applications.
   284 
   285 Concrete syntax is attached to specified constants in internal form,
   286 independently of name spaces.  The parse tree representation is
   287 slightly different -- use 'notation' instead of raw 'syntax', and
   288 'translations' with explicit "CONST" markup to accommodate this.
   289 
   290 * Isar/locales: improved parameter handling:
   291 - use of locales "var" and "struct" no longer necessary;
   292 - parameter renamings are no longer required to be injective.
   293   This enables, for example, to define a locale for endomorphisms thus:
   294   locale endom = homom mult mult h.
   295 
   296 * Isar/locales: changed the way locales with predicates are defined.
   297 Instead of accumulating the specification, the imported expression is
   298 now an interpretation.
   299 INCOMPATIBILITY: different normal form of locale expressions.
   300 In particular, in interpretations of locales with predicates,
   301 goals repesenting already interpreted fragments are not removed
   302 automatically.  Use methods `intro_locales' and `unfold_locales'; see below.
   303 
   304 * Isar/locales: new methods `intro_locales' and `unfold_locales' provide
   305 backward reasoning on locales predicates.  The methods are aware of
   306 interpretations and discharge corresponding goals.  `intro_locales' is
   307 less aggressive then `unfold_locales' and does not unfold predicates to
   308 assumptions.
   309 
   310 * Isar/locales: the order in which locale fragments are accumulated
   311 has changed.  This enables to override declarations from fragments
   312 due to interpretations -- for example, unwanted simp rules.
   313 
   314 * Provers/induct: improved internal context management to support
   315 local fixes and defines on-the-fly.  Thus explicit meta-level
   316 connectives !! and ==> are rarely required anymore in inductive goals
   317 (using object-logic connectives for this purpose has been long
   318 obsolete anyway).  The subsequent proof patterns illustrate advanced
   319 techniques of natural induction; general datatypes and inductive sets
   320 work analogously (see also src/HOL/Lambda for realistic examples).
   321 
   322 (1) This is how to ``strengthen'' an inductive goal wrt. certain
   323 parameters:
   324 
   325   lemma
   326     fixes n :: nat and x :: 'a
   327     assumes a: "A n x"
   328     shows "P n x"
   329     using a                     -- {* make induct insert fact a *}
   330   proof (induct n arbitrary: x) -- {* generalize goal to "!!x. A n x ==> P n x" *}
   331     case 0
   332     show ?case sorry
   333   next
   334     case (Suc n)
   335     note `!!x. A n x ==> P n x` -- {* induction hypothesis, according to induction rule *}
   336     note `A (Suc n) x`          -- {* induction premise, stemming from fact a *}
   337     show ?case sorry
   338   qed
   339 
   340 (2) This is how to perform induction over ``expressions of a certain
   341 form'', using a locally defined inductive parameter n == "a x"
   342 together with strengthening (the latter is usually required to get
   343 sufficiently flexible induction hypotheses):
   344 
   345   lemma
   346     fixes a :: "'a => nat"
   347     assumes a: "A (a x)"
   348     shows "P (a x)"
   349     using a
   350   proof (induct n == "a x" arbitrary: x)
   351     ...
   352 
   353 See also HOL/Isar_examples/Puzzle.thy for an application of the this
   354 particular technique.
   355 
   356 (3) This is how to perform existential reasoning ('obtains' or
   357 'obtain') by induction, while avoiding explicit object-logic
   358 encodings:
   359 
   360   lemma
   361     fixes n :: nat
   362     obtains x :: 'a where "P n x" and "Q n x"
   363   proof (induct n arbitrary: thesis)
   364     case 0
   365     obtain x where "P 0 x" and "Q 0 x" sorry
   366     then show thesis by (rule 0)
   367   next
   368     case (Suc n)
   369     obtain x where "P n x" and "Q n x" by (rule Suc.hyps)
   370     obtain x where "P (Suc n) x" and "Q (Suc n) x" sorry
   371     then show thesis by (rule Suc.prems)
   372   qed
   373 
   374 Here the 'arbitrary: thesis' specification essentially modifies the
   375 scope of the formal thesis parameter, in order to the get the whole
   376 existence statement through the induction as expected.
   377 
   378 * Provers/induct: mutual induction rules are now specified as a list
   379 of rule sharing the same induction cases.  HOL packages usually
   380 provide foo_bar.inducts for mutually defined items foo and bar
   381 (e.g. inductive sets or datatypes).  INCOMPATIBILITY, users need to
   382 specify mutual induction rules differently, i.e. like this:
   383 
   384   (induct rule: foo_bar.inducts)
   385   (induct set: foo bar)
   386   (induct type: foo bar)
   387 
   388 The ML function ProjectRule.projections turns old-style rules into the
   389 new format.
   390 
   391 * Provers/induct: improved handling of simultaneous goals.  Instead of
   392 introducing object-level conjunction, the statement is now split into
   393 several conclusions, while the corresponding symbolic cases are
   394 nested accordingly.  INCOMPATIBILITY, proofs need to be structured
   395 explicitly.  For example:
   396 
   397   lemma
   398     fixes n :: nat
   399     shows "P n" and "Q n"
   400   proof (induct n)
   401     case 0 case 1
   402     show "P 0" sorry
   403   next
   404     case 0 case 2
   405     show "Q 0" sorry
   406   next
   407     case (Suc n) case 1
   408     note `P n` and `Q n`
   409     show "P (Suc n)" sorry
   410   next
   411     case (Suc n) case 2
   412     note `P n` and `Q n`
   413     show "Q (Suc n)" sorry
   414   qed
   415 
   416 The split into subcases may be deferred as follows -- this is
   417 particularly relevant for goal statements with local premises.
   418 
   419   lemma
   420     fixes n :: nat
   421     shows "A n ==> P n" and "B n ==> Q n"
   422   proof (induct n)
   423     case 0
   424     {
   425       case 1
   426       note `A 0`
   427       show "P 0" sorry
   428     next
   429       case 2
   430       note `B 0`
   431       show "Q 0" sorry
   432     }
   433   next
   434     case (Suc n)
   435     note `A n ==> P n` and `B n ==> Q n`
   436     {
   437       case 1
   438       note `A (Suc n)`
   439       show "P (Suc n)" sorry
   440     next
   441       case 2
   442       note `B (Suc n)`
   443       show "Q (Suc n)" sorry
   444     }
   445   qed
   446 
   447 If simultaneous goals are to be used with mutual rules, the statement
   448 needs to be structured carefully as a two-level conjunction, using
   449 lists of propositions separated by 'and':
   450 
   451   lemma
   452     shows "a : A ==> P1 a"
   453           "a : A ==> P2 a"
   454       and "b : B ==> Q1 b"
   455           "b : B ==> Q2 b"
   456           "b : B ==> Q3 b"
   457   proof (induct set: A B)
   458 
   459 * Provers/induct: support coinduction as well.  See
   460 src/HOL/Library/Coinductive_List.thy for various examples.
   461 
   462 * Attribute "symmetric" produces result with standardized schematic
   463 variables (index 0).  Potential INCOMPATIBILITY.
   464 
   465 * Simplifier: by default the simplifier trace only shows top level rewrites
   466 now. That is, trace_simp_depth_limit is set to 1 by default. Thus there is
   467 less danger of being flooded by the trace. The trace indicates where parts
   468 have been suppressed.
   469   
   470 * Provers/classical: removed obsolete classical version of elim_format
   471 attribute; classical elim/dest rules are now treated uniformly when
   472 manipulating the claset.
   473 
   474 * Provers/classical: stricter checks to ensure that supplied intro,
   475 dest and elim rules are well-formed; dest and elim rules must have at
   476 least one premise.
   477 
   478 * Provers/classical: attributes dest/elim/intro take an optional
   479 weight argument for the rule (just as the Pure versions).  Weights are
   480 ignored by automated tools, but determine the search order of single
   481 rule steps.
   482 
   483 * Syntax: input syntax now supports dummy variable binding "%_. b",
   484 where the body does not mention the bound variable.  Note that dummy
   485 patterns implicitly depend on their context of bounds, which makes
   486 "{_. _}" match any set comprehension as expected.  Potential
   487 INCOMPATIBILITY -- parse translations need to cope with syntactic
   488 constant "_idtdummy" in the binding position.
   489 
   490 * Syntax: removed obsolete syntactic constant "_K" and its associated
   491 parse translation.  INCOMPATIBILITY -- use dummy abstraction instead,
   492 for example "A -> B" => "Pi A (%_. B)".
   493 
   494 * Pure: 'class_deps' command visualizes the subclass relation, using
   495 the graph browser tool.
   496 
   497 * Pure: 'print_theory' now suppresses entities with internal name
   498 (trailing "_") by default; use '!' option for full details.
   499 
   500 
   501 *** HOL ***
   502 
   503 * New syntactic class "size"; overloaded constant "size" now
   504 has type "'a::size ==> bool"
   505 
   506 * Constants "Divides.op div", "Divides.op mod" and "Divides.op dvd" no named
   507   "Divides.div", "Divides.mod" and "Divides.dvd"
   508   INCOMPATIBILITY for ML code directly refering to constant names.
   509 
   510 * Replaced "auto_term" by the conceptually simpler method "relation",
   511 which just applies the instantiated termination rule with no further
   512 simplifications.
   513 INCOMPATIBILITY: 
   514 Replace 
   515         termination by (auto_term "MYREL")
   516 with 
   517         termination by (relation "MYREL") auto
   518 
   519 * Automated termination proofs "by lexicographic_order" are now
   520 included in the abbreviated function command "fun". No explicit
   521 "termination" command is necessary anymore.
   522 INCOMPATIBILITY: If a "fun"-definition cannot be proved terminating by
   523 a lexicographic size order, then the command fails. Use the expanded
   524 version "function" for these cases.
   525 
   526 * New method "lexicographic_order" automatically synthesizes
   527 termination relations as lexicographic combinations of size measures. 
   528 Usage for (function package) termination proofs:
   529 
   530 termination 
   531 by lexicographic_order
   532 
   533 * Records: generalised field-update to take a function on the field 
   534 rather than the new value: r(|A := x|) is translated to A_update (K x) r
   535 The K-combinator that is internally used is called K_record.
   536 INCOMPATIBILITY: Usage of the plain update functions has to be
   537 adapted.
   538  
   539 * axclass "semiring_0" now contains annihilation axioms 
   540 ("x * 0 = 0","0 * x = 0"), which are required for a semiring. Richer 
   541 structures do not inherit from semiring_0 anymore, because this property 
   542 is a theorem there, not an axiom.
   543 INCOMPATIBILITY: In instances of semiring_0, there is more to prove, but 
   544 this is mostly trivial.
   545 
   546 * axclass "recpower" was generalized to arbitrary monoids, not just 
   547 commutative semirings.
   548 INCOMPATIBILITY: If you use recpower and need commutativity or a semiring
   549 property, add the corresponding classes.
   550 
   551 * Locale Lattic_Locales.partial_order changed (to achieve consistency with
   552   axclass order):
   553   - moved to Orderings.partial_order
   554   - additional parameter ``less''
   555   INCOMPATIBILITY.
   556 
   557 * Constant "List.list_all2" in List.thy now uses authentic syntax.
   558 INCOMPATIBILITY: translations containing list_all2 may go wrong. On Isar
   559 level, use abbreviations instead.
   560 
   561 * Constant "List.op mem" in List.thy now has proper name: "List.memberl"
   562 INCOMPATIBILITY: rarely occuring name references (e.g. ``List.op mem.simps'')
   563 require renaming (e.g. ``List.memberl.simps'').
   564 
   565 * Renamed constants in HOL.thy:
   566     0      ~> HOL.zero
   567     1      ~> HOL.one
   568 INCOMPATIBILITY: ML code directly refering to constant names may need adaption
   569 This in general only affects hand-written proof tactics, simprocs and so on.
   570 
   571 * New theory Code_Generator providing class 'eq',
   572 allowing for code generation with polymorphic equality.
   573 
   574 * Numeral syntax: type 'bin' which was a mere type copy of 'int' has been
   575 abandoned in favour of plain 'int'. INCOMPATIBILITY -- Significant changes
   576 for setting up numeral syntax for types:
   577 
   578   - new constants Numeral.pred and Numeral.succ instead
   579       of former Numeral.bin_pred and Numeral.bin_succ.
   580   - Use integer operations instead of bin_add, bin_mult and so on.
   581   - Numeral simplification theorems named Numeral.numeral_simps instead of Bin_simps.
   582   - ML structure Bin_Simprocs now named Int_Numeral_Base_Simprocs.
   583 
   584 See HOL/Integ/IntArith.thy for an example setup.
   585 
   586 * New top level command 'normal_form' computes the normal form of a term
   587 that may contain free variables. For example 'normal_form "rev[a,b,c]"'
   588 prints '[b,c,a]'. This command is suitable for heavy-duty computations
   589 because the functions are compiled to ML first.
   590 INCOMPATIBILITY: new keywords 'normal_form' must quoted when used as
   591 an identifier.
   592 
   593 * Alternative iff syntax "A <-> B" for equality on bool (with priority
   594 25 like -->); output depends on the "iff" print_mode, the default is
   595 "A = B" (with priority 50).
   596 
   597 * Renamed constants in HOL.thy and Orderings.thy:
   598     op +   ~> HOL.plus
   599     op -   ~> HOL.minus
   600     uminus ~> HOL.uminus
   601     op *   ~> HOL.times
   602     op <   ~> Orderings.less
   603     op <=  ~> Orderings.less_eq
   604 
   605 Adaptions may be required in the following cases:
   606 
   607 a) User-defined constants using any of the names "plus", "minus", "times",
   608 "less" or "less_eq". The standard syntax translations for "+", "-" and "*"
   609 may go wrong.
   610 INCOMPATIBILITY: use more specific names.
   611 
   612 b) Variables named "plus", "minus", "times", "less", "less_eq"
   613 INCOMPATIBILITY: use more specific names.
   614 
   615 c) Permutative equations (e.g. "a + b = b + a")
   616 Since the change of names also changes the order of terms, permutative
   617 rewrite rules may get applied in a different order. Experience shows that
   618 this is rarely the case (only two adaptions in the whole Isabelle
   619 distribution).
   620 INCOMPATIBILITY: rewrite proofs
   621 
   622 d) ML code directly refering to constant names
   623 This in general only affects hand-written proof tactics, simprocs and so on.
   624 INCOMPATIBILITY: grep your sourcecode and replace names.
   625 
   626 * Relations less (<) and less_eq (<=) are also available on type bool.
   627 Modified syntax to disallow nesting without explicit parentheses,
   628 e.g. "(x < y) < z" or "x < (y < z)", but NOT "x < y < z".
   629 
   630 * "LEAST x:A. P" expands to "LEAST x. x:A & P" (input only).
   631 
   632 * Relation composition operator "op O" now has precedence 75 and binds
   633 stronger than union and intersection. INCOMPATIBILITY.
   634 
   635 * The old set interval syntax "{m..n(}" (and relatives) has been removed.
   636   Use "{m..<n}" (and relatives) instead.
   637 
   638 * In the context of the assumption "~(s = t)" the Simplifier rewrites
   639 "t = s" to False (by simproc "neq_simproc").  For backward
   640 compatibility this can be disabled by ML "reset use_neq_simproc".
   641 
   642 * "m dvd n" where m and n are numbers is evaluated to True/False by simp.
   643 
   644 * Theorem Cons_eq_map_conv no longer has attribute `simp'.
   645 
   646 * Theorem setsum_mult renamed to setsum_right_distrib.
   647 
   648 * Prefer ex1I over ex_ex1I in single-step reasoning, e.g. by the
   649 'rule' method.
   650 
   651 * Tactics 'sat' and 'satx' reimplemented, several improvements: goals
   652 no longer need to be stated as "<prems> ==> False", equivalences (i.e.
   653 "=" on type bool) are handled, variable names of the form "lit_<n>"
   654 are no longer reserved, significant speedup.
   655 
   656 * Tactics 'sat' and 'satx' can now replay MiniSat proof traces.  zChaff is
   657 still supported as well.
   658 
   659 * inductive and datatype: provide projections of mutual rules, bundled
   660 as foo_bar.inducts;
   661 
   662 * Library: moved theories Parity, GCD, Binomial, Infinite_Set to Library.
   663 
   664 * Library: moved theory Accessible_Part to main HOL.
   665 
   666 * Library: added theory Coinductive_List of potentially infinite lists
   667 as greatest fixed-point.
   668 
   669 * Library: added theory AssocList which implements (finite) maps as
   670 association lists.
   671 
   672 * New proof method "evaluation" for efficiently solving a goal (i.e. a
   673 boolean expression) by compiling it to ML. The goal is "proved" (via
   674 the oracle "Evaluation") if it evaluates to True.
   675 
   676 * Linear arithmetic now splits certain operators (e.g. min, max, abs)
   677 also when invoked by the simplifier.  This results in the simplifier
   678 being more powerful on arithmetic goals.  INCOMPATIBILITY.  Set
   679 fast_arith_split_limit to 0 to obtain the old behavior.
   680 
   681 * Support for hex (0x20) and binary (0b1001) numerals. 
   682 
   683 * New method: reify eqs (t), where eqs are equations for an
   684 interpretation I :: 'a list => 'b => 'c and t::'c is an optional
   685 parameter, computes a term s::'b and a list xs::'a list and proves the
   686 theorem I xs s = t. This is also known as reification or quoting. The
   687 resulting theorem is applied to the subgoal to substitute t with I xs
   688 s.  If t is omitted, the subgoal itself is reified.
   689 
   690 * New method: reflection corr_thm eqs (t). The parameters eqs and (t)
   691 are as explained above. corr_thm is a theorem for I vs (f t) = I vs t,
   692 where f is supposed to be a computable function (in the sense of code
   693 generattion). The method uses reify to compute s and xs as above then
   694 applies corr_thm and uses normalization by evaluation to "prove" f s =
   695 r and finally gets the theorem t = r, which is again applied to the
   696 subgoal. An Example is available in HOL/ex/ReflectionEx.thy.
   697 
   698 * Reflection: Automatic refification now handels binding, an example
   699 is available in HOL/ex/ReflectionEx.thy
   700 
   701 
   702 *** HOL-Algebra ***
   703 
   704 * Formalisation of ideals and the quotient construction over rings.
   705 
   706 * Order and lattice theory no longer based on records.
   707 INCOMPATIBILITY.
   708 
   709 * Method algebra is now set up via an attribute.  For examples see
   710 CRing.thy.  INCOMPATIBILITY: the method is now weaker on combinations
   711 of algebraic structures.
   712 
   713 * Renamed `CRing.thy' to `Ring.thy'.  INCOMPATIBILITY.
   714 
   715 
   716 *** HOL-Complex ***
   717 
   718 * Theory Real: new method ferrack implements quantifier elimination
   719 for linear arithmetic over the reals. The quantifier elimination
   720 feature is used only for decision, for compatibility with arith. This
   721 means a goal is either solved or left unchanged, no simplification.
   722 
   723 
   724 *** ML ***
   725 
   726 * Pure/table:
   727 
   728 Function `...tab.foldl` removed.
   729 INCOMPATIBILITY: use `...tabfold` instead
   730 
   731 * Pure/library:
   732 
   733 `gen_rem(s)` abandoned in favour of `remove` / `subtract`.
   734 INCOMPATIBILITY:
   735 rewrite "gen_rem eq (xs, x)" to "remove (eq o swap) x xs"
   736 rewrite "gen_rems eq (xs, ys)" to "subtract (eq o swap) ys xs"
   737 drop "swap" if "eq" is symmetric.
   738 
   739 * Pure/library:
   740 
   741 infixes `ins` `ins_string` `ins_int` have been abandoned in favour of `insert`.
   742 INCOMPATIBILITY: rewrite "x ins(_...) xs" to "insert (op =) x xs"
   743 
   744 * Pure/General/susp.ML:
   745 
   746 New module for delayed evaluations.
   747 
   748 * Pure/library:
   749 
   750 Semantically identical functions "equal_list" and "eq_list" have been
   751 unified to "eq_list".
   752 
   753 * Pure/library:
   754 
   755   val burrow: ('a list -> 'b list) -> 'a list list -> 'b list list
   756   val fold_burrow: ('a list -> 'c -> 'b list * 'd) -> 'a list list -> 'c -> 'b list list * 'd
   757 
   758 The semantics of "burrow" is: "take a function with *simulatanously*
   759 transforms a list of value, and apply it *simulatanously* to a list of
   760 list of values of the appropriate type". Confer this with "map" which
   761 would *not* apply its argument function simulatanously but in
   762 sequence. "fold_burrow" has an additional context.
   763 
   764 Both actually avoid the usage of "unflat" since they hide away
   765 "unflat" from the user.
   766 
   767 * Pure/library: functions map2 and fold2 with curried syntax for
   768 simultanous mapping and folding:
   769 
   770     val map2: ('a -> 'b -> 'c) -> 'a list -> 'b list -> 'c list
   771     val fold2: ('a -> 'b -> 'c -> 'c) -> 'a list -> 'b list -> 'c -> 'c
   772 
   773 * Pure/library: indexed lists - some functions in the Isabelle library
   774 treating lists over 'a as finite mappings from [0...n] to 'a have been
   775 given more convenient names and signatures reminiscent of similar
   776 functions for alists, tables, etc:
   777 
   778   val nth: 'a list -> int -> 'a 
   779   val nth_map: int -> ('a -> 'a) -> 'a list -> 'a list
   780   val fold_index: (int * 'a -> 'b -> 'b) -> 'a list -> 'b -> 'b
   781 
   782 Note that fold_index starts counting at index 0, not 1 like foldln
   783 used to.
   784 
   785 * Pure/library: general ``divide_and_conquer'' combinator on lists.
   786 
   787 * Pure/General/name_mangler.ML provides a functor for generic name
   788 mangling (bijective mapping from expression values to strings).
   789 
   790 * Pure/General/rat.ML implements rational numbers.
   791 
   792 * Pure/General/table.ML: the join operations now works via exceptions
   793 DUP/SAME instead of type option.  This is simpler in simple cases, and
   794 admits slightly more efficient complex applications.
   795 
   796 * Pure: datatype Context.generic joins theory/Proof.context and
   797 provides some facilities for code that works in either kind of
   798 context, notably GenericDataFun for uniform theory and proof data.
   799 
   800 * Pure: 'advanced' translation functions (parse_translation etc.) now
   801 use Context.generic instead of just theory.
   802 
   803 * Pure: simplified internal attribute type, which is now always
   804 Context.generic * thm -> Context.generic * thm.  Global (theory)
   805 vs. local (Proof.context) attributes have been discontinued, while
   806 minimizing code duplication.  Thm.rule_attribute and
   807 Thm.declaration_attribute build canonical attributes; see also
   808 structure Context for further operations on Context.generic, notably
   809 GenericDataFun.  INCOMPATIBILITY, need to adapt attribute type
   810 declarations and definitions.
   811 
   812 * Pure/kernel: consts certification ignores sort constraints given in
   813 signature declarations.  (This information is not relevant to the
   814 logic, but only for type inference.)  IMPORTANT INTERNAL CHANGE.
   815 
   816 * Pure: axiomatic type classes are now purely definitional, with
   817 explicit proofs of class axioms and super class relations performed
   818 internally.  See Pure/axclass.ML for the main internal interfaces --
   819 notably AxClass.define_class supercedes AxClass.add_axclass, and
   820 AxClass.axiomatize_class/classrel/arity supercede
   821 Sign.add_classes/classrel/arities.
   822 
   823 * Pure/Isar: Args/Attrib parsers operate on Context.generic --
   824 global/local versions on theory vs. Proof.context have been
   825 discontinued; Attrib.syntax and Method.syntax have been adapted
   826 accordingly.  INCOMPATIBILITY, need to adapt parser expressions for
   827 attributes, methods, etc.
   828 
   829 * Pure: several functions of signature "... -> theory -> theory * ..."
   830 have been reoriented to "... -> theory -> ... * theory" in order to
   831 allow natural usage in combination with the ||>, ||>>, |-> and
   832 fold_map combinators.
   833 
   834 * Pure: official theorem names (closed derivations) and additional
   835 comments (tags) are now strictly separate.  Name hints -- which are
   836 maintained as tags -- may be attached any time without affecting the
   837 derivation.
   838 
   839 * Pure: primitive rule lift_rule now takes goal cterm instead of an
   840 actual goal state (thm).  Use Thm.lift_rule (Thm.cprem_of st i) to
   841 achieve the old behaviour.
   842 
   843 * Pure: the "Goal" constant is now called "prop", supporting a
   844 slightly more general idea of ``protecting'' meta-level rule
   845 statements.
   846 
   847 * Pure: Logic.(un)varify only works in a global context, which is now
   848 enforced instead of silently assumed.  INCOMPATIBILITY, may use
   849 Logic.legacy_(un)varify as temporary workaround.
   850 
   851 * Pure: structure Name provides scalable operations for generating
   852 internal variable names, notably Name.variants etc.  This replaces
   853 some popular functions from term.ML:
   854 
   855   Term.variant		->  Name.variant
   856   Term.variantlist	->  Name.variant_list  (*canonical argument order*)
   857   Term.invent_names	->  Name.invent_list
   858 
   859 Note that low-level renaming rarely occurs in new code -- operations
   860 from structure Variable are used instead (see below).
   861 
   862 * Pure: structure Variable provides fundamental operations for proper
   863 treatment of fixed/schematic variables in a context.  For example,
   864 Variable.import introduces fixes for schematics of given facts and
   865 Variable.export reverses the effect (up to renaming) -- this replaces
   866 various freeze_thaw operations.
   867 
   868 * Pure: structure Goal provides simple interfaces for
   869 init/conclude/finish and tactical prove operations (replacing former
   870 Tactic.prove).  Goal.prove is the canonical way to prove results
   871 within a given context; Goal.prove_global is a degraded version for
   872 theory level goals, including a global Drule.standard.  Note that
   873 OldGoals.prove_goalw_cterm has long been obsolete, since it is
   874 ill-behaved in a local proof context (e.g. with local fixes/assumes or
   875 in a locale context).
   876 
   877 * Isar: simplified treatment of user-level errors, using exception
   878 ERROR of string uniformly.  Function error now merely raises ERROR,
   879 without any side effect on output channels.  The Isar toplevel takes
   880 care of proper display of ERROR exceptions.  ML code may use plain
   881 handle/can/try; cat_error may be used to concatenate errors like this:
   882 
   883   ... handle ERROR msg => cat_error msg "..."
   884 
   885 Toplevel ML code (run directly or through the Isar toplevel) may be
   886 embedded into the Isar toplevel with exception display/debug like
   887 this:
   888 
   889   Isar.toplevel (fn () => ...)
   890 
   891 INCOMPATIBILITY, removed special transform_error facilities, removed
   892 obsolete variants of user-level exceptions (ERROR_MESSAGE,
   893 Context.PROOF, ProofContext.CONTEXT, Proof.STATE, ProofHistory.FAIL)
   894 -- use plain ERROR instead.
   895 
   896 * Isar: theory setup now has type (theory -> theory), instead of a
   897 list.  INCOMPATIBILITY, may use #> to compose setup functions.
   898 
   899 * Isar: installed ML toplevel pretty printer for type Proof.context,
   900 subject to ProofContext.debug/verbose flags.
   901 
   902 * Isar: Toplevel.theory_to_proof admits transactions that modify the
   903 theory before entering a proof state.  Transactions now always see a
   904 quasi-functional intermediate checkpoint, both in interactive and
   905 batch mode.
   906 
   907 * Simplifier: the simpset of a running simplification process now
   908 contains a proof context (cf. Simplifier.the_context), which is the
   909 very context that the initial simpset has been retrieved from (by
   910 simpset_of/local_simpset_of).  Consequently, all plug-in components
   911 (solver, looper etc.) may depend on arbitrary proof data.
   912 
   913 * Simplifier.inherit_context inherits the proof context (plus the
   914 local bounds) of the current simplification process; any simproc
   915 etc. that calls the Simplifier recursively should do this!  Removed
   916 former Simplifier.inherit_bounds, which is already included here --
   917 INCOMPATIBILITY.  Tools based on low-level rewriting may even have to
   918 specify an explicit context using Simplifier.context/theory_context.
   919 
   920 * Simplifier/Classical Reasoner: more abstract interfaces
   921 change_simpset/claset for modifying the simpset/claset reference of a
   922 theory; raw versions simpset/claset_ref etc. have been discontinued --
   923 INCOMPATIBILITY.
   924 
   925 * Provers: more generic wrt. syntax of object-logics, avoid hardwired
   926 "Trueprop" etc.
   927 
   928 
   929 *** System ***
   930 
   931 * settings: ML_IDENTIFIER -- which is appended to user specific heap
   932 locations -- now includes the Isabelle version identifier as well.
   933 This simplifies use of multiple Isabelle installations.
   934 
   935 * isabelle-process: option -S (secure mode) disables some critical
   936 operations, notably runtime compilation and evaluation of ML source
   937 code.
   938 
   939 
   940 New in Isabelle2005 (October 2005)
   941 ----------------------------------
   942 
   943 *** General ***
   944 
   945 * Theory headers: the new header syntax for Isar theories is
   946 
   947   theory <name>
   948   imports <theory1> ... <theoryN>
   949   uses <file1> ... <fileM>
   950   begin
   951 
   952 where the 'uses' part is optional.  The previous syntax
   953 
   954   theory <name> = <theory1> + ... + <theoryN>:
   955 
   956 will disappear in the next release.  Use isatool fixheaders to convert
   957 existing theory files.  Note that there is no change in ancient
   958 non-Isar theories now, but these will disappear soon.
   959 
   960 * Theory loader: parent theories can now also be referred to via
   961 relative and absolute paths.
   962 
   963 * Command 'find_theorems' searches for a list of criteria instead of a
   964 list of constants. Known criteria are: intro, elim, dest, name:string,
   965 simp:term, and any term. Criteria can be preceded by '-' to select
   966 theorems that do not match. Intro, elim, dest select theorems that
   967 match the current goal, name:s selects theorems whose fully qualified
   968 name contain s, and simp:term selects all simplification rules whose
   969 lhs match term.  Any other term is interpreted as pattern and selects
   970 all theorems matching the pattern. Available in ProofGeneral under
   971 'ProofGeneral -> Find Theorems' or C-c C-f.  Example:
   972 
   973   C-c C-f (100) "(_::nat) + _ + _" intro -name: "HOL."
   974 
   975 prints the last 100 theorems matching the pattern "(_::nat) + _ + _",
   976 matching the current goal as introduction rule and not having "HOL."
   977 in their name (i.e. not being defined in theory HOL).
   978 
   979 * Command 'thms_containing' has been discontinued in favour of
   980 'find_theorems'; INCOMPATIBILITY.
   981 
   982 * Communication with Proof General is now 8bit clean, which means that
   983 Unicode text in UTF-8 encoding may be used within theory texts (both
   984 formal and informal parts).  Cf. option -U of the Isabelle Proof
   985 General interface.  Here are some simple examples (cf. src/HOL/ex):
   986 
   987   http://isabelle.in.tum.de/library/HOL/ex/Hebrew.html
   988   http://isabelle.in.tum.de/library/HOL/ex/Chinese.html
   989 
   990 * Improved efficiency of the Simplifier and, to a lesser degree, the
   991 Classical Reasoner.  Typical big applications run around 2 times
   992 faster.
   993 
   994 
   995 *** Document preparation ***
   996 
   997 * Commands 'display_drafts' and 'print_drafts' perform simple output
   998 of raw sources.  Only those symbols that do not require additional
   999 LaTeX packages (depending on comments in isabellesym.sty) are
  1000 displayed properly, everything else is left verbatim.  isatool display
  1001 and isatool print are used as front ends (these are subject to the
  1002 DVI/PDF_VIEWER and PRINT_COMMAND settings, respectively).
  1003 
  1004 * Command tags control specific markup of certain regions of text,
  1005 notably folding and hiding.  Predefined tags include "theory" (for
  1006 theory begin and end), "proof" for proof commands, and "ML" for
  1007 commands involving ML code; the additional tags "visible" and
  1008 "invisible" are unused by default.  Users may give explicit tag
  1009 specifications in the text, e.g. ''by %invisible (auto)''.  The
  1010 interpretation of tags is determined by the LaTeX job during document
  1011 preparation: see option -V of isatool usedir, or options -n and -t of
  1012 isatool document, or even the LaTeX macros \isakeeptag, \isafoldtag,
  1013 \isadroptag.
  1014 
  1015 Several document versions may be produced at the same time via isatool
  1016 usedir (the generated index.html will link all of them).  Typical
  1017 specifications include ''-V document=theory,proof,ML'' to present
  1018 theory/proof/ML parts faithfully, ''-V outline=/proof,/ML'' to fold
  1019 proof and ML commands, and ''-V mutilated=-theory,-proof,-ML'' to omit
  1020 these parts without any formal replacement text.  The Isabelle site
  1021 default settings produce ''document'' and ''outline'' versions as
  1022 specified above.
  1023 
  1024 * Several new antiquotations:
  1025 
  1026   @{term_type term} prints a term with its type annotated;
  1027 
  1028   @{typeof term} prints the type of a term;
  1029 
  1030   @{const const} is the same as @{term const}, but checks that the
  1031   argument is a known logical constant;
  1032 
  1033   @{term_style style term} and @{thm_style style thm} print a term or
  1034   theorem applying a "style" to it
  1035 
  1036   @{ML text}
  1037 
  1038 Predefined styles are 'lhs' and 'rhs' printing the lhs/rhs of
  1039 definitions, equations, inequations etc., 'concl' printing only the
  1040 conclusion of a meta-logical statement theorem, and 'prem1' .. 'prem19'
  1041 to print the specified premise.  TermStyle.add_style provides an ML
  1042 interface for introducing further styles.  See also the "LaTeX Sugar"
  1043 document practical applications.  The ML antiquotation prints
  1044 type-checked ML expressions verbatim.
  1045 
  1046 * Markup commands 'chapter', 'section', 'subsection', 'subsubsection',
  1047 and 'text' support optional locale specification '(in loc)', which
  1048 specifies the default context for interpreting antiquotations.  For
  1049 example: 'text (in lattice) {* @{thm inf_assoc}*}'.
  1050 
  1051 * Option 'locale=NAME' of antiquotations specifies an alternative
  1052 context interpreting the subsequent argument.  For example: @{thm
  1053 [locale=lattice] inf_assoc}.
  1054 
  1055 * Proper output of proof terms (@{prf ...} and @{full_prf ...}) within
  1056 a proof context.
  1057 
  1058 * Proper output of antiquotations for theory commands involving a
  1059 proof context (such as 'locale' or 'theorem (in loc) ...').
  1060 
  1061 * Delimiters of outer tokens (string etc.) now produce separate LaTeX
  1062 macros (\isachardoublequoteopen, isachardoublequoteclose etc.).
  1063 
  1064 * isatool usedir: new option -C (default true) controls whether option
  1065 -D should include a copy of the original document directory; -C false
  1066 prevents unwanted effects such as copying of administrative CVS data.
  1067 
  1068 
  1069 *** Pure ***
  1070 
  1071 * Considerably improved version of 'constdefs' command.  Now performs
  1072 automatic type-inference of declared constants; additional support for
  1073 local structure declarations (cf. locales and HOL records), see also
  1074 isar-ref manual.  Potential INCOMPATIBILITY: need to observe strictly
  1075 sequential dependencies of definitions within a single 'constdefs'
  1076 section; moreover, the declared name needs to be an identifier.  If
  1077 all fails, consider to fall back on 'consts' and 'defs' separately.
  1078 
  1079 * Improved indexed syntax and implicit structures.  First of all,
  1080 indexed syntax provides a notational device for subscripted
  1081 application, using the new syntax \<^bsub>term\<^esub> for arbitrary
  1082 expressions.  Secondly, in a local context with structure
  1083 declarations, number indexes \<^sub>n or the empty index (default
  1084 number 1) refer to a certain fixed variable implicitly; option
  1085 show_structs controls printing of implicit structures.  Typical
  1086 applications of these concepts involve record types and locales.
  1087 
  1088 * New command 'no_syntax' removes grammar declarations (and
  1089 translations) resulting from the given syntax specification, which is
  1090 interpreted in the same manner as for the 'syntax' command.
  1091 
  1092 * 'Advanced' translation functions (parse_translation etc.) may depend
  1093 on the signature of the theory context being presently used for
  1094 parsing/printing, see also isar-ref manual.
  1095 
  1096 * Improved 'oracle' command provides a type-safe interface to turn an
  1097 ML expression of type theory -> T -> term into a primitive rule of
  1098 type theory -> T -> thm (i.e. the functionality of Thm.invoke_oracle
  1099 is already included here); see also FOL/ex/IffExample.thy;
  1100 INCOMPATIBILITY.
  1101 
  1102 * axclass: name space prefix for class "c" is now "c_class" (was "c"
  1103 before); "cI" is no longer bound, use "c.intro" instead.
  1104 INCOMPATIBILITY.  This change avoids clashes of fact bindings for
  1105 axclasses vs. locales.
  1106 
  1107 * Improved internal renaming of symbolic identifiers -- attach primes
  1108 instead of base 26 numbers.
  1109 
  1110 * New flag show_question_marks controls printing of leading question
  1111 marks in schematic variable names.
  1112 
  1113 * In schematic variable names, *any* symbol following \<^isub> or
  1114 \<^isup> is now treated as part of the base name.  For example, the
  1115 following works without printing of awkward ".0" indexes:
  1116 
  1117   lemma "x\<^isub>1 = x\<^isub>2 ==> x\<^isub>2 = x\<^isub>1"
  1118     by simp
  1119 
  1120 * Inner syntax includes (*(*nested*) comments*).
  1121 
  1122 * Pretty printer now supports unbreakable blocks, specified in mixfix
  1123 annotations as "(00...)".
  1124 
  1125 * Clear separation of logical types and nonterminals, where the latter
  1126 may only occur in 'syntax' specifications or type abbreviations.
  1127 Before that distinction was only partially implemented via type class
  1128 "logic" vs. "{}".  Potential INCOMPATIBILITY in rare cases of improper
  1129 use of 'types'/'consts' instead of 'nonterminals'/'syntax'.  Some very
  1130 exotic syntax specifications may require further adaption
  1131 (e.g. Cube/Cube.thy).
  1132 
  1133 * Removed obsolete type class "logic", use the top sort {} instead.
  1134 Note that non-logical types should be declared as 'nonterminals'
  1135 rather than 'types'.  INCOMPATIBILITY for new object-logic
  1136 specifications.
  1137 
  1138 * Attributes 'induct' and 'cases': type or set names may now be
  1139 locally fixed variables as well.
  1140 
  1141 * Simplifier: can now control the depth to which conditional rewriting
  1142 is traced via the PG menu Isabelle -> Settings -> Trace Simp Depth
  1143 Limit.
  1144 
  1145 * Simplifier: simplification procedures may now take the current
  1146 simpset into account (cf. Simplifier.simproc(_i) / mk_simproc
  1147 interface), which is very useful for calling the Simplifier
  1148 recursively.  Minor INCOMPATIBILITY: the 'prems' argument of simprocs
  1149 is gone -- use prems_of_ss on the simpset instead.  Moreover, the
  1150 low-level mk_simproc no longer applies Logic.varify internally, to
  1151 allow for use in a context of fixed variables.
  1152 
  1153 * thin_tac now works even if the assumption being deleted contains !!
  1154 or ==>.  More generally, erule now works even if the major premise of
  1155 the elimination rule contains !! or ==>.
  1156 
  1157 * Method 'rules' has been renamed to 'iprover'. INCOMPATIBILITY.
  1158 
  1159 * Reorganized bootstrapping of the Pure theories; CPure is now derived
  1160 from Pure, which contains all common declarations already.  Both
  1161 theories are defined via plain Isabelle/Isar .thy files.
  1162 INCOMPATIBILITY: elements of CPure (such as the CPure.intro /
  1163 CPure.elim / CPure.dest attributes) now appear in the Pure name space;
  1164 use isatool fixcpure to adapt your theory and ML sources.
  1165 
  1166 * New syntax 'name(i-j, i-, i, ...)' for referring to specific
  1167 selections of theorems in named facts via index ranges.
  1168 
  1169 * 'print_theorems': in theory mode, really print the difference
  1170 wrt. the last state (works for interactive theory development only),
  1171 in proof mode print all local facts (cf. 'print_facts');
  1172 
  1173 * 'hide': option '(open)' hides only base names.
  1174 
  1175 * More efficient treatment of intermediate checkpoints in interactive
  1176 theory development.
  1177 
  1178 * Code generator is now invoked via code_module (incremental code
  1179 generation) and code_library (modular code generation, ML structures
  1180 for each theory).  INCOMPATIBILITY: new keywords 'file' and 'contains'
  1181 must be quoted when used as identifiers.
  1182 
  1183 * New 'value' command for reading, evaluating and printing terms using
  1184 the code generator.  INCOMPATIBILITY: command keyword 'value' must be
  1185 quoted when used as identifier.
  1186 
  1187 
  1188 *** Locales ***
  1189 
  1190 * New commands for the interpretation of locale expressions in
  1191 theories (1), locales (2) and proof contexts (3).  These generate
  1192 proof obligations from the expression specification.  After the
  1193 obligations have been discharged, theorems of the expression are added
  1194 to the theory, target locale or proof context.  The synopsis of the
  1195 commands is a follows:
  1196 
  1197   (1) interpretation expr inst
  1198   (2) interpretation target < expr
  1199   (3) interpret expr inst
  1200 
  1201 Interpretation in theories and proof contexts require a parameter
  1202 instantiation of terms from the current context.  This is applied to
  1203 specifications and theorems of the interpreted expression.
  1204 Interpretation in locales only permits parameter renaming through the
  1205 locale expression.  Interpretation is smart in that interpretations
  1206 that are active already do not occur in proof obligations, neither are
  1207 instantiated theorems stored in duplicate.  Use 'print_interps' to
  1208 inspect active interpretations of a particular locale.  For details,
  1209 see the Isar Reference manual.  Examples can be found in
  1210 HOL/Finite_Set.thy and HOL/Algebra/UnivPoly.thy.
  1211 
  1212 INCOMPATIBILITY: former 'instantiate' has been withdrawn, use
  1213 'interpret' instead.
  1214 
  1215 * New context element 'constrains' for adding type constraints to
  1216 parameters.
  1217 
  1218 * Context expressions: renaming of parameters with syntax
  1219 redeclaration.
  1220 
  1221 * Locale declaration: 'includes' disallowed.
  1222 
  1223 * Proper static binding of attribute syntax -- i.e. types / terms /
  1224 facts mentioned as arguments are always those of the locale definition
  1225 context, independently of the context of later invocations.  Moreover,
  1226 locale operations (renaming and type / term instantiation) are applied
  1227 to attribute arguments as expected.
  1228 
  1229 INCOMPATIBILITY of the ML interface: always pass Attrib.src instead of
  1230 actual attributes; rare situations may require Attrib.attribute to
  1231 embed those attributes into Attrib.src that lack concrete syntax.
  1232 Attribute implementations need to cooperate properly with the static
  1233 binding mechanism.  Basic parsers Args.XXX_typ/term/prop and
  1234 Attrib.XXX_thm etc. already do the right thing without further
  1235 intervention.  Only unusual applications -- such as "where" or "of"
  1236 (cf. src/Pure/Isar/attrib.ML), which process arguments depending both
  1237 on the context and the facts involved -- may have to assign parsed
  1238 values to argument tokens explicitly.
  1239 
  1240 * Changed parameter management in theorem generation for long goal
  1241 statements with 'includes'.  INCOMPATIBILITY: produces a different
  1242 theorem statement in rare situations.
  1243 
  1244 * Locale inspection command 'print_locale' omits notes elements.  Use
  1245 'print_locale!' to have them included in the output.
  1246 
  1247 
  1248 *** Provers ***
  1249 
  1250 * Provers/hypsubst.ML: improved version of the subst method, for
  1251 single-step rewriting: it now works in bound variable contexts. New is
  1252 'subst (asm)', for rewriting an assumption.  INCOMPATIBILITY: may
  1253 rewrite a different subterm than the original subst method, which is
  1254 still available as 'simplesubst'.
  1255 
  1256 * Provers/quasi.ML: new transitivity reasoners for transitivity only
  1257 and quasi orders.
  1258 
  1259 * Provers/trancl.ML: new transitivity reasoner for transitive and
  1260 reflexive-transitive closure of relations.
  1261 
  1262 * Provers/blast.ML: new reference depth_limit to make blast's depth
  1263 limit (previously hard-coded with a value of 20) user-definable.
  1264 
  1265 * Provers/simplifier.ML has been moved to Pure, where Simplifier.setup
  1266 is peformed already.  Object-logics merely need to finish their
  1267 initial simpset configuration as before.  INCOMPATIBILITY.
  1268 
  1269 
  1270 *** HOL ***
  1271 
  1272 * Symbolic syntax of Hilbert Choice Operator is now as follows:
  1273 
  1274   syntax (epsilon)
  1275     "_Eps" :: "[pttrn, bool] => 'a"    ("(3\<some>_./ _)" [0, 10] 10)
  1276 
  1277 The symbol \<some> is displayed as the alternative epsilon of LaTeX
  1278 and x-symbol; use option '-m epsilon' to get it actually printed.
  1279 Moreover, the mathematically important symbolic identifier \<epsilon>
  1280 becomes available as variable, constant etc.  INCOMPATIBILITY,
  1281 
  1282 * "x > y" abbreviates "y < x" and "x >= y" abbreviates "y <= x".
  1283 Similarly for all quantifiers: "ALL x > y" etc.  The x-symbol for >=
  1284 is \<ge>. New transitivity rules have been added to HOL/Orderings.thy to
  1285 support corresponding Isar calculations.
  1286 
  1287 * "{x:A. P}" abbreviates "{x. x:A & P}", and similarly for "\<in>"
  1288 instead of ":".
  1289 
  1290 * theory SetInterval: changed the syntax for open intervals:
  1291 
  1292   Old       New
  1293   {..n(}    {..<n}
  1294   {)n..}    {n<..}
  1295   {m..n(}   {m..<n}
  1296   {)m..n}   {m<..n}
  1297   {)m..n(}  {m<..<n}
  1298 
  1299 The old syntax is still supported but will disappear in the next
  1300 release.  For conversion use the following Emacs search and replace
  1301 patterns (these are not perfect but work quite well):
  1302 
  1303   {)\([^\.]*\)\.\.  ->  {\1<\.\.}
  1304   \.\.\([^(}]*\)(}  ->  \.\.<\1}
  1305 
  1306 * Theory Commutative_Ring (in Library): method comm_ring for proving
  1307 equalities in commutative rings; method 'algebra' provides a generic
  1308 interface.
  1309 
  1310 * Theory Finite_Set: changed the syntax for 'setsum', summation over
  1311 finite sets: "setsum (%x. e) A", which used to be "\<Sum>x:A. e", is
  1312 now either "SUM x:A. e" or "\<Sum>x \<in> A. e". The bound variable can
  1313 be a tuple pattern.
  1314 
  1315 Some new syntax forms are available:
  1316 
  1317   "\<Sum>x | P. e"      for     "setsum (%x. e) {x. P}"
  1318   "\<Sum>x = a..b. e"   for     "setsum (%x. e) {a..b}"
  1319   "\<Sum>x = a..<b. e"  for     "setsum (%x. e) {a..<b}"
  1320   "\<Sum>x < k. e"      for     "setsum (%x. e) {..<k}"
  1321 
  1322 The latter form "\<Sum>x < k. e" used to be based on a separate
  1323 function "Summation", which has been discontinued.
  1324 
  1325 * theory Finite_Set: in structured induction proofs, the insert case
  1326 is now 'case (insert x F)' instead of the old counterintuitive 'case
  1327 (insert F x)'.
  1328 
  1329 * The 'refute' command has been extended to support a much larger
  1330 fragment of HOL, including axiomatic type classes, constdefs and
  1331 typedefs, inductive datatypes and recursion.
  1332 
  1333 * New tactics 'sat' and 'satx' to prove propositional tautologies.
  1334 Requires zChaff with proof generation to be installed.  See
  1335 HOL/ex/SAT_Examples.thy for examples.
  1336 
  1337 * Datatype induction via method 'induct' now preserves the name of the
  1338 induction variable. For example, when proving P(xs::'a list) by
  1339 induction on xs, the induction step is now P(xs) ==> P(a#xs) rather
  1340 than P(list) ==> P(a#list) as previously.  Potential INCOMPATIBILITY
  1341 in unstructured proof scripts.
  1342 
  1343 * Reworked implementation of records.  Improved scalability for
  1344 records with many fields, avoiding performance problems for type
  1345 inference. Records are no longer composed of nested field types, but
  1346 of nested extension types. Therefore the record type only grows linear
  1347 in the number of extensions and not in the number of fields.  The
  1348 top-level (users) view on records is preserved.  Potential
  1349 INCOMPATIBILITY only in strange cases, where the theory depends on the
  1350 old record representation. The type generated for a record is called
  1351 <record_name>_ext_type.
  1352 
  1353 Flag record_quick_and_dirty_sensitive can be enabled to skip the
  1354 proofs triggered by a record definition or a simproc (if
  1355 quick_and_dirty is enabled).  Definitions of large records can take
  1356 quite long.
  1357 
  1358 New simproc record_upd_simproc for simplification of multiple record
  1359 updates enabled by default.  Moreover, trivial updates are also
  1360 removed: r(|x := x r|) = r.  INCOMPATIBILITY: old proofs break
  1361 occasionally, since simplification is more powerful by default.
  1362 
  1363 * typedef: proper support for polymorphic sets, which contain extra
  1364 type-variables in the term.
  1365 
  1366 * Simplifier: automatically reasons about transitivity chains
  1367 involving "trancl" (r^+) and "rtrancl" (r^*) by setting up tactics
  1368 provided by Provers/trancl.ML as additional solvers.  INCOMPATIBILITY:
  1369 old proofs break occasionally as simplification may now solve more
  1370 goals than previously.
  1371 
  1372 * Simplifier: converts x <= y into x = y if assumption y <= x is
  1373 present.  Works for all partial orders (class "order"), in particular
  1374 numbers and sets.  For linear orders (e.g. numbers) it treats ~ x < y
  1375 just like y <= x.
  1376 
  1377 * Simplifier: new simproc for "let x = a in f x".  If a is a free or
  1378 bound variable or a constant then the let is unfolded.  Otherwise
  1379 first a is simplified to b, and then f b is simplified to g. If
  1380 possible we abstract b from g arriving at "let x = b in h x",
  1381 otherwise we unfold the let and arrive at g.  The simproc can be
  1382 enabled/disabled by the reference use_let_simproc.  Potential
  1383 INCOMPATIBILITY since simplification is more powerful by default.
  1384 
  1385 * Classical reasoning: the meson method now accepts theorems as arguments.
  1386 
  1387 * Prover support: pre-release of the Isabelle-ATP linkup, which runs background
  1388 jobs to provide advice on the provability of subgoals.
  1389 
  1390 * Theory OrderedGroup and Ring_and_Field: various additions and
  1391 improvements to faciliate calculations involving equalities and
  1392 inequalities.
  1393 
  1394 The following theorems have been eliminated or modified
  1395 (INCOMPATIBILITY):
  1396 
  1397   abs_eq             now named abs_of_nonneg
  1398   abs_of_ge_0        now named abs_of_nonneg
  1399   abs_minus_eq       now named abs_of_nonpos
  1400   imp_abs_id         now named abs_of_nonneg
  1401   imp_abs_neg_id     now named abs_of_nonpos
  1402   mult_pos           now named mult_pos_pos
  1403   mult_pos_le        now named mult_nonneg_nonneg
  1404   mult_pos_neg_le    now named mult_nonneg_nonpos
  1405   mult_pos_neg2_le   now named mult_nonneg_nonpos2
  1406   mult_neg           now named mult_neg_neg
  1407   mult_neg_le        now named mult_nonpos_nonpos
  1408 
  1409 * Theory Parity: added rules for simplifying exponents.
  1410 
  1411 * Theory List:
  1412 
  1413 The following theorems have been eliminated or modified
  1414 (INCOMPATIBILITY):
  1415 
  1416   list_all_Nil       now named list_all.simps(1)
  1417   list_all_Cons      now named list_all.simps(2)
  1418   list_all_conv      now named list_all_iff
  1419   set_mem_eq         now named mem_iff
  1420 
  1421 * Theories SetsAndFunctions and BigO (see HOL/Library) support
  1422 asymptotic "big O" calculations.  See the notes in BigO.thy.
  1423 
  1424 
  1425 *** HOL-Complex ***
  1426 
  1427 * Theory RealDef: better support for embedding natural numbers and
  1428 integers in the reals.
  1429 
  1430 The following theorems have been eliminated or modified
  1431 (INCOMPATIBILITY):
  1432 
  1433   exp_ge_add_one_self  now requires no hypotheses
  1434   real_of_int_add      reversed direction of equality (use [symmetric])
  1435   real_of_int_minus    reversed direction of equality (use [symmetric])
  1436   real_of_int_diff     reversed direction of equality (use [symmetric])
  1437   real_of_int_mult     reversed direction of equality (use [symmetric])
  1438 
  1439 * Theory RComplete: expanded support for floor and ceiling functions.
  1440 
  1441 * Theory Ln is new, with properties of the natural logarithm
  1442 
  1443 * Hyperreal: There is a new type constructor "star" for making
  1444 nonstandard types.  The old type names are now type synonyms:
  1445 
  1446   hypreal = real star
  1447   hypnat = nat star
  1448   hcomplex = complex star
  1449 
  1450 * Hyperreal: Many groups of similarly-defined constants have been
  1451 replaced by polymorphic versions (INCOMPATIBILITY):
  1452 
  1453   star_of <-- hypreal_of_real, hypnat_of_nat, hcomplex_of_complex
  1454 
  1455   starset      <-- starsetNat, starsetC
  1456   *s*          <-- *sNat*, *sc*
  1457   starset_n    <-- starsetNat_n, starsetC_n
  1458   *sn*         <-- *sNatn*, *scn*
  1459   InternalSets <-- InternalNatSets, InternalCSets
  1460 
  1461   starfun      <-- starfun{Nat,Nat2,C,RC,CR}
  1462   *f*          <-- *fNat*, *fNat2*, *fc*, *fRc*, *fcR*
  1463   starfun_n    <-- starfun{Nat,Nat2,C,RC,CR}_n
  1464   *fn*         <-- *fNatn*, *fNat2n*, *fcn*, *fRcn*, *fcRn*
  1465   InternalFuns <-- InternalNatFuns, InternalNatFuns2, Internal{C,RC,CR}Funs
  1466 
  1467 * Hyperreal: Many type-specific theorems have been removed in favor of
  1468 theorems specific to various axiomatic type classes (INCOMPATIBILITY):
  1469 
  1470   add_commute <-- {hypreal,hypnat,hcomplex}_add_commute
  1471   add_assoc   <-- {hypreal,hypnat,hcomplex}_add_assocs
  1472   OrderedGroup.add_0 <-- {hypreal,hypnat,hcomplex}_add_zero_left
  1473   OrderedGroup.add_0_right <-- {hypreal,hcomplex}_add_zero_right
  1474   right_minus <-- hypreal_add_minus
  1475   left_minus <-- {hypreal,hcomplex}_add_minus_left
  1476   mult_commute <-- {hypreal,hypnat,hcomplex}_mult_commute
  1477   mult_assoc <-- {hypreal,hypnat,hcomplex}_mult_assoc
  1478   mult_1_left <-- {hypreal,hypnat}_mult_1, hcomplex_mult_one_left
  1479   mult_1_right <-- hcomplex_mult_one_right
  1480   mult_zero_left <-- hcomplex_mult_zero_left
  1481   left_distrib <-- {hypreal,hypnat,hcomplex}_add_mult_distrib
  1482   right_distrib <-- hypnat_add_mult_distrib2
  1483   zero_neq_one <-- {hypreal,hypnat,hcomplex}_zero_not_eq_one
  1484   right_inverse <-- hypreal_mult_inverse
  1485   left_inverse <-- hypreal_mult_inverse_left, hcomplex_mult_inv_left
  1486   order_refl <-- {hypreal,hypnat}_le_refl
  1487   order_trans <-- {hypreal,hypnat}_le_trans
  1488   order_antisym <-- {hypreal,hypnat}_le_anti_sym
  1489   order_less_le <-- {hypreal,hypnat}_less_le
  1490   linorder_linear <-- {hypreal,hypnat}_le_linear
  1491   add_left_mono <-- {hypreal,hypnat}_add_left_mono
  1492   mult_strict_left_mono <-- {hypreal,hypnat}_mult_less_mono2
  1493   add_nonneg_nonneg <-- hypreal_le_add_order
  1494 
  1495 * Hyperreal: Separate theorems having to do with type-specific
  1496 versions of constants have been merged into theorems that apply to the
  1497 new polymorphic constants (INCOMPATIBILITY):
  1498 
  1499   STAR_UNIV_set <-- {STAR_real,NatStar_real,STARC_complex}_set
  1500   STAR_empty_set <-- {STAR,NatStar,STARC}_empty_set
  1501   STAR_Un <-- {STAR,NatStar,STARC}_Un
  1502   STAR_Int <-- {STAR,NatStar,STARC}_Int
  1503   STAR_Compl <-- {STAR,NatStar,STARC}_Compl
  1504   STAR_subset <-- {STAR,NatStar,STARC}_subset
  1505   STAR_mem <-- {STAR,NatStar,STARC}_mem
  1506   STAR_mem_Compl <-- {STAR,STARC}_mem_Compl
  1507   STAR_diff <-- {STAR,STARC}_diff
  1508   STAR_star_of_image_subset <-- {STAR_hypreal_of_real, NatStar_hypreal_of_real,
  1509     STARC_hcomplex_of_complex}_image_subset
  1510   starset_n_Un <-- starset{Nat,C}_n_Un
  1511   starset_n_Int <-- starset{Nat,C}_n_Int
  1512   starset_n_Compl <-- starset{Nat,C}_n_Compl
  1513   starset_n_diff <-- starset{Nat,C}_n_diff
  1514   InternalSets_Un <-- Internal{Nat,C}Sets_Un
  1515   InternalSets_Int <-- Internal{Nat,C}Sets_Int
  1516   InternalSets_Compl <-- Internal{Nat,C}Sets_Compl
  1517   InternalSets_diff <-- Internal{Nat,C}Sets_diff
  1518   InternalSets_UNIV_diff <-- Internal{Nat,C}Sets_UNIV_diff
  1519   InternalSets_starset_n <-- Internal{Nat,C}Sets_starset{Nat,C}_n
  1520   starset_starset_n_eq <-- starset{Nat,C}_starset{Nat,C}_n_eq
  1521   starset_n_starset <-- starset{Nat,C}_n_starset{Nat,C}
  1522   starfun_n_starfun <-- starfun{Nat,Nat2,C,RC,CR}_n_starfun{Nat,Nat2,C,RC,CR}
  1523   starfun <-- starfun{Nat,Nat2,C,RC,CR}
  1524   starfun_mult <-- starfun{Nat,Nat2,C,RC,CR}_mult
  1525   starfun_add <-- starfun{Nat,Nat2,C,RC,CR}_add
  1526   starfun_minus <-- starfun{Nat,Nat2,C,RC,CR}_minus
  1527   starfun_diff <-- starfun{C,RC,CR}_diff
  1528   starfun_o <-- starfun{NatNat2,Nat2,_stafunNat,C,C_starfunRC,_starfunCR}_o
  1529   starfun_o2 <-- starfun{NatNat2,_stafunNat,C,C_starfunRC,_starfunCR}_o2
  1530   starfun_const_fun <-- starfun{Nat,Nat2,C,RC,CR}_const_fun
  1531   starfun_inverse <-- starfun{Nat,C,RC,CR}_inverse
  1532   starfun_eq <-- starfun{Nat,Nat2,C,RC,CR}_eq
  1533   starfun_eq_iff <-- starfun{C,RC,CR}_eq_iff
  1534   starfun_Id <-- starfunC_Id
  1535   starfun_approx <-- starfun{Nat,CR}_approx
  1536   starfun_capprox <-- starfun{C,RC}_capprox
  1537   starfun_abs <-- starfunNat_rabs
  1538   starfun_lambda_cancel <-- starfun{C,CR,RC}_lambda_cancel
  1539   starfun_lambda_cancel2 <-- starfun{C,CR,RC}_lambda_cancel2
  1540   starfun_mult_HFinite_approx <-- starfunCR_mult_HFinite_capprox
  1541   starfun_mult_CFinite_capprox <-- starfun{C,RC}_mult_CFinite_capprox
  1542   starfun_add_capprox <-- starfun{C,RC}_add_capprox
  1543   starfun_add_approx <-- starfunCR_add_approx
  1544   starfun_inverse_inverse <-- starfunC_inverse_inverse
  1545   starfun_divide <-- starfun{C,CR,RC}_divide
  1546   starfun_n <-- starfun{Nat,C}_n
  1547   starfun_n_mult <-- starfun{Nat,C}_n_mult
  1548   starfun_n_add <-- starfun{Nat,C}_n_add
  1549   starfun_n_add_minus <-- starfunNat_n_add_minus
  1550   starfun_n_const_fun <-- starfun{Nat,C}_n_const_fun
  1551   starfun_n_minus <-- starfun{Nat,C}_n_minus
  1552   starfun_n_eq <-- starfun{Nat,C}_n_eq
  1553 
  1554   star_n_add <-- {hypreal,hypnat,hcomplex}_add
  1555   star_n_minus <-- {hypreal,hcomplex}_minus
  1556   star_n_diff <-- {hypreal,hcomplex}_diff
  1557   star_n_mult <-- {hypreal,hcomplex}_mult
  1558   star_n_inverse <-- {hypreal,hcomplex}_inverse
  1559   star_n_le <-- {hypreal,hypnat}_le
  1560   star_n_less <-- {hypreal,hypnat}_less
  1561   star_n_zero_num <-- {hypreal,hypnat,hcomplex}_zero_num
  1562   star_n_one_num <-- {hypreal,hypnat,hcomplex}_one_num
  1563   star_n_abs <-- hypreal_hrabs
  1564   star_n_divide <-- hcomplex_divide
  1565 
  1566   star_of_add <-- {hypreal_of_real,hypnat_of_nat,hcomplex_of_complex}_add
  1567   star_of_minus <-- {hypreal_of_real,hcomplex_of_complex}_minus
  1568   star_of_diff <-- hypreal_of_real_diff
  1569   star_of_mult <-- {hypreal_of_real,hypnat_of_nat,hcomplex_of_complex}_mult
  1570   star_of_one <-- {hypreal_of_real,hcomplex_of_complex}_one
  1571   star_of_zero <-- {hypreal_of_real,hypnat_of_nat,hcomplex_of_complex}_zero
  1572   star_of_le <-- {hypreal_of_real,hypnat_of_nat}_le_iff
  1573   star_of_less <-- {hypreal_of_real,hypnat_of_nat}_less_iff
  1574   star_of_eq <-- {hypreal_of_real,hypnat_of_nat,hcomplex_of_complex}_eq_iff
  1575   star_of_inverse <-- {hypreal_of_real,hcomplex_of_complex}_inverse
  1576   star_of_divide <-- {hypreal_of_real,hcomplex_of_complex}_divide
  1577   star_of_of_nat <-- {hypreal_of_real,hcomplex_of_complex}_of_nat
  1578   star_of_of_int <-- {hypreal_of_real,hcomplex_of_complex}_of_int
  1579   star_of_number_of <-- {hypreal,hcomplex}_number_of
  1580   star_of_number_less <-- number_of_less_hypreal_of_real_iff
  1581   star_of_number_le <-- number_of_le_hypreal_of_real_iff
  1582   star_of_eq_number <-- hypreal_of_real_eq_number_of_iff
  1583   star_of_less_number <-- hypreal_of_real_less_number_of_iff
  1584   star_of_le_number <-- hypreal_of_real_le_number_of_iff
  1585   star_of_power <-- hypreal_of_real_power
  1586   star_of_eq_0 <-- hcomplex_of_complex_zero_iff
  1587 
  1588 * Hyperreal: new method "transfer" that implements the transfer
  1589 principle of nonstandard analysis. With a subgoal that mentions
  1590 nonstandard types like "'a star", the command "apply transfer"
  1591 replaces it with an equivalent one that mentions only standard types.
  1592 To be successful, all free variables must have standard types; non-
  1593 standard variables must have explicit universal quantifiers.
  1594 
  1595 * Hyperreal: A theory of Taylor series.
  1596 
  1597 
  1598 *** HOLCF ***
  1599 
  1600 * Discontinued special version of 'constdefs' (which used to support
  1601 continuous functions) in favor of the general Pure one with full
  1602 type-inference.
  1603 
  1604 * New simplification procedure for solving continuity conditions; it
  1605 is much faster on terms with many nested lambda abstractions (cubic
  1606 instead of exponential time).
  1607 
  1608 * New syntax for domain package: selector names are now optional.
  1609 Parentheses should be omitted unless argument is lazy, for example:
  1610 
  1611   domain 'a stream = cons "'a" (lazy "'a stream")
  1612 
  1613 * New command 'fixrec' for defining recursive functions with pattern
  1614 matching; defining multiple functions with mutual recursion is also
  1615 supported.  Patterns may include the constants cpair, spair, up, sinl,
  1616 sinr, or any data constructor defined by the domain package. The given
  1617 equations are proven as rewrite rules. See HOLCF/ex/Fixrec_ex.thy for
  1618 syntax and examples.
  1619 
  1620 * New commands 'cpodef' and 'pcpodef' for defining predicate subtypes
  1621 of cpo and pcpo types. Syntax is exactly like the 'typedef' command,
  1622 but the proof obligation additionally includes an admissibility
  1623 requirement. The packages generate instances of class cpo or pcpo,
  1624 with continuity and strictness theorems for Rep and Abs.
  1625 
  1626 * HOLCF: Many theorems have been renamed according to a more standard naming
  1627 scheme (INCOMPATIBILITY):
  1628 
  1629   foo_inject:  "foo$x = foo$y ==> x = y"
  1630   foo_eq:      "(foo$x = foo$y) = (x = y)"
  1631   foo_less:    "(foo$x << foo$y) = (x << y)"
  1632   foo_strict:  "foo$UU = UU"
  1633   foo_defined: "... ==> foo$x ~= UU"
  1634   foo_defined_iff: "(foo$x = UU) = (x = UU)"
  1635 
  1636 
  1637 *** ZF ***
  1638 
  1639 * ZF/ex: theories Group and Ring provide examples in abstract algebra,
  1640 including the First Isomorphism Theorem (on quotienting by the kernel
  1641 of a homomorphism).
  1642 
  1643 * ZF/Simplifier: install second copy of type solver that actually
  1644 makes use of TC rules declared to Isar proof contexts (or locales);
  1645 the old version is still required for ML proof scripts.
  1646 
  1647 
  1648 *** Cube ***
  1649 
  1650 * Converted to Isar theory format; use locales instead of axiomatic
  1651 theories.
  1652 
  1653 
  1654 *** ML ***
  1655 
  1656 * Pure/library.ML: added ##>, ##>>, #>> -- higher-order counterparts
  1657 for ||>, ||>>, |>>,
  1658 
  1659 * Pure/library.ML no longer defines its own option datatype, but uses
  1660 that of the SML basis, which has constructors NONE and SOME instead of
  1661 None and Some, as well as exception Option.Option instead of OPTION.
  1662 The functions the, if_none, is_some, is_none have been adapted
  1663 accordingly, while Option.map replaces apsome.
  1664 
  1665 * Pure/library.ML: the exception LIST has been given up in favour of
  1666 the standard exceptions Empty and Subscript, as well as
  1667 Library.UnequalLengths.  Function like Library.hd and Library.tl are
  1668 superceded by the standard hd and tl functions etc.
  1669 
  1670 A number of basic list functions are no longer exported to the ML
  1671 toplevel, as they are variants of predefined functions.  The following
  1672 suggests how one can translate existing code:
  1673 
  1674     rev_append xs ys = List.revAppend (xs, ys)
  1675     nth_elem (i, xs) = List.nth (xs, i)
  1676     last_elem xs = List.last xs
  1677     flat xss = List.concat xss
  1678     seq fs = List.app fs
  1679     partition P xs = List.partition P xs
  1680     mapfilter f xs = List.mapPartial f xs
  1681 
  1682 * Pure/library.ML: several combinators for linear functional
  1683 transformations, notably reverse application and composition:
  1684 
  1685   x |> f                f #> g
  1686   (x, y) |-> f          f #-> g
  1687 
  1688 * Pure/library.ML: introduced/changed precedence of infix operators:
  1689 
  1690   infix 1 |> |-> ||> ||>> |>> |>>> #> #->;
  1691   infix 2 ?;
  1692   infix 3 o oo ooo oooo;
  1693   infix 4 ~~ upto downto;
  1694 
  1695 Maybe INCOMPATIBILITY when any of those is used in conjunction with other
  1696 infix operators.
  1697 
  1698 * Pure/library.ML: natural list combinators fold, fold_rev, and
  1699 fold_map support linear functional transformations and nesting.  For
  1700 example:
  1701 
  1702   fold f [x1, ..., xN] y =
  1703     y |> f x1 |> ... |> f xN
  1704 
  1705   (fold o fold) f [xs1, ..., xsN] y =
  1706     y |> fold f xs1 |> ... |> fold f xsN
  1707 
  1708   fold f [x1, ..., xN] =
  1709     f x1 #> ... #> f xN
  1710 
  1711   (fold o fold) f [xs1, ..., xsN] =
  1712     fold f xs1 #> ... #> fold f xsN
  1713 
  1714 * Pure/library.ML: the following selectors on type 'a option are
  1715 available:
  1716 
  1717   the:               'a option -> 'a  (*partial*)
  1718   these:             'a option -> 'a  where 'a = 'b list
  1719   the_default: 'a -> 'a option -> 'a
  1720   the_list:          'a option -> 'a list
  1721 
  1722 * Pure/General: structure AList (cf. Pure/General/alist.ML) provides
  1723 basic operations for association lists, following natural argument
  1724 order; moreover the explicit equality predicate passed here avoids
  1725 potentially expensive polymorphic runtime equality checks.
  1726 The old functions may be expressed as follows:
  1727 
  1728   assoc = uncurry (AList.lookup (op =))
  1729   assocs = these oo AList.lookup (op =)
  1730   overwrite = uncurry (AList.update (op =)) o swap
  1731 
  1732 * Pure/General: structure AList (cf. Pure/General/alist.ML) provides
  1733 
  1734   val make: ('a -> 'b) -> 'a list -> ('a * 'b) list
  1735   val find: ('a * 'b -> bool) -> ('c * 'b) list -> 'a -> 'c list
  1736 
  1737 replacing make_keylist and keyfilter (occassionally used)
  1738 Naive rewrites:
  1739 
  1740   make_keylist = AList.make
  1741   keyfilter = AList.find (op =)
  1742 
  1743 * eq_fst and eq_snd now take explicit equality parameter, thus
  1744   avoiding eqtypes. Naive rewrites:
  1745 
  1746     eq_fst = eq_fst (op =)
  1747     eq_snd = eq_snd (op =)
  1748 
  1749 * Removed deprecated apl and apr (rarely used).
  1750   Naive rewrites:
  1751 
  1752     apl (n, op) =>>= curry op n
  1753     apr (op, m) =>>= fn n => op (n, m)
  1754 
  1755 * Pure/General: structure OrdList (cf. Pure/General/ord_list.ML)
  1756 provides a reasonably efficient light-weight implementation of sets as
  1757 lists.
  1758 
  1759 * Pure/General: generic tables (cf. Pure/General/table.ML) provide a
  1760 few new operations; existing lookup and update are now curried to
  1761 follow natural argument order (for use with fold etc.);
  1762 INCOMPATIBILITY, use (uncurry Symtab.lookup) etc. as last resort.
  1763 
  1764 * Pure/General: output via the Isabelle channels of
  1765 writeln/warning/error etc. is now passed through Output.output, with a
  1766 hook for arbitrary transformations depending on the print_mode
  1767 (cf. Output.add_mode -- the first active mode that provides a output
  1768 function wins).  Already formatted output may be embedded into further
  1769 text via Output.raw; the result of Pretty.string_of/str_of and derived
  1770 functions (string_of_term/cterm/thm etc.) is already marked raw to
  1771 accommodate easy composition of diagnostic messages etc.  Programmers
  1772 rarely need to care about Output.output or Output.raw at all, with
  1773 some notable exceptions: Output.output is required when bypassing the
  1774 standard channels (writeln etc.), or in token translations to produce
  1775 properly formatted results; Output.raw is required when capturing
  1776 already output material that will eventually be presented to the user
  1777 a second time.  For the default print mode, both Output.output and
  1778 Output.raw have no effect.
  1779 
  1780 * Pure/General: Output.time_accumulator NAME creates an operator ('a
  1781 -> 'b) -> 'a -> 'b to measure runtime and count invocations; the
  1782 cumulative results are displayed at the end of a batch session.
  1783 
  1784 * Pure/General: File.sysify_path and File.quote_sysify path have been
  1785 replaced by File.platform_path and File.shell_path (with appropriate
  1786 hooks).  This provides a clean interface for unusual systems where the
  1787 internal and external process view of file names are different.
  1788 
  1789 * Pure: more efficient orders for basic syntactic entities: added
  1790 fast_string_ord, fast_indexname_ord, fast_term_ord; changed sort_ord
  1791 and typ_ord to use fast_string_ord and fast_indexname_ord (term_ord is
  1792 NOT affected); structures Symtab, Vartab, Typtab, Termtab use the fast
  1793 orders now -- potential INCOMPATIBILITY for code that depends on a
  1794 particular order for Symtab.keys, Symtab.dest, etc. (consider using
  1795 Library.sort_strings on result).
  1796 
  1797 * Pure/term.ML: combinators fold_atyps, fold_aterms, fold_term_types,
  1798 fold_types traverse types/terms from left to right, observing natural
  1799 argument order.  Supercedes previous foldl_XXX versions, add_frees,
  1800 add_vars etc. have been adapted as well: INCOMPATIBILITY.
  1801 
  1802 * Pure: name spaces have been refined, with significant changes of the
  1803 internal interfaces -- INCOMPATIBILITY.  Renamed cond_extern(_table)
  1804 to extern(_table).  The plain name entry path is superceded by a
  1805 general 'naming' context, which also includes the 'policy' to produce
  1806 a fully qualified name and external accesses of a fully qualified
  1807 name; NameSpace.extend is superceded by context dependent
  1808 Sign.declare_name.  Several theory and proof context operations modify
  1809 the naming context.  Especially note Theory.restore_naming and
  1810 ProofContext.restore_naming to get back to a sane state; note that
  1811 Theory.add_path is no longer sufficient to recover from
  1812 Theory.absolute_path in particular.
  1813 
  1814 * Pure: new flags short_names (default false) and unique_names
  1815 (default true) for controlling output of qualified names.  If
  1816 short_names is set, names are printed unqualified.  If unique_names is
  1817 reset, the name prefix is reduced to the minimum required to achieve
  1818 the original result when interning again, even if there is an overlap
  1819 with earlier declarations.
  1820 
  1821 * Pure/TheoryDataFun: change of the argument structure; 'prep_ext' is
  1822 now 'extend', and 'merge' gets an additional Pretty.pp argument
  1823 (useful for printing error messages).  INCOMPATIBILITY.
  1824 
  1825 * Pure: major reorganization of the theory context.  Type Sign.sg and
  1826 Theory.theory are now identified, referring to the universal
  1827 Context.theory (see Pure/context.ML).  Actual signature and theory
  1828 content is managed as theory data.  The old code and interfaces were
  1829 spread over many files and structures; the new arrangement introduces
  1830 considerable INCOMPATIBILITY to gain more clarity:
  1831 
  1832   Context -- theory management operations (name, identity, inclusion,
  1833     parents, ancestors, merge, etc.), plus generic theory data;
  1834 
  1835   Sign -- logical signature and syntax operations (declaring consts,
  1836     types, etc.), plus certify/read for common entities;
  1837 
  1838   Theory -- logical theory operations (stating axioms, definitions,
  1839     oracles), plus a copy of logical signature operations (consts,
  1840     types, etc.); also a few basic management operations (Theory.copy,
  1841     Theory.merge, etc.)
  1842 
  1843 The most basic sign_of operations (Theory.sign_of, Thm.sign_of_thm
  1844 etc.) as well as the sign field in Thm.rep_thm etc. have been retained
  1845 for convenience -- they merely return the theory.
  1846 
  1847 * Pure: type Type.tsig is superceded by theory in most interfaces.
  1848 
  1849 * Pure: the Isar proof context type is already defined early in Pure
  1850 as Context.proof (note that ProofContext.context and Proof.context are
  1851 aliases, where the latter is the preferred name).  This enables other
  1852 Isabelle components to refer to that type even before Isar is present.
  1853 
  1854 * Pure/sign/theory: discontinued named name spaces (i.e. classK,
  1855 typeK, constK, axiomK, oracleK), but provide explicit operations for
  1856 any of these kinds.  For example, Sign.intern typeK is now
  1857 Sign.intern_type, Theory.hide_space Sign.typeK is now
  1858 Theory.hide_types.  Also note that former
  1859 Theory.hide_classes/types/consts are now
  1860 Theory.hide_classes_i/types_i/consts_i, while the non '_i' versions
  1861 internalize their arguments!  INCOMPATIBILITY.
  1862 
  1863 * Pure: get_thm interface (of PureThy and ProofContext) expects
  1864 datatype thmref (with constructors Name and NameSelection) instead of
  1865 plain string -- INCOMPATIBILITY;
  1866 
  1867 * Pure: cases produced by proof methods specify options, where NONE
  1868 means to remove case bindings -- INCOMPATIBILITY in
  1869 (RAW_)METHOD_CASES.
  1870 
  1871 * Pure: the following operations retrieve axioms or theorems from a
  1872 theory node or theory hierarchy, respectively:
  1873 
  1874   Theory.axioms_of: theory -> (string * term) list
  1875   Theory.all_axioms_of: theory -> (string * term) list
  1876   PureThy.thms_of: theory -> (string * thm) list
  1877   PureThy.all_thms_of: theory -> (string * thm) list
  1878 
  1879 * Pure: print_tac now outputs the goal through the trace channel.
  1880 
  1881 * Isar toplevel: improved diagnostics, mostly for Poly/ML only.
  1882 Reference Toplevel.debug (default false) controls detailed printing
  1883 and tracing of low-level exceptions; Toplevel.profiling (default 0)
  1884 controls execution profiling -- set to 1 for time and 2 for space
  1885 (both increase the runtime).
  1886 
  1887 * Isar session: The initial use of ROOT.ML is now always timed,
  1888 i.e. the log will show the actual process times, in contrast to the
  1889 elapsed wall-clock time that the outer shell wrapper produces.
  1890 
  1891 * Simplifier: improved handling of bound variables (nameless
  1892 representation, avoid allocating new strings).  Simprocs that invoke
  1893 the Simplifier recursively should use Simplifier.inherit_bounds to
  1894 avoid local name clashes.  Failure to do so produces warnings
  1895 "Simplifier: renamed bound variable ..."; set Simplifier.debug_bounds
  1896 for further details.
  1897 
  1898 * ML functions legacy_bindings and use_legacy_bindings produce ML fact
  1899 bindings for all theorems stored within a given theory; this may help
  1900 in porting non-Isar theories to Isar ones, while keeping ML proof
  1901 scripts for the time being.
  1902 
  1903 * ML operator HTML.with_charset specifies the charset begin used for
  1904 generated HTML files.  For example:
  1905 
  1906   HTML.with_charset "utf-8" use_thy "Hebrew";
  1907   HTML.with_charset "utf-8" use_thy "Chinese";
  1908 
  1909 
  1910 *** System ***
  1911 
  1912 * Allow symlinks to all proper Isabelle executables (Isabelle,
  1913 isabelle, isatool etc.).
  1914 
  1915 * ISABELLE_DOC_FORMAT setting specifies preferred document format (for
  1916 isatool doc, isatool mkdir, display_drafts etc.).
  1917 
  1918 * isatool usedir: option -f allows specification of the ML file to be
  1919 used by Isabelle; default is ROOT.ML.
  1920 
  1921 * New isatool version outputs the version identifier of the Isabelle
  1922 distribution being used.
  1923 
  1924 * HOL: new isatool dimacs2hol converts files in DIMACS CNF format
  1925 (containing Boolean satisfiability problems) into Isabelle/HOL
  1926 theories.
  1927 
  1928 
  1929 
  1930 New in Isabelle2004 (April 2004)
  1931 --------------------------------
  1932 
  1933 *** General ***
  1934 
  1935 * Provers/order.ML:  new efficient reasoner for partial and linear orders.
  1936   Replaces linorder.ML.
  1937 
  1938 * Pure: Greek letters (except small lambda, \<lambda>), as well as Gothic
  1939   (\<aa>...\<zz>\<AA>...\<ZZ>), calligraphic (\<A>...\<Z>), and Euler
  1940   (\<a>...\<z>), are now considered normal letters, and can therefore
  1941   be used anywhere where an ASCII letter (a...zA...Z) has until
  1942   now. COMPATIBILITY: This obviously changes the parsing of some
  1943   terms, especially where a symbol has been used as a binder, say
  1944   '\<Pi>x. ...', which is now a type error since \<Pi>x will be parsed
  1945   as an identifier.  Fix it by inserting a space around former
  1946   symbols.  Call 'isatool fixgreek' to try to fix parsing errors in
  1947   existing theory and ML files.
  1948 
  1949 * Pure: Macintosh and Windows line-breaks are now allowed in theory files.
  1950 
  1951 * Pure: single letter sub/superscripts (\<^isub> and \<^isup>) are now
  1952   allowed in identifiers. Similar to Greek letters \<^isub> is now considered
  1953   a normal (but invisible) letter. For multiple letter subscripts repeat
  1954   \<^isub> like this: x\<^isub>1\<^isub>2.
  1955 
  1956 * Pure: There are now sub-/superscripts that can span more than one
  1957   character. Text between \<^bsub> and \<^esub> is set in subscript in
  1958   ProofGeneral and LaTeX, text between \<^bsup> and \<^esup> in
  1959   superscript. The new control characters are not identifier parts.
  1960 
  1961 * Pure: Control-symbols of the form \<^raw:...> will literally print the
  1962   content of "..." to the latex file instead of \isacntrl... . The "..."
  1963   may consist of any printable characters excluding the end bracket >.
  1964 
  1965 * Pure: Using new Isar command "finalconsts" (or the ML functions
  1966   Theory.add_finals or Theory.add_finals_i) it is now possible to
  1967   declare constants "final", which prevents their being given a definition
  1968   later.  It is useful for constants whose behaviour is fixed axiomatically
  1969   rather than definitionally, such as the meta-logic connectives.
  1970 
  1971 * Pure: 'instance' now handles general arities with general sorts
  1972   (i.e. intersections of classes),
  1973 
  1974 * Presentation: generated HTML now uses a CSS style sheet to make layout
  1975   (somewhat) independent of content. It is copied from lib/html/isabelle.css.
  1976   It can be changed to alter the colors/layout of generated pages.
  1977 
  1978 
  1979 *** Isar ***
  1980 
  1981 * Tactic emulation methods rule_tac, erule_tac, drule_tac, frule_tac,
  1982   cut_tac, subgoal_tac and thin_tac:
  1983   - Now understand static (Isar) contexts.  As a consequence, users of Isar
  1984     locales are no longer forced to write Isar proof scripts.
  1985     For details see Isar Reference Manual, paragraph 4.3.2: Further tactic
  1986     emulations.
  1987   - INCOMPATIBILITY: names of variables to be instantiated may no
  1988     longer be enclosed in quotes.  Instead, precede variable name with `?'.
  1989     This is consistent with the instantiation attribute "where".
  1990 
  1991 * Attributes "where" and "of":
  1992   - Now take type variables of instantiated theorem into account when reading
  1993     the instantiation string.  This fixes a bug that caused instantiated
  1994     theorems to have too special types in some circumstances.
  1995   - "where" permits explicit instantiations of type variables.
  1996 
  1997 * Calculation commands "moreover" and "also" no longer interfere with
  1998   current facts ("this"), admitting arbitrary combinations with "then"
  1999   and derived forms.
  2000 
  2001 * Locales:
  2002   - Goal statements involving the context element "includes" no longer
  2003     generate theorems with internal delta predicates (those ending on
  2004     "_axioms") in the premise.
  2005     Resolve particular premise with <locale>.intro to obtain old form.
  2006   - Fixed bug in type inference ("unify_frozen") that prevented mix of target
  2007     specification and "includes" elements in goal statement.
  2008   - Rule sets <locale>.intro and <locale>.axioms no longer declared as
  2009     [intro?] and [elim?] (respectively) by default.
  2010   - Experimental command for instantiation of locales in proof contexts:
  2011         instantiate <label>[<attrs>]: <loc>
  2012     Instantiates locale <loc> and adds all its theorems to the current context
  2013     taking into account their attributes.  Label and attrs are optional
  2014     modifiers, like in theorem declarations.  If present, names of
  2015     instantiated theorems are qualified with <label>, and the attributes
  2016     <attrs> are applied after any attributes these theorems might have already.
  2017       If the locale has assumptions, a chained fact of the form
  2018     "<loc> t1 ... tn" is expected from which instantiations of the parameters
  2019     are derived.  The command does not support old-style locales declared
  2020     with "locale (open)".
  2021       A few (very simple) examples can be found in FOL/ex/LocaleInst.thy.
  2022 
  2023 * HOL: Tactic emulation methods induct_tac and case_tac understand static
  2024   (Isar) contexts.
  2025 
  2026 
  2027 *** HOL ***
  2028 
  2029 * Proof import: new image HOL4 contains the imported library from
  2030   the HOL4 system with about 2500 theorems. It is imported by
  2031   replaying proof terms produced by HOL4 in Isabelle. The HOL4 image
  2032   can be used like any other Isabelle image.  See
  2033   HOL/Import/HOL/README for more information.
  2034 
  2035 * Simplifier:
  2036   - Much improved handling of linear and partial orders.
  2037     Reasoners for linear and partial orders are set up for type classes
  2038     "linorder" and "order" respectively, and are added to the default simpset
  2039     as solvers.  This means that the simplifier can build transitivity chains
  2040     to solve goals from the assumptions.
  2041   - INCOMPATIBILITY: old proofs break occasionally.  Typically, applications
  2042     of blast or auto after simplification become unnecessary because the goal
  2043     is solved by simplification already.
  2044 
  2045 * Numerics: new theory Ring_and_Field contains over 250 basic numerical laws,
  2046     all proved in axiomatic type classes for semirings, rings and fields.
  2047 
  2048 * Numerics:
  2049   - Numeric types (nat, int, and in HOL-Complex rat, real, complex, etc.) are
  2050     now formalized using the Ring_and_Field theory mentioned above.
  2051   - INCOMPATIBILITY: simplification and arithmetic behaves somewhat differently
  2052     than before, because now they are set up once in a generic manner.
  2053   - INCOMPATIBILITY: many type-specific arithmetic laws have gone.
  2054     Look for the general versions in Ring_and_Field (and Power if they concern
  2055     exponentiation).
  2056 
  2057 * Type "rat" of the rational numbers is now available in HOL-Complex.
  2058 
  2059 * Records:
  2060   - Record types are now by default printed with their type abbreviation
  2061     instead of the list of all field types. This can be configured via
  2062     the reference "print_record_type_abbr".
  2063   - Simproc "record_upd_simproc" for simplification of multiple updates added
  2064     (not enabled by default).
  2065   - Simproc "record_ex_sel_eq_simproc" to simplify EX x. sel r = x resp.
  2066     EX x. x = sel r to True (not enabled by default).
  2067   - Tactic "record_split_simp_tac" to split and simplify records added.
  2068 
  2069 * 'specification' command added, allowing for definition by
  2070   specification.  There is also an 'ax_specification' command that
  2071   introduces the new constants axiomatically.
  2072 
  2073 * arith(_tac) is now able to generate counterexamples for reals as well.
  2074 
  2075 * HOL-Algebra: new locale "ring" for non-commutative rings.
  2076 
  2077 * HOL-ex: InductiveInvariant_examples illustrates advanced recursive function
  2078   definitions, thanks to Sava Krsti\'{c} and John Matthews.
  2079 
  2080 * HOL-Matrix: a first theory for matrices in HOL with an application of
  2081   matrix theory to linear programming.
  2082 
  2083 * Unions and Intersections:
  2084   The latex output syntax of UN and INT has been changed
  2085   from "\Union x \in A. B" to "\Union_{x \in A} B"
  2086   i.e. the index formulae has become a subscript.
  2087   Similarly for "\Union x. B", and for \Inter instead of \Union.
  2088 
  2089 * Unions and Intersections over Intervals:
  2090   There is new short syntax "UN i<=n. A" for "UN i:{0..n}. A". There is
  2091   also an x-symbol version with subscripts "\<Union>\<^bsub>i <= n\<^esub>. A"
  2092   like in normal math, and corresponding versions for < and for intersection.
  2093 
  2094 * HOL/List: Ordering "lexico" is renamed "lenlex" and the standard
  2095   lexicographic dictonary ordering has been added as "lexord".
  2096 
  2097 * ML: the legacy theory structures Int and List have been removed. They had
  2098   conflicted with ML Basis Library structures having the same names.
  2099 
  2100 * 'refute' command added to search for (finite) countermodels.  Only works
  2101   for a fragment of HOL.  The installation of an external SAT solver is
  2102   highly recommended.  See "HOL/Refute.thy" for details.
  2103 
  2104 * 'quickcheck' command: Allows to find counterexamples by evaluating
  2105   formulae under an assignment of free variables to random values.
  2106   In contrast to 'refute', it can deal with inductive datatypes,
  2107   but cannot handle quantifiers. See "HOL/ex/Quickcheck_Examples.thy"
  2108   for examples.
  2109 
  2110 
  2111 *** HOLCF ***
  2112 
  2113 * Streams now come with concatenation and are part of the HOLCF image
  2114 
  2115 
  2116 
  2117 New in Isabelle2003 (May 2003)
  2118 ------------------------------
  2119 
  2120 *** General ***
  2121 
  2122 * Provers/simplifier:
  2123 
  2124   - Completely reimplemented method simp (ML: Asm_full_simp_tac):
  2125     Assumptions are now subject to complete mutual simplification,
  2126     not just from left to right. The simplifier now preserves
  2127     the order of assumptions.
  2128 
  2129     Potential INCOMPATIBILITY:
  2130 
  2131     -- simp sometimes diverges where the old version did
  2132        not, e.g. invoking simp on the goal
  2133 
  2134         [| P (f x); y = x; f x = f y |] ==> Q
  2135 
  2136        now gives rise to the infinite reduction sequence
  2137 
  2138         P(f x) --(f x = f y)--> P(f y) --(y = x)--> P(f x) --(f x = f y)--> ...
  2139 
  2140        Using "simp (asm_lr)" (ML: Asm_lr_simp_tac) instead often solves this
  2141        kind of problem.
  2142 
  2143     -- Tactics combining classical reasoner and simplification (such as auto)
  2144        are also affected by this change, because many of them rely on
  2145        simp. They may sometimes diverge as well or yield a different numbers
  2146        of subgoals. Try to use e.g. force, fastsimp, or safe instead of auto
  2147        in case of problems. Sometimes subsequent calls to the classical
  2148        reasoner will fail because a preceeding call to the simplifier too
  2149        eagerly simplified the goal, e.g. deleted redundant premises.
  2150 
  2151   - The simplifier trace now shows the names of the applied rewrite rules
  2152 
  2153   - You can limit the number of recursive invocations of the simplifier
  2154     during conditional rewriting (where the simplifie tries to solve the
  2155     conditions before applying the rewrite rule):
  2156     ML "simp_depth_limit := n"
  2157     where n is an integer. Thus you can force termination where previously
  2158     the simplifier would diverge.
  2159 
  2160   - Accepts free variables as head terms in congruence rules.  Useful in Isar.
  2161 
  2162   - No longer aborts on failed congruence proof.  Instead, the
  2163     congruence is ignored.
  2164 
  2165 * Pure: New generic framework for extracting programs from constructive
  2166   proofs. See HOL/Extraction.thy for an example instantiation, as well
  2167   as HOL/Extraction for some case studies.
  2168 
  2169 * Pure: The main goal of the proof state is no longer shown by default, only
  2170 the subgoals. This behaviour is controlled by a new flag.
  2171    PG menu: Isabelle/Isar -> Settings -> Show Main Goal
  2172 (ML: Proof.show_main_goal).
  2173 
  2174 * Pure: You can find all matching introduction rules for subgoal 1, i.e. all
  2175 rules whose conclusion matches subgoal 1:
  2176       PG menu: Isabelle/Isar -> Show me -> matching rules
  2177 The rules are ordered by how closely they match the subgoal.
  2178 In particular, rules that solve a subgoal outright are displayed first
  2179 (or rather last, the way they are printed).
  2180 (ML: ProofGeneral.print_intros())
  2181 
  2182 * Pure: New flag trace_unify_fail causes unification to print
  2183 diagnostic information (PG: in trace buffer) when it fails. This is
  2184 useful for figuring out why single step proofs like rule, erule or
  2185 assumption failed.
  2186 
  2187 * Pure: Locale specifications now produce predicate definitions
  2188 according to the body of text (covering assumptions modulo local
  2189 definitions); predicate "loc_axioms" covers newly introduced text,
  2190 while "loc" is cumulative wrt. all included locale expressions; the
  2191 latter view is presented only on export into the global theory
  2192 context; potential INCOMPATIBILITY, use "(open)" option to fall back
  2193 on the old view without predicates;
  2194 
  2195 * Pure: predefined locales "var" and "struct" are useful for sharing
  2196 parameters (as in CASL, for example); just specify something like
  2197 ``var x + var y + struct M'' as import;
  2198 
  2199 * Pure: improved thms_containing: proper indexing of facts instead of
  2200 raw theorems; check validity of results wrt. current name space;
  2201 include local facts of proof configuration (also covers active
  2202 locales), cover fixed variables in index; may use "_" in term
  2203 specification; an optional limit for the number of printed facts may
  2204 be given (the default is 40);
  2205 
  2206 * Pure: disallow duplicate fact bindings within new-style theory files
  2207 (batch-mode only);
  2208 
  2209 * Provers: improved induct method: assumptions introduced by case
  2210 "foo" are split into "foo.hyps" (from the rule) and "foo.prems" (from
  2211 the goal statement); "foo" still refers to all facts collectively;
  2212 
  2213 * Provers: the function blast.overloaded has been removed: all constants
  2214 are regarded as potentially overloaded, which improves robustness in exchange
  2215 for slight decrease in efficiency;
  2216 
  2217 * Provers/linorder: New generic prover for transitivity reasoning over
  2218 linear orders.  Note: this prover is not efficient!
  2219 
  2220 * Isar: preview of problems to finish 'show' now produce an error
  2221 rather than just a warning (in interactive mode);
  2222 
  2223 
  2224 *** HOL ***
  2225 
  2226 * arith(_tac)
  2227 
  2228  - Produces a counter example if it cannot prove a goal.
  2229    Note that the counter example may be spurious if the goal is not a formula
  2230    of quantifier-free linear arithmetic.
  2231    In ProofGeneral the counter example appears in the trace buffer.
  2232 
  2233  - Knows about div k and mod k where k is a numeral of type nat or int.
  2234 
  2235  - Calls full Presburger arithmetic (by Amine Chaieb) if quantifier-free
  2236    linear arithmetic fails. This takes account of quantifiers and divisibility.
  2237    Presburger arithmetic can also be called explicitly via presburger(_tac).
  2238 
  2239 * simp's arithmetic capabilities have been enhanced a bit: it now
  2240 takes ~= in premises into account (by performing a case split);
  2241 
  2242 * simp reduces "m*(n div m) + n mod m" to n, even if the two summands
  2243 are distributed over a sum of terms;
  2244 
  2245 * New tactic "trans_tac" and method "trans" instantiate
  2246 Provers/linorder.ML for axclasses "order" and "linorder" (predicates
  2247 "<=", "<" and "=").
  2248 
  2249 * function INCOMPATIBILITIES: Pi-sets have been redefined and moved from main
  2250 HOL to Library/FuncSet; constant "Fun.op o" is now called "Fun.comp";
  2251 
  2252 * 'typedef' command has new option "open" to suppress the set
  2253 definition;
  2254 
  2255 * functions Min and Max on finite sets have been introduced (theory
  2256 Finite_Set);
  2257 
  2258 * attribute [symmetric] now works for relations as well; it turns
  2259 (x,y) : R^-1 into (y,x) : R, and vice versa;
  2260 
  2261 * induct over a !!-quantified statement (say !!x1..xn):
  2262   each "case" automatically performs "fix x1 .. xn" with exactly those names.
  2263 
  2264 * Map: `empty' is no longer a constant but a syntactic abbreviation for
  2265 %x. None. Warning: empty_def now refers to the previously hidden definition
  2266 of the empty set.
  2267 
  2268 * Algebra: formalization of classical algebra.  Intended as base for
  2269 any algebraic development in Isabelle.  Currently covers group theory
  2270 (up to Sylow's theorem) and ring theory (Universal Property of
  2271 Univariate Polynomials).  Contributions welcome;
  2272 
  2273 * GroupTheory: deleted, since its material has been moved to Algebra;
  2274 
  2275 * Complex: new directory of the complex numbers with numeric constants,
  2276 nonstandard complex numbers, and some complex analysis, standard and
  2277 nonstandard (Jacques Fleuriot);
  2278 
  2279 * HOL-Complex: new image for analysis, replacing HOL-Real and HOL-Hyperreal;
  2280 
  2281 * Hyperreal: introduced Gauge integration and hyperreal logarithms (Jacques
  2282 Fleuriot);
  2283 
  2284 * Real/HahnBanach: updated and adapted to locales;
  2285 
  2286 * NumberTheory: added Gauss's law of quadratic reciprocity (by Avigad,
  2287 Gray and Kramer);
  2288 
  2289 * UNITY: added the Meier-Sanders theory of progress sets;
  2290 
  2291 * MicroJava: bytecode verifier and lightweight bytecode verifier
  2292 as abstract algorithms, instantiated to the JVM;
  2293 
  2294 * Bali: Java source language formalization. Type system, operational
  2295 semantics, axiomatic semantics. Supported language features:
  2296 classes, interfaces, objects,virtual methods, static methods,
  2297 static/instance fields, arrays, access modifiers, definite
  2298 assignment, exceptions.
  2299 
  2300 
  2301 *** ZF ***
  2302 
  2303 * ZF/Constructible: consistency proof for AC (Gdel's constructible
  2304 universe, etc.);
  2305 
  2306 * Main ZF: virtually all theories converted to new-style format;
  2307 
  2308 
  2309 *** ML ***
  2310 
  2311 * Pure: Tactic.prove provides sane interface for internal proofs;
  2312 omits the infamous "standard" operation, so this is more appropriate
  2313 than prove_goalw_cterm in many situations (e.g. in simprocs);
  2314 
  2315 * Pure: improved error reporting of simprocs;
  2316 
  2317 * Provers: Simplifier.simproc(_i) provides sane interface for setting
  2318 up simprocs;
  2319 
  2320 
  2321 *** Document preparation ***
  2322 
  2323 * uses \par instead of \\ for line breaks in theory text. This may
  2324 shift some page breaks in large documents. To get the old behaviour
  2325 use \renewcommand{\isanewline}{\mbox{}\\\mbox{}} in root.tex.
  2326 
  2327 * minimized dependencies of isabelle.sty and isabellesym.sty on
  2328 other packages
  2329 
  2330 * \<euro> now needs package babel/greek instead of marvosym (which
  2331 broke \Rightarrow)
  2332 
  2333 * normal size for \<zero>...\<nine> (uses \mathbf instead of
  2334 textcomp package)
  2335 
  2336 
  2337 
  2338 New in Isabelle2002 (March 2002)
  2339 --------------------------------
  2340 
  2341 *** Document preparation ***
  2342 
  2343 * greatly simplified document preparation setup, including more
  2344 graceful interpretation of isatool usedir -i/-d/-D options, and more
  2345 instructive isatool mkdir; users should basically be able to get
  2346 started with "isatool mkdir HOL Test && isatool make"; alternatively,
  2347 users may run a separate document processing stage manually like this:
  2348 "isatool usedir -D output HOL Test && isatool document Test/output";
  2349 
  2350 * theory dependency graph may now be incorporated into documents;
  2351 isatool usedir -g true will produce session_graph.eps/.pdf for use
  2352 with \includegraphics of LaTeX;
  2353 
  2354 * proper spacing of consecutive markup elements, especially text
  2355 blocks after section headings;
  2356 
  2357 * support bold style (for single symbols only), input syntax is like
  2358 this: "\<^bold>\<alpha>" or "\<^bold>A";
  2359 
  2360 * \<bullet> is now output as bold \cdot by default, which looks much
  2361 better in printed text;
  2362 
  2363 * added default LaTeX bindings for \<tturnstile> and \<TTurnstile>;
  2364 note that these symbols are currently unavailable in Proof General /
  2365 X-Symbol; new symbols \<zero>, \<one>, ..., \<nine>, and \<euro>;
  2366 
  2367 * isatool latex no longer depends on changed TEXINPUTS, instead
  2368 isatool document copies the Isabelle style files to the target
  2369 location;
  2370 
  2371 
  2372 *** Isar ***
  2373 
  2374 * Pure/Provers: improved proof by cases and induction;
  2375   - 'case' command admits impromptu naming of parameters (such as
  2376     "case (Suc n)");
  2377   - 'induct' method divinates rule instantiation from the inductive
  2378     claim; no longer requires excessive ?P bindings for proper
  2379     instantiation of cases;
  2380   - 'induct' method properly enumerates all possibilities of set/type
  2381     rules; as a consequence facts may be also passed through *type*
  2382     rules without further ado;
  2383   - 'induct' method now derives symbolic cases from the *rulified*
  2384     rule (before it used to rulify cases stemming from the internal
  2385     atomized version); this means that the context of a non-atomic
  2386     statement becomes is included in the hypothesis, avoiding the
  2387     slightly cumbersome show "PROP ?case" form;
  2388   - 'induct' may now use elim-style induction rules without chaining
  2389     facts, using ``missing'' premises from the goal state; this allows
  2390     rules stemming from inductive sets to be applied in unstructured
  2391     scripts, while still benefitting from proper handling of non-atomic
  2392     statements; NB: major inductive premises need to be put first, all
  2393     the rest of the goal is passed through the induction;
  2394   - 'induct' proper support for mutual induction involving non-atomic
  2395     rule statements (uses the new concept of simultaneous goals, see
  2396     below);
  2397   - append all possible rule selections, but only use the first
  2398     success (no backtracking);
  2399   - removed obsolete "(simplified)" and "(stripped)" options of methods;
  2400   - undeclared rule case names default to numbers 1, 2, 3, ...;
  2401   - added 'print_induct_rules' (covered by help item in recent Proof
  2402     General versions);
  2403   - moved induct/cases attributes to Pure, methods to Provers;
  2404   - generic method setup instantiated for FOL and HOL;
  2405 
  2406 * Pure: support multiple simultaneous goal statements, for example
  2407 "have a: A and b: B" (same for 'theorem' etc.); being a pure
  2408 meta-level mechanism, this acts as if several individual goals had
  2409 been stated separately; in particular common proof methods need to be
  2410 repeated in order to cover all claims; note that a single elimination
  2411 step is *not* sufficient to establish the two conjunctions, so this
  2412 fails:
  2413 
  2414   assume "A & B" then have A and B ..   (*".." fails*)
  2415 
  2416 better use "obtain" in situations as above; alternative refer to
  2417 multi-step methods like 'auto', 'simp_all', 'blast+' etc.;
  2418 
  2419 * Pure: proper integration with ``locales''; unlike the original
  2420 version by Florian Kammller, Isar locales package high-level proof
  2421 contexts rather than raw logical ones (e.g. we admit to include
  2422 attributes everywhere); operations on locales include merge and
  2423 rename; support for implicit arguments (``structures''); simultaneous
  2424 type-inference over imports and text; see also HOL/ex/Locales.thy for
  2425 some examples;
  2426 
  2427 * Pure: the following commands have been ``localized'', supporting a
  2428 target locale specification "(in name)": 'lemma', 'theorem',
  2429 'corollary', 'lemmas', 'theorems', 'declare'; the results will be
  2430 stored both within the locale and at the theory level (exported and
  2431 qualified by the locale name);
  2432 
  2433 * Pure: theory goals may now be specified in ``long'' form, with
  2434 ad-hoc contexts consisting of arbitrary locale elements. for example
  2435 ``lemma foo: fixes x assumes "A x" shows "B x"'' (local syntax and
  2436 definitions may be given, too); the result is a meta-level rule with
  2437 the context elements being discharged in the obvious way;
  2438 
  2439 * Pure: new proof command 'using' allows to augment currently used
  2440 facts after a goal statement ('using' is syntactically analogous to
  2441 'apply', but acts on the goal's facts only); this allows chained facts
  2442 to be separated into parts given before and after a claim, as in
  2443 ``from a and b have C using d and e <proof>'';
  2444 
  2445 * Pure: renamed "antecedent" case to "rule_context";
  2446 
  2447 * Pure: new 'judgment' command records explicit information about the
  2448 object-logic embedding (used by several tools internally); no longer
  2449 use hard-wired "Trueprop";
  2450 
  2451 * Pure: added 'corollary' command;
  2452 
  2453 * Pure: fixed 'token_translation' command;
  2454 
  2455 * Pure: removed obsolete 'exported' attribute;
  2456 
  2457 * Pure: dummy pattern "_" in is/let is now automatically lifted over
  2458 bound variables: "ALL x. P x --> Q x" (is "ALL x. _ --> ?C x")
  2459 supersedes more cumbersome ... (is "ALL x. _ x --> ?C x");
  2460 
  2461 * Pure: method 'atomize' presents local goal premises as object-level
  2462 statements (atomic meta-level propositions); setup controlled via
  2463 rewrite rules declarations of 'atomize' attribute; example
  2464 application: 'induct' method with proper rule statements in improper
  2465 proof *scripts*;
  2466 
  2467 * Pure: emulation of instantiation tactics (rule_tac, cut_tac, etc.)
  2468 now consider the syntactic context of assumptions, giving a better
  2469 chance to get type-inference of the arguments right (this is
  2470 especially important for locales);
  2471 
  2472 * Pure: "sorry" no longer requires quick_and_dirty in interactive
  2473 mode;
  2474 
  2475 * Pure/obtain: the formal conclusion "thesis", being marked as
  2476 ``internal'', may no longer be reference directly in the text;
  2477 potential INCOMPATIBILITY, may need to use "?thesis" in rare
  2478 situations;
  2479 
  2480 * Pure: generic 'sym' attribute which declares a rule both as pure
  2481 'elim?' and for the 'symmetric' operation;
  2482 
  2483 * Pure: marginal comments ``--'' may now occur just anywhere in the
  2484 text; the fixed correlation with particular command syntax has been
  2485 discontinued;
  2486 
  2487 * Pure: new method 'rules' is particularly well-suited for proof
  2488 search in intuitionistic logic; a bit slower than 'blast' or 'fast',
  2489 but often produces more compact proof terms with less detours;
  2490 
  2491 * Pure/Provers/classical: simplified integration with pure rule
  2492 attributes and methods; the classical "intro?/elim?/dest?"
  2493 declarations coincide with the pure ones; the "rule" method no longer
  2494 includes classically swapped intros; "intro" and "elim" methods no
  2495 longer pick rules from the context; also got rid of ML declarations
  2496 AddXIs/AddXEs/AddXDs; all of this has some potential for
  2497 INCOMPATIBILITY;
  2498 
  2499 * Provers/classical: attribute 'swapped' produces classical inversions
  2500 of introduction rules;
  2501 
  2502 * Provers/simplifier: 'simplified' attribute may refer to explicit
  2503 rules instead of full simplifier context; 'iff' attribute handles
  2504 conditional rules;
  2505 
  2506 * HOL: 'typedef' now allows alternative names for Rep/Abs morphisms;
  2507 
  2508 * HOL: 'recdef' now fails on unfinished automated proofs, use
  2509 "(permissive)" option to recover old behavior;
  2510 
  2511 * HOL: 'inductive' no longer features separate (collective) attributes
  2512 for 'intros' (was found too confusing);
  2513 
  2514 * HOL: properly declared induction rules less_induct and
  2515 wf_induct_rule;
  2516 
  2517 
  2518 *** HOL ***
  2519 
  2520 * HOL: moved over to sane numeral syntax; the new policy is as
  2521 follows:
  2522 
  2523   - 0 and 1 are polymorphic constants, which are defined on any
  2524   numeric type (nat, int, real etc.);
  2525 
  2526   - 2, 3, 4, ... and -1, -2, -3, ... are polymorphic numerals, based
  2527   binary representation internally;
  2528 
  2529   - type nat has special constructor Suc, and generally prefers Suc 0
  2530   over 1::nat and Suc (Suc 0) over 2::nat;
  2531 
  2532 This change may cause significant problems of INCOMPATIBILITY; here
  2533 are some hints on converting existing sources:
  2534 
  2535   - due to the new "num" token, "-0" and "-1" etc. are now atomic
  2536   entities, so expressions involving "-" (unary or binary minus) need
  2537   to be spaced properly;
  2538 
  2539   - existing occurrences of "1" may need to be constraint "1::nat" or
  2540   even replaced by Suc 0; similar for old "2";
  2541 
  2542   - replace "#nnn" by "nnn", and "#-nnn" by "-nnn";
  2543 
  2544   - remove all special provisions on numerals in proofs;
  2545 
  2546 * HOL: simp rules nat_number expand numerals on nat to Suc/0
  2547 representation (depends on bin_arith_simps in the default context);
  2548 
  2549 * HOL: symbolic syntax for x^2 (numeral 2);
  2550 
  2551 * HOL: the class of all HOL types is now called "type" rather than
  2552 "term"; INCOMPATIBILITY, need to adapt references to this type class
  2553 in axclass/classes, instance/arities, and (usually rare) occurrences
  2554 in typings (of consts etc.); internally the class is called
  2555 "HOL.type", ML programs should refer to HOLogic.typeS;
  2556 
  2557 * HOL/record package improvements:
  2558   - new derived operations "fields" to build a partial record section,
  2559     "extend" to promote a fixed record to a record scheme, and
  2560     "truncate" for the reverse; cf. theorems "xxx.defs", which are *not*
  2561     declared as simp by default;
  2562   - shared operations ("more", "fields", etc.) now need to be always
  2563     qualified) --- potential INCOMPATIBILITY;
  2564   - removed "make_scheme" operations (use "make" with "extend") --
  2565     INCOMPATIBILITY;
  2566   - removed "more" class (simply use "term") -- INCOMPATIBILITY;
  2567   - provides cases/induct rules for use with corresponding Isar
  2568     methods (for concrete records, record schemes, concrete more
  2569     parts, and schematic more parts -- in that order);
  2570   - internal definitions directly based on a light-weight abstract
  2571     theory of product types over typedef rather than datatype;
  2572 
  2573 * HOL: generic code generator for generating executable ML code from
  2574 specifications; specific support for HOL constructs such as inductive
  2575 datatypes and sets, as well as recursive functions; can be invoked
  2576 via 'generate_code' theory section;
  2577 
  2578 * HOL: canonical cases/induct rules for n-tuples (n = 3..7);
  2579 
  2580 * HOL: consolidated and renamed several theories.  In particular:
  2581         Ord.thy has been absorbed into HOL.thy
  2582         String.thy has been absorbed into List.thy
  2583 
  2584 * HOL: concrete setsum syntax "\<Sum>i:A. b" == "setsum (%i. b) A"
  2585 (beware of argument permutation!);
  2586 
  2587 * HOL: linorder_less_split superseded by linorder_cases;
  2588 
  2589 * HOL/List: "nodups" renamed to "distinct";
  2590 
  2591 * HOL: added "The" definite description operator; move Hilbert's "Eps"
  2592 to peripheral theory "Hilbert_Choice"; some INCOMPATIBILITIES:
  2593   - Ex_def has changed, now need to use some_eq_ex
  2594 
  2595 * HOL: made split_all_tac safe; EXISTING PROOFS MAY FAIL OR LOOP, so
  2596 in this (rare) case use:
  2597 
  2598   delSWrapper "split_all_tac"
  2599   addSbefore ("unsafe_split_all_tac", unsafe_split_all_tac)
  2600 
  2601 * HOL: added safe wrapper "split_conv_tac" to claset; EXISTING PROOFS
  2602 MAY FAIL;
  2603 
  2604 * HOL: introduced f^n = f o ... o f; warning: due to the limits of
  2605 Isabelle's type classes, ^ on functions and relations has too general
  2606 a domain, namely ('a * 'b) set and 'a => 'b; this means that it may be
  2607 necessary to attach explicit type constraints;
  2608 
  2609 * HOL/Relation: the prefix name of the infix "O" has been changed from
  2610 "comp" to "rel_comp"; INCOMPATIBILITY: a few theorems have been
  2611 renamed accordingly (eg "compI" -> "rel_compI").
  2612 
  2613 * HOL: syntax translations now work properly with numerals and records
  2614 expressions;
  2615 
  2616 * HOL: bounded abstraction now uses syntax "%" / "\<lambda>" instead
  2617 of "lam" -- INCOMPATIBILITY;
  2618 
  2619 * HOL: got rid of some global declarations (potential INCOMPATIBILITY
  2620 for ML tools): const "()" renamed "Product_Type.Unity", type "unit"
  2621 renamed "Product_Type.unit";
  2622 
  2623 * HOL: renamed rtrancl_into_rtrancl2 to converse_rtrancl_into_rtrancl
  2624 
  2625 * HOL: removed obsolete theorem "optionE" (use "option.exhaust", or
  2626 the "cases" method);
  2627 
  2628 * HOL/GroupTheory: group theory examples including Sylow's theorem (by
  2629 Florian Kammller);
  2630 
  2631 * HOL/IMP: updated and converted to new-style theory format; several
  2632 parts turned into readable document, with proper Isar proof texts and
  2633 some explanations (by Gerwin Klein);
  2634 
  2635 * HOL-Real: added Complex_Numbers (by Gertrud Bauer);
  2636 
  2637 * HOL-Hyperreal is now a logic image;
  2638 
  2639 
  2640 *** HOLCF ***
  2641 
  2642 * Isar: consts/constdefs supports mixfix syntax for continuous
  2643 operations;
  2644 
  2645 * Isar: domain package adapted to new-style theory format, e.g. see
  2646 HOLCF/ex/Dnat.thy;
  2647 
  2648 * theory Lift: proper use of rep_datatype lift instead of ML hacks --
  2649 potential INCOMPATIBILITY; now use plain induct_tac instead of former
  2650 lift.induct_tac, always use UU instead of Undef;
  2651 
  2652 * HOLCF/IMP: updated and converted to new-style theory;
  2653 
  2654 
  2655 *** ZF ***
  2656 
  2657 * Isar: proper integration of logic-specific tools and packages,
  2658 including theory commands '(co)inductive', '(co)datatype',
  2659 'rep_datatype', 'inductive_cases', as well as methods 'ind_cases',
  2660 'induct_tac', 'case_tac', and 'typecheck' (with attribute 'TC');
  2661 
  2662 * theory Main no longer includes AC; for the Axiom of Choice, base
  2663 your theory on Main_ZFC;
  2664 
  2665 * the integer library now covers quotients and remainders, with many
  2666 laws relating division to addition, multiplication, etc.;
  2667 
  2668 * ZF/UNITY: Chandy and Misra's UNITY is now available in ZF, giving a
  2669 typeless version of the formalism;
  2670 
  2671 * ZF/AC, Coind, IMP, Resid: updated and converted to new-style theory
  2672 format;
  2673 
  2674 * ZF/Induct: new directory for examples of inductive definitions,
  2675 including theory Multiset for multiset orderings; converted to
  2676 new-style theory format;
  2677 
  2678 * ZF: many new theorems about lists, ordinals, etc.;
  2679 
  2680 
  2681 *** General ***
  2682 
  2683 * Pure/kernel: meta-level proof terms (by Stefan Berghofer); reference
  2684 variable proof controls level of detail: 0 = no proofs (only oracle
  2685 dependencies), 1 = lemma dependencies, 2 = compact proof terms; see
  2686 also ref manual for further ML interfaces;
  2687 
  2688 * Pure/axclass: removed obsolete ML interface
  2689 goal_subclass/goal_arity;
  2690 
  2691 * Pure/syntax: new token syntax "num" for plain numerals (without "#"
  2692 of "xnum"); potential INCOMPATIBILITY, since -0, -1 etc. are now
  2693 separate tokens, so expressions involving minus need to be spaced
  2694 properly;
  2695 
  2696 * Pure/syntax: support non-oriented infixes, using keyword "infix"
  2697 rather than "infixl" or "infixr";
  2698 
  2699 * Pure/syntax: concrete syntax for dummy type variables admits genuine
  2700 sort constraint specifications in type inference; e.g. "x::_::foo"
  2701 ensures that the type of "x" is of sort "foo" (but not necessarily a
  2702 type variable);
  2703 
  2704 * Pure/syntax: print modes "type_brackets" and "no_type_brackets"
  2705 control output of nested => (types); the default behavior is
  2706 "type_brackets";
  2707 
  2708 * Pure/syntax: builtin parse translation for "_constify" turns valued
  2709 tokens into AST constants;
  2710 
  2711 * Pure/syntax: prefer later declarations of translations and print
  2712 translation functions; potential INCOMPATIBILITY: need to reverse
  2713 multiple declarations for same syntax element constant;
  2714 
  2715 * Pure/show_hyps reset by default (in accordance to existing Isar
  2716 practice);
  2717 
  2718 * Provers/classical: renamed addaltern to addafter, addSaltern to
  2719 addSafter;
  2720 
  2721 * Provers/clasimp: ``iff'' declarations now handle conditional rules
  2722 as well;
  2723 
  2724 * system: tested support for MacOS X; should be able to get Isabelle +
  2725 Proof General to work in a plain Terminal after installing Poly/ML
  2726 (e.g. from the Isabelle distribution area) and GNU bash alone
  2727 (e.g. from http://www.apple.com); full X11, XEmacs and X-Symbol
  2728 support requires further installations, e.g. from
  2729 http://fink.sourceforge.net/);
  2730 
  2731 * system: support Poly/ML 4.1.1 (able to manage larger heaps);
  2732 
  2733 * system: reduced base memory usage by Poly/ML (approx. 20 MB instead
  2734 of 40 MB), cf. ML_OPTIONS;
  2735 
  2736 * system: Proof General keywords specification is now part of the
  2737 Isabelle distribution (see etc/isar-keywords.el);
  2738 
  2739 * system: support for persistent Proof General sessions (refrain from
  2740 outdating all loaded theories on startup); user may create writable
  2741 logic images like this: ``isabelle -q HOL Test'';
  2742 
  2743 * system: smart selection of Isabelle process versus Isabelle
  2744 interface, accommodates case-insensitive file systems (e.g. HFS+); may
  2745 run both "isabelle" and "Isabelle" even if file names are badly
  2746 damaged (executable inspects the case of the first letter of its own
  2747 name); added separate "isabelle-process" and "isabelle-interface";
  2748 
  2749 * system: refrain from any attempt at filtering input streams; no
  2750 longer support ``8bit'' encoding of old isabelle font, instead proper
  2751 iso-latin characters may now be used; the related isatools
  2752 "symbolinput" and "nonascii" have disappeared as well;
  2753 
  2754 * system: removed old "xterm" interface (the print modes "xterm" and
  2755 "xterm_color" are still available for direct use in a suitable
  2756 terminal);
  2757 
  2758 
  2759 
  2760 New in Isabelle99-2 (February 2001)
  2761 -----------------------------------
  2762 
  2763 *** Overview of INCOMPATIBILITIES ***
  2764 
  2765 * HOL: please note that theories in the Library and elsewhere often use the
  2766 new-style (Isar) format; to refer to their theorems in an ML script you must
  2767 bind them to ML identifers by e.g.      val thm_name = thm "thm_name";
  2768 
  2769 * HOL: inductive package no longer splits induction rule aggressively,
  2770 but only as far as specified by the introductions given; the old
  2771 format may be recovered via ML function complete_split_rule or attribute
  2772 'split_rule (complete)';
  2773 
  2774 * HOL: induct renamed to lfp_induct, lfp_Tarski to lfp_unfold,
  2775 gfp_Tarski to gfp_unfold;
  2776 
  2777 * HOL: contrapos, contrapos2 renamed to contrapos_nn, contrapos_pp;
  2778 
  2779 * HOL: infix "dvd" now has priority 50 rather than 70 (because it is a
  2780 relation); infix "^^" has been renamed "``"; infix "``" has been
  2781 renamed "`"; "univalent" has been renamed "single_valued";
  2782 
  2783 * HOL/Real: "rinv" and "hrinv" replaced by overloaded "inverse"
  2784 operation;
  2785 
  2786 * HOLCF: infix "`" has been renamed "$"; the symbol syntax is \<cdot>;
  2787 
  2788 * Isar: 'obtain' no longer declares "that" fact as simp/intro;
  2789 
  2790 * Isar/HOL: method 'induct' now handles non-atomic goals; as a
  2791 consequence, it is no longer monotonic wrt. the local goal context
  2792 (which is now passed through the inductive cases);
  2793 
  2794 * Document preparation: renamed standard symbols \<ll> to \<lless> and
  2795 \<gg> to \<ggreater>;
  2796 
  2797 
  2798 *** Document preparation ***
  2799 
  2800 * \isabellestyle{NAME} selects version of Isabelle output (currently
  2801 available: are "it" for near math-mode best-style output, "sl" for
  2802 slanted text style, and "tt" for plain type-writer; if no
  2803 \isabellestyle command is given, output is according to slanted
  2804 type-writer);
  2805 
  2806 * support sub/super scripts (for single symbols only), input syntax is
  2807 like this: "A\<^sup>*" or "A\<^sup>\<star>";
  2808 
  2809 * some more standard symbols; see Appendix A of the system manual for
  2810 the complete list of symbols defined in isabellesym.sty;
  2811 
  2812 * improved isabelle style files; more abstract symbol implementation
  2813 (should now use \isamath{...} and \isatext{...} in custom symbol
  2814 definitions);
  2815 
  2816 * antiquotation @{goals} and @{subgoals} for output of *dynamic* goals
  2817 state; Note that presentation of goal states does not conform to
  2818 actual human-readable proof documents.  Please do not include goal
  2819 states into document output unless you really know what you are doing!
  2820 
  2821 * proper indentation of antiquoted output with proportional LaTeX
  2822 fonts;
  2823 
  2824 * no_document ML operator temporarily disables LaTeX document
  2825 generation;
  2826 
  2827 * isatool unsymbolize tunes sources for plain ASCII communication;
  2828 
  2829 
  2830 *** Isar ***
  2831 
  2832 * Pure: Isar now suffers initial goal statements to contain unbound
  2833 schematic variables (this does not conform to actual readable proof
  2834 documents, due to unpredictable outcome and non-compositional proof
  2835 checking); users who know what they are doing may use schematic goals
  2836 for Prolog-style synthesis of proven results;
  2837 
  2838 * Pure: assumption method (an implicit finishing) now handles actual
  2839 rules as well;
  2840 
  2841 * Pure: improved 'obtain' --- moved to Pure, insert "that" into
  2842 initial goal, declare "that" only as Pure intro (only for single
  2843 steps); the "that" rule assumption may now be involved in implicit
  2844 finishing, thus ".." becomes a feasible for trivial obtains;
  2845 
  2846 * Pure: default proof step now includes 'intro_classes'; thus trivial
  2847 instance proofs may be performed by "..";
  2848 
  2849 * Pure: ?thesis / ?this / "..." now work for pure meta-level
  2850 statements as well;
  2851 
  2852 * Pure: more robust selection of calculational rules;
  2853 
  2854 * Pure: the builtin notion of 'finished' goal now includes the ==-refl
  2855 rule (as well as the assumption rule);
  2856 
  2857 * Pure: 'thm_deps' command visualizes dependencies of theorems and
  2858 lemmas, using the graph browser tool;
  2859 
  2860 * Pure: predict failure of "show" in interactive mode;
  2861 
  2862 * Pure: 'thms_containing' now takes actual terms as arguments;
  2863 
  2864 * HOL: improved method 'induct' --- now handles non-atomic goals
  2865 (potential INCOMPATIBILITY); tuned error handling;
  2866 
  2867 * HOL: cases and induct rules now provide explicit hints about the
  2868 number of facts to be consumed (0 for "type" and 1 for "set" rules);
  2869 any remaining facts are inserted into the goal verbatim;
  2870 
  2871 * HOL: local contexts (aka cases) may now contain term bindings as
  2872 well; the 'cases' and 'induct' methods new provide a ?case binding for
  2873 the result to be shown in each case;
  2874 
  2875 * HOL: added 'recdef_tc' command;
  2876 
  2877 * isatool convert assists in eliminating legacy ML scripts;
  2878 
  2879 
  2880 *** HOL ***
  2881 
  2882 * HOL/Library: a collection of generic theories to be used together
  2883 with main HOL; the theory loader path already includes this directory
  2884 by default; the following existing theories have been moved here:
  2885 HOL/Induct/Multiset, HOL/Induct/Acc (as Accessible_Part), HOL/While
  2886 (as While_Combinator), HOL/Lex/Prefix (as List_Prefix);
  2887 
  2888 * HOL/Unix: "Some aspects of Unix file-system security", a typical
  2889 modelling and verification task performed in Isabelle/HOL +
  2890 Isabelle/Isar + Isabelle document preparation (by Markus Wenzel).
  2891 
  2892 * HOL/Algebra: special summation operator SUM no longer exists, it has
  2893 been replaced by setsum; infix 'assoc' now has priority 50 (like
  2894 'dvd'); axiom 'one_not_zero' has been moved from axclass 'ring' to
  2895 'domain', this makes the theory consistent with mathematical
  2896 literature;
  2897 
  2898 * HOL basics: added overloaded operations "inverse" and "divide"
  2899 (infix "/"), syntax for generic "abs" operation, generic summation
  2900 operator \<Sum>;
  2901 
  2902 * HOL/typedef: simplified package, provide more useful rules (see also
  2903 HOL/subset.thy);
  2904 
  2905 * HOL/datatype: induction rule for arbitrarily branching datatypes is
  2906 now expressed as a proper nested rule (old-style tactic scripts may
  2907 require atomize_strip_tac to cope with non-atomic premises);
  2908 
  2909 * HOL: renamed theory "Prod" to "Product_Type", renamed "split" rule
  2910 to "split_conv" (old name still available for compatibility);
  2911 
  2912 * HOL: improved concrete syntax for strings (e.g. allows translation
  2913 rules with string literals);
  2914 
  2915 * HOL-Real-Hyperreal: this extends HOL-Real with the hyperreals
  2916  and Fleuriot's mechanization of analysis, including the transcendental
  2917  functions for the reals;
  2918 
  2919 * HOL/Real, HOL/Hyperreal: improved arithmetic simplification;
  2920 
  2921 
  2922 *** CTT ***
  2923 
  2924 * CTT: x-symbol support for Pi, Sigma, -->, : (membership); note that
  2925 "lam" is displayed as TWO lambda-symbols
  2926 
  2927 * CTT: theory Main now available, containing everything (that is, Bool
  2928 and Arith);
  2929 
  2930 
  2931 *** General ***
  2932 
  2933 * Pure: the Simplifier has been implemented properly as a derived rule
  2934 outside of the actual kernel (at last!); the overall performance
  2935 penalty in practical applications is about 50%, while reliability of
  2936 the Isabelle inference kernel has been greatly improved;
  2937 
  2938 * print modes "brackets" and "no_brackets" control output of nested =>
  2939 (types) and ==> (props); the default behaviour is "brackets";
  2940 
  2941 * Provers: fast_tac (and friends) now handle actual object-logic rules
  2942 as assumptions as well;
  2943 
  2944 * system: support Poly/ML 4.0;
  2945 
  2946 * system: isatool install handles KDE version 1 or 2;
  2947 
  2948 
  2949 
  2950 New in Isabelle99-1 (October 2000)
  2951 ----------------------------------
  2952 
  2953 *** Overview of INCOMPATIBILITIES ***
  2954 
  2955 * HOL: simplification of natural numbers is much changed; to partly
  2956 recover the old behaviour (e.g. to prevent n+n rewriting to #2*n)
  2957 issue the following ML commands:
  2958 
  2959   Delsimprocs Nat_Numeral_Simprocs.cancel_numerals;
  2960   Delsimprocs [Nat_Numeral_Simprocs.combine_numerals];
  2961 
  2962 * HOL: simplification no longer dives into case-expressions; this is
  2963 controlled by "t.weak_case_cong" for each datatype t;
  2964 
  2965 * HOL: nat_less_induct renamed to less_induct;
  2966 
  2967 * HOL: systematic renaming of the SOME (Eps) rules, may use isatool
  2968 fixsome to patch .thy and .ML sources automatically;
  2969 
  2970   select_equality  -> some_equality
  2971   select_eq_Ex     -> some_eq_ex
  2972   selectI2EX       -> someI2_ex
  2973   selectI2         -> someI2
  2974   selectI          -> someI
  2975   select1_equality -> some1_equality
  2976   Eps_sym_eq       -> some_sym_eq_trivial
  2977   Eps_eq           -> some_eq_trivial
  2978 
  2979 * HOL: exhaust_tac on datatypes superceded by new generic case_tac;
  2980 
  2981 * HOL: removed obsolete theorem binding expand_if (refer to split_if
  2982 instead);
  2983 
  2984 * HOL: the recursion equations generated by 'recdef' are now called
  2985 f.simps instead of f.rules;
  2986 
  2987 * HOL: qed_spec_mp now also handles bounded ALL as well;
  2988 
  2989 * HOL: 0 is now overloaded, so the type constraint ":: nat" may
  2990 sometimes be needed;
  2991 
  2992 * HOL: the constant for "f``x" is now "image" rather than "op ``";
  2993 
  2994 * HOL: the constant for "f-``x" is now "vimage" rather than "op -``";
  2995 
  2996 * HOL: the disjoint sum is now "<+>" instead of "Plus"; the cartesian
  2997 product is now "<*>" instead of "Times"; the lexicographic product is
  2998 now "<*lex*>" instead of "**";
  2999 
  3000 * HOL: theory Sexp is now in HOL/Induct examples (it used to be part
  3001 of main HOL, but was unused); better use HOL's datatype package;
  3002 
  3003 * HOL: removed "symbols" syntax for constant "override" of theory Map;
  3004 the old syntax may be recovered as follows:
  3005 
  3006   syntax (symbols)
  3007     override  :: "('a ~=> 'b) => ('a ~=> 'b) => ('a ~=> 'b)"
  3008       (infixl "\\<oplus>" 100)
  3009 
  3010 * HOL/Real: "rabs" replaced by overloaded "abs" function;
  3011 
  3012 * HOL/ML: even fewer consts are declared as global (see theories Ord,
  3013 Lfp, Gfp, WF); this only affects ML packages that refer to const names
  3014 internally;
  3015 
  3016 * HOL and ZF: syntax for quotienting wrt an equivalence relation
  3017 changed from A/r to A//r;
  3018 
  3019 * ZF: new treatment of arithmetic (nat & int) may break some old
  3020 proofs;
  3021 
  3022 * Isar: renamed some attributes (RS -> THEN, simplify -> simplified,
  3023 rulify -> rule_format, elimify -> elim_format, ...);
  3024 
  3025 * Isar/Provers: intro/elim/dest attributes changed; renamed
  3026 intro/intro!/intro!! flags to intro!/intro/intro? (in most cases, one
  3027 should have to change intro!! to intro? only); replaced "delrule" by
  3028 "rule del";
  3029 
  3030 * Isar/HOL: renamed "intrs" to "intros" in inductive definitions;
  3031 
  3032 * Provers: strengthened force_tac by using new first_best_tac;
  3033 
  3034 * LaTeX document preparation: several changes of isabelle.sty (see
  3035 lib/texinputs);
  3036 
  3037 
  3038 *** Document preparation ***
  3039 
  3040 * formal comments (text blocks etc.) in new-style theories may now
  3041 contain antiquotations of thm/prop/term/typ/text to be presented
  3042 according to latex print mode; concrete syntax is like this:
  3043 @{term[show_types] "f(x) = a + x"};
  3044 
  3045 * isatool mkdir provides easy setup of Isabelle session directories,
  3046 including proper document sources;
  3047 
  3048 * generated LaTeX sources are now deleted after successful run
  3049 (isatool document -c); may retain a copy somewhere else via -D option
  3050 of isatool usedir;
  3051 
  3052 * isatool usedir -D now lets isatool latex -o sty update the Isabelle
  3053 style files, achieving self-contained LaTeX sources and simplifying
  3054 LaTeX debugging;
  3055 
  3056 * old-style theories now produce (crude) LaTeX output as well;
  3057 
  3058 * browser info session directories are now self-contained (may be put
  3059 on WWW server seperately); improved graphs of nested sessions; removed
  3060 graph for 'all sessions';
  3061 
  3062 * several improvements in isabelle style files; \isabellestyle{it}
  3063 produces fake math mode output; \isamarkupheader is now \section by
  3064 default; see lib/texinputs/isabelle.sty etc.;
  3065 
  3066 
  3067 *** Isar ***
  3068 
  3069 * Isar/Pure: local results and corresponding term bindings are now
  3070 subject to Hindley-Milner polymorphism (similar to ML); this
  3071 accommodates incremental type-inference very nicely;
  3072 
  3073 * Isar/Pure: new derived language element 'obtain' supports
  3074 generalized existence reasoning;
  3075 
  3076 * Isar/Pure: new calculational elements 'moreover' and 'ultimately'
  3077 support accumulation of results, without applying any rules yet;
  3078 useful to collect intermediate results without explicit name
  3079 references, and for use with transitivity rules with more than 2
  3080 premises;
  3081 
  3082 * Isar/Pure: scalable support for case-analysis type proofs: new
  3083 'case' language element refers to local contexts symbolically, as
  3084 produced by certain proof methods; internally, case names are attached
  3085 to theorems as "tags";
  3086 
  3087 * Isar/Pure: theory command 'hide' removes declarations from
  3088 class/type/const name spaces;
  3089 
  3090 * Isar/Pure: theory command 'defs' supports option "(overloaded)" to
  3091 indicate potential overloading;
  3092 
  3093 * Isar/Pure: changed syntax of local blocks from {{ }} to { };
  3094 
  3095 * Isar/Pure: syntax of sorts made 'inner', i.e. have to write
  3096 "{a,b,c}" instead of {a,b,c};
  3097 
  3098 * Isar/Pure now provides its own version of intro/elim/dest
  3099 attributes; useful for building new logics, but beware of confusion
  3100 with the version in Provers/classical;
  3101 
  3102 * Isar/Pure: the local context of (non-atomic) goals is provided via
  3103 case name 'antecedent';
  3104 
  3105 * Isar/Pure: removed obsolete 'transfer' attribute (transfer of thms
  3106 to the current context is now done automatically);
  3107 
  3108 * Isar/Pure: theory command 'method_setup' provides a simple interface
  3109 for definining proof methods in ML;
  3110 
  3111 * Isar/Provers: intro/elim/dest attributes changed; renamed
  3112 intro/intro!/intro!! flags to intro!/intro/intro? (INCOMPATIBILITY, in
  3113 most cases, one should have to change intro!! to intro? only);
  3114 replaced "delrule" by "rule del";
  3115 
  3116 * Isar/Provers: new 'hypsubst' method, plain 'subst' method and
  3117 'symmetric' attribute (the latter supercedes [RS sym]);
  3118 
  3119 * Isar/Provers: splitter support (via 'split' attribute and 'simp'
  3120 method modifier); 'simp' method: 'only:' modifier removes loopers as
  3121 well (including splits);
  3122 
  3123 * Isar/Provers: Simplifier and Classical methods now support all kind
  3124 of modifiers used in the past, including 'cong', 'iff', etc.
  3125 
  3126 * Isar/Provers: added 'fastsimp' and 'clarsimp' methods (combination
  3127 of Simplifier and Classical reasoner);
  3128 
  3129 * Isar/HOL: new proof method 'cases' and improved version of 'induct'
  3130 now support named cases; major packages (inductive, datatype, primrec,
  3131 recdef) support case names and properly name parameters;
  3132 
  3133 * Isar/HOL: new transitivity rules for substitution in inequalities --
  3134 monotonicity conditions are extracted to be proven at end of
  3135 calculations;
  3136 
  3137 * Isar/HOL: removed 'case_split' thm binding, should use 'cases' proof
  3138 method anyway;
  3139 
  3140 * Isar/HOL: removed old expand_if = split_if; theorems if_splits =
  3141 split_if split_if_asm; datatype package provides theorems foo.splits =
  3142 foo.split foo.split_asm for each datatype;
  3143 
  3144 * Isar/HOL: tuned inductive package, rename "intrs" to "intros"
  3145 (potential INCOMPATIBILITY), emulation of mk_cases feature for proof
  3146 scripts: new 'inductive_cases' command and 'ind_cases' method; (Note:
  3147 use "(cases (simplified))" method in proper proof texts);
  3148 
  3149 * Isar/HOL: added global 'arith_split' attribute for 'arith' method;
  3150 
  3151 * Isar: names of theorems etc. may be natural numbers as well;
  3152 
  3153 * Isar: 'pr' command: optional arguments for goals_limit and
  3154 ProofContext.prems_limit; no longer prints theory contexts, but only
  3155 proof states;
  3156 
  3157 * Isar: diagnostic commands 'pr', 'thm', 'prop', 'term', 'typ' admit
  3158 additional print modes to be specified; e.g. "pr(latex)" will print
  3159 proof state according to the Isabelle LaTeX style;
  3160 
  3161 * Isar: improved support for emulating tactic scripts, including proof
  3162 methods 'rule_tac' etc., 'cut_tac', 'thin_tac', 'subgoal_tac',
  3163 'rename_tac', 'rotate_tac', 'tactic', and 'case_tac' / 'induct_tac'
  3164 (for HOL datatypes);
  3165 
  3166 * Isar: simplified (more robust) goal selection of proof methods: 1st
  3167 goal, all goals, or explicit goal specifier (tactic emulation); thus
  3168 'proof method scripts' have to be in depth-first order;
  3169 
  3170 * Isar: tuned 'let' syntax: replaced 'as' keyword by 'and';
  3171 
  3172 * Isar: removed 'help' command, which hasn't been too helpful anyway;
  3173 should instead use individual commands for printing items
  3174 (print_commands, print_methods etc.);
  3175 
  3176 * Isar: added 'nothing' --- the empty list of theorems;
  3177 
  3178 
  3179 *** HOL ***
  3180 
  3181 * HOL/MicroJava: formalization of a fragment of Java, together with a
  3182 corresponding virtual machine and a specification of its bytecode
  3183 verifier and a lightweight bytecode verifier, including proofs of
  3184 type-safety; by Gerwin Klein, Tobias Nipkow, David von Oheimb, and
  3185 Cornelia Pusch (see also the homepage of project Bali at
  3186 http://isabelle.in.tum.de/Bali/);
  3187 
  3188 * HOL/Algebra: new theory of rings and univariate polynomials, by
  3189 Clemens Ballarin;
  3190 
  3191 * HOL/NumberTheory: fundamental Theorem of Arithmetic, Chinese
  3192 Remainder Theorem, Fermat/Euler Theorem, Wilson's Theorem, by Thomas M
  3193 Rasmussen;
  3194 
  3195 * HOL/Lattice: fundamental concepts of lattice theory and order
  3196 structures, including duals, properties of bounds versus algebraic
  3197 laws, lattice operations versus set-theoretic ones, the Knaster-Tarski
  3198 Theorem for complete lattices etc.; may also serve as a demonstration
  3199 for abstract algebraic reasoning using axiomatic type classes, and
  3200 mathematics-style proof in Isabelle/Isar; by Markus Wenzel;
  3201 
  3202 * HOL/Prolog: a (bare-bones) implementation of Lambda-Prolog, by David
  3203 von Oheimb;
  3204 
  3205 * HOL/IMPP: extension of IMP with local variables and mutually
  3206 recursive procedures, by David von Oheimb;
  3207 
  3208 * HOL/Lambda: converted into new-style theory and document;
  3209 
  3210 * HOL/ex/Multiquote: example of multiple nested quotations and
  3211 anti-quotations -- basically a generalized version of de-Bruijn
  3212 representation; very useful in avoiding lifting of operations;
  3213 
  3214 * HOL/record: added general record equality rule to simpset; fixed
  3215 select-update simplification procedure to handle extended records as
  3216 well; admit "r" as field name;
  3217 
  3218 * HOL: 0 is now overloaded over the new sort "zero", allowing its use with
  3219 other numeric types and also as the identity of groups, rings, etc.;
  3220 
  3221 * HOL: new axclass plus_ac0 for addition with the AC-laws and 0 as identity.
  3222 Types nat and int belong to this axclass;
  3223 
  3224 * HOL: greatly improved simplification involving numerals of type nat, int, real:
  3225    (i + #8 + j) = Suc k simplifies to  #7 + (i + j) = k
  3226    i*j + k + j*#3*i     simplifies to  #4*(i*j) + k
  3227   two terms #m*u and #n*u are replaced by #(m+n)*u
  3228     (where #m, #n and u can implicitly be 1; this is simproc combine_numerals)
  3229   and the term/formula #m*u+x ~~ #n*u+y simplifies simplifies to #(m-n)+x ~~ y
  3230     or x ~~ #(n-m)+y, where ~~ is one of = < <= or - (simproc cancel_numerals);
  3231 
  3232 * HOL: meson_tac is available (previously in ex/meson.ML); it is a
  3233 powerful prover for predicate logic but knows nothing of clasets; see
  3234 ex/mesontest.ML and ex/mesontest2.ML for example applications;
  3235 
  3236 * HOL: new version of "case_tac" subsumes both boolean case split and
  3237 "exhaust_tac" on datatypes; INCOMPATIBILITY: exhaust_tac no longer
  3238 exists, may define val exhaust_tac = case_tac for ad-hoc portability;
  3239 
  3240 * HOL: simplification no longer dives into case-expressions: only the
  3241 selector expression is simplified, but not the remaining arms; to
  3242 enable full simplification of case-expressions for datatype t, you may
  3243 remove t.weak_case_cong from the simpset, either globally (Delcongs
  3244 [thm"t.weak_case_cong"];) or locally (delcongs [...]).
  3245 
  3246 * HOL/recdef: the recursion equations generated by 'recdef' for
  3247 function 'f' are now called f.simps instead of f.rules; if all
  3248 termination conditions are proved automatically, these simplification
  3249 rules are added to the simpset, as in primrec; rules may be named
  3250 individually as well, resulting in a separate list of theorems for
  3251 each equation;
  3252 
  3253 * HOL/While is a new theory that provides a while-combinator. It
  3254 permits the definition of tail-recursive functions without the
  3255 provision of a termination measure. The latter is necessary once the
  3256 invariant proof rule for while is applied.
  3257 
  3258 * HOL: new (overloaded) notation for the set of elements below/above
  3259 some element: {..u}, {..u(}, {l..}, {)l..}. See theory SetInterval.
  3260 
  3261 * HOL: theorems impI, allI, ballI bound as "strip";
  3262 
  3263 * HOL: new tactic induct_thm_tac: thm -> string -> int -> tactic
  3264 induct_tac th "x1 ... xn" expects th to have a conclusion of the form
  3265 P v1 ... vn and abbreviates res_inst_tac [("v1","x1"),...,("vn","xn")] th;
  3266 
  3267 * HOL/Real: "rabs" replaced by overloaded "abs" function;
  3268 
  3269 * HOL: theory Sexp now in HOL/Induct examples (it used to be part of
  3270 main HOL, but was unused);
  3271 
  3272 * HOL: fewer consts declared as global (e.g. have to refer to
  3273 "Lfp.lfp" instead of "lfp" internally; affects ML packages only);
  3274 
  3275 * HOL: tuned AST representation of nested pairs, avoiding bogus output
  3276 in case of overlap with user translations (e.g. judgements over
  3277 tuples); (note that the underlying logical represenation is still
  3278 bogus);
  3279 
  3280 
  3281 *** ZF ***
  3282 
  3283 * ZF: simplification automatically cancels common terms in arithmetic
  3284 expressions over nat and int;
  3285 
  3286 * ZF: new treatment of nat to minimize type-checking: all operators
  3287 coerce their operands to a natural number using the function natify,
  3288 making the algebraic laws unconditional;
  3289 
  3290 * ZF: as above, for int: operators coerce their operands to an integer
  3291 using the function intify;
  3292 
  3293 * ZF: the integer library now contains many of the usual laws for the
  3294 orderings, including $<=, and monotonicity laws for $+ and $*;
  3295 
  3296 * ZF: new example ZF/ex/NatSum to demonstrate integer arithmetic
  3297 simplification;
  3298 
  3299 * FOL and ZF: AddIffs now available, giving theorems of the form P<->Q
  3300 to the simplifier and classical reasoner simultaneously;
  3301 
  3302 
  3303 *** General ***
  3304 
  3305 * Provers: blast_tac now handles actual object-logic rules as
  3306 assumptions; note that auto_tac uses blast_tac internally as well;
  3307 
  3308 * Provers: new functions rulify/rulify_no_asm: thm -> thm for turning
  3309 outer -->/All/Ball into ==>/!!; qed_spec_mp now uses rulify_no_asm;
  3310 
  3311 * Provers: delrules now handles destruct rules as well (no longer need
  3312 explicit make_elim);
  3313 
  3314 * Provers: Blast_tac now warns of and ignores "weak elimination rules" e.g.
  3315   [| inj ?f;          ?f ?x = ?f ?y; ?x = ?y ==> ?W |] ==> ?W
  3316 use instead the strong form,
  3317   [| inj ?f; ~ ?W ==> ?f ?x = ?f ?y; ?x = ?y ==> ?W |] ==> ?W
  3318 in HOL, FOL and ZF the function cla_make_elim will create such rules
  3319 from destruct-rules;
  3320 
  3321 * Provers: Simplifier.easy_setup provides a fast path to basic
  3322 Simplifier setup for new object-logics;
  3323 
  3324 * Pure: AST translation rules no longer require constant head on LHS;
  3325 
  3326 * Pure: improved name spaces: ambiguous output is qualified; support
  3327 for hiding of names;
  3328 
  3329 * system: smart setup of canonical ML_HOME, ISABELLE_INTERFACE, and
  3330 XSYMBOL_HOME; no longer need to do manual configuration in most
  3331 situations;
  3332 
  3333 * system: compression of ML heaps images may now be controlled via -c
  3334 option of isabelle and isatool usedir (currently only observed by
  3335 Poly/ML);
  3336 
  3337 * system: isatool installfonts may handle X-Symbol fonts as well (very
  3338 useful for remote X11);
  3339 
  3340 * system: provide TAGS file for Isabelle sources;
  3341 
  3342 * ML: infix 'OF' is a version of 'MRS' with more appropriate argument
  3343 order;
  3344 
  3345 * ML: renamed flags Syntax.trace_norm_ast to Syntax.trace_ast; global
  3346 timing flag supersedes proof_timing and Toplevel.trace;
  3347 
  3348 * ML: new combinators |>> and |>>> for incremental transformations
  3349 with secondary results (e.g. certain theory extensions):
  3350 
  3351 * ML: PureThy.add_defs gets additional argument to indicate potential
  3352 overloading (usually false);
  3353 
  3354 * ML: PureThy.add_thms/add_axioms/add_defs now return theorems as
  3355 results;
  3356 
  3357 
  3358 
  3359 New in Isabelle99 (October 1999)
  3360 --------------------------------
  3361 
  3362 *** Overview of INCOMPATIBILITIES (see below for more details) ***
  3363 
  3364 * HOL: The THEN and ELSE parts of conditional expressions (if P then x else y)
  3365 are no longer simplified.  (This allows the simplifier to unfold recursive
  3366 functional programs.)  To restore the old behaviour, declare
  3367 
  3368     Delcongs [if_weak_cong];
  3369 
  3370 * HOL: Removed the obsolete syntax "Compl A"; use -A for set
  3371 complement;
  3372 
  3373 * HOL: the predicate "inj" is now defined by translation to "inj_on";
  3374 
  3375 * HOL/datatype: mutual_induct_tac no longer exists --
  3376   use induct_tac "x_1 ... x_n" instead of mutual_induct_tac ["x_1", ..., "x_n"]
  3377 
  3378 * HOL/typedef: fixed type inference for representing set; type
  3379 arguments now have to occur explicitly on the rhs as type constraints;
  3380 
  3381 * ZF: The con_defs part of an inductive definition may no longer refer
  3382 to constants declared in the same theory;
  3383 
  3384 * HOL, ZF: the function mk_cases, generated by the inductive
  3385 definition package, has lost an argument.  To simplify its result, it
  3386 uses the default simpset instead of a supplied list of theorems.
  3387 
  3388 * HOL/List: the constructors of type list are now Nil and Cons;
  3389 
  3390 * Simplifier: the type of the infix ML functions
  3391         setSSolver addSSolver setSolver addSolver
  3392 is now  simpset * solver -> simpset  where `solver' is a new abstract type
  3393 for packaging solvers. A solver is created via
  3394         mk_solver: string -> (thm list -> int -> tactic) -> solver
  3395 where the string argument is only a comment.
  3396 
  3397 
  3398 *** Proof tools ***
  3399 
  3400 * Provers/Arith/fast_lin_arith.ML contains a functor for creating a
  3401 decision procedure for linear arithmetic. Currently it is used for
  3402 types `nat', `int', and `real' in HOL (see below); it can, should and
  3403 will be instantiated for other types and logics as well.
  3404 
  3405 * The simplifier now accepts rewrite rules with flexible heads, eg
  3406      hom ?f ==> ?f(?x+?y) = ?f ?x + ?f ?y
  3407   They are applied like any rule with a non-pattern lhs, i.e. by first-order
  3408   matching.
  3409 
  3410 
  3411 *** General ***
  3412 
  3413 * New Isabelle/Isar subsystem provides an alternative to traditional
  3414 tactical theorem proving; together with the ProofGeneral/isar user
  3415 interface it offers an interactive environment for developing human
  3416 readable proof documents (Isar == Intelligible semi-automated
  3417 reasoning); for further information see isatool doc isar-ref,
  3418 src/HOL/Isar_examples and http://isabelle.in.tum.de/Isar/
  3419 
  3420 * improved and simplified presentation of theories: better HTML markup
  3421 (including colors), graph views in several sizes; isatool usedir now
  3422 provides a proper interface for user theories (via -P option); actual
  3423 document preparation based on (PDF)LaTeX is available as well (for
  3424 new-style theories only); see isatool doc system for more information;
  3425 
  3426 * native support for Proof General, both for classic Isabelle and
  3427 Isabelle/Isar;
  3428 
  3429 * ML function thm_deps visualizes dependencies of theorems and lemmas,
  3430 using the graph browser tool;
  3431 
  3432 * Isabelle manuals now also available as PDF;
  3433 
  3434 * theory loader rewritten from scratch (may not be fully
  3435 bug-compatible); old loadpath variable has been replaced by show_path,
  3436 add_path, del_path, reset_path functions; new operations such as
  3437 update_thy, touch_thy, remove_thy, use/update_thy_only (see also
  3438 isatool doc ref);
  3439 
  3440 * improved isatool install: option -k creates KDE application icon,
  3441 option -p DIR installs standalone binaries;
  3442 
  3443 * added ML_PLATFORM setting (useful for cross-platform installations);
  3444 more robust handling of platform specific ML images for SML/NJ;
  3445 
  3446 * the settings environment is now statically scoped, i.e. it is never
  3447 created again in sub-processes invoked from isabelle, isatool, or
  3448 Isabelle;
  3449 
  3450 * path element specification '~~' refers to '$ISABELLE_HOME';
  3451 
  3452 * in locales, the "assumes" and "defines" parts may be omitted if
  3453 empty;
  3454 
  3455 * new print_mode "xsymbols" for extended symbol support (e.g. genuine
  3456 long arrows);
  3457 
  3458 * new print_mode "HTML";
  3459 
  3460 * new flag show_tags controls display of tags of theorems (which are
  3461 basically just comments that may be attached by some tools);
  3462 
  3463 * Isamode 2.6 requires patch to accomodate change of Isabelle font
  3464 mode and goal output format:
  3465 
  3466 diff -r Isamode-2.6/elisp/isa-load.el Isamode/elisp/isa-load.el
  3467 244c244
  3468 <       (list (isa-getenv "ISABELLE") "-msymbols" logic-name)
  3469 ---
  3470 >       (list (isa-getenv "ISABELLE") "-misabelle_font" "-msymbols" logic-name)
  3471 diff -r Isabelle-2.6/elisp/isa-proofstate.el Isamode/elisp/isa-proofstate.el
  3472 181c181
  3473 < (defconst proofstate-proofstart-regexp "^Level [0-9]+$"
  3474 ---
  3475 > (defconst proofstate-proofstart-regexp "^Level [0-9]+"
  3476 
  3477 * function bind_thms stores lists of theorems (cf. bind_thm);
  3478 
  3479 * new shorthand tactics ftac, eatac, datac, fatac;
  3480 
  3481 * qed (and friends) now accept "" as result name; in that case the
  3482 theorem is not stored, but proper checks and presentation of the
  3483 result still apply;
  3484 
  3485 * theorem database now also indexes constants "Trueprop", "all",
  3486 "==>", "=="; thus thms_containing, findI etc. may retrieve more rules;
  3487 
  3488 
  3489 *** HOL ***
  3490 
  3491 ** HOL arithmetic **
  3492 
  3493 * There are now decision procedures for linear arithmetic over nat and
  3494 int:
  3495 
  3496 1. arith_tac copes with arbitrary formulae involving `=', `<', `<=',
  3497 `+', `-', `Suc', `min', `max' and numerical constants; other subterms
  3498 are treated as atomic; subformulae not involving type `nat' or `int'
  3499 are ignored; quantified subformulae are ignored unless they are
  3500 positive universal or negative existential. The tactic has to be
  3501 invoked by hand and can be a little bit slow. In particular, the
  3502 running time is exponential in the number of occurrences of `min' and
  3503 `max', and `-' on `nat'.
  3504 
  3505 2. fast_arith_tac is a cut-down version of arith_tac: it only takes
  3506 (negated) (in)equalities among the premises and the conclusion into
  3507 account (i.e. no compound formulae) and does not know about `min' and
  3508 `max', and `-' on `nat'. It is fast and is used automatically by the
  3509 simplifier.
  3510 
  3511 NB: At the moment, these decision procedures do not cope with mixed
  3512 nat/int formulae where the two parts interact, such as `m < n ==>
  3513 int(m) < int(n)'.
  3514 
  3515 * HOL/Numeral provides a generic theory of numerals (encoded
  3516 efficiently as bit strings); setup for types nat/int/real is in place;
  3517 INCOMPATIBILITY: since numeral syntax is now polymorphic, rather than
  3518 int, existing theories and proof scripts may require a few additional
  3519 type constraints;
  3520 
  3521 * integer division and remainder can now be performed on constant
  3522 arguments;
  3523 
  3524 * many properties of integer multiplication, division and remainder
  3525 are now available;
  3526 
  3527 * An interface to the Stanford Validity Checker (SVC) is available through the
  3528 tactic svc_tac.  Propositional tautologies and theorems of linear arithmetic
  3529 are proved automatically.  SVC must be installed separately, and its results
  3530 must be TAKEN ON TRUST (Isabelle does not check the proofs, but tags any
  3531 invocation of the underlying oracle).  For SVC see
  3532   http://verify.stanford.edu/SVC
  3533 
  3534 * IsaMakefile: the HOL-Real target now builds an actual image;
  3535 
  3536 
  3537 ** HOL misc **
  3538 
  3539 * HOL/Real/HahnBanach: the Hahn-Banach theorem for real vector spaces
  3540 (in Isabelle/Isar) -- by Gertrud Bauer;
  3541 
  3542 * HOL/BCV: generic model of bytecode verification, i.e. data-flow
  3543 analysis for assembly languages with subtypes;
  3544 
  3545 * HOL/TLA (Lamport's Temporal Logic of Actions): major reorganization
  3546 -- avoids syntactic ambiguities and treats state, transition, and
  3547 temporal levels more uniformly; introduces INCOMPATIBILITIES due to
  3548 changed syntax and (many) tactics;
  3549 
  3550 * HOL/inductive: Now also handles more general introduction rules such
  3551   as "ALL y. (y, x) : r --> y : acc r ==> x : acc r"; monotonicity
  3552   theorems are now maintained within the theory (maintained via the
  3553   "mono" attribute);
  3554 
  3555 * HOL/datatype: Now also handles arbitrarily branching datatypes
  3556   (using function types) such as
  3557 
  3558   datatype 'a tree = Atom 'a | Branch "nat => 'a tree"
  3559 
  3560 * HOL/record: record_simproc (part of the default simpset) takes care
  3561 of selectors applied to updated records; record_split_tac is no longer
  3562 part of the default claset; update_defs may now be removed from the
  3563 simpset in many cases; COMPATIBILITY: old behavior achieved by
  3564 
  3565   claset_ref () := claset() addSWrapper record_split_wrapper;
  3566   Delsimprocs [record_simproc]
  3567 
  3568 * HOL/typedef: fixed type inference for representing set; type
  3569 arguments now have to occur explicitly on the rhs as type constraints;
  3570 
  3571 * HOL/recdef (TFL): 'congs' syntax now expects comma separated list of theorem
  3572 names rather than an ML expression;
  3573 
  3574 * HOL/defer_recdef (TFL): like recdef but the well-founded relation can be
  3575 supplied later.  Program schemes can be defined, such as
  3576     "While B C s = (if B s then While B C (C s) else s)"
  3577 where the well-founded relation can be chosen after B and C have been given.
  3578 
  3579 * HOL/List: the constructors of type list are now Nil and Cons;
  3580 INCOMPATIBILITY: while [] and infix # syntax is still there, of
  3581 course, ML tools referring to List.list.op # etc. have to be adapted;
  3582 
  3583 * HOL_quantifiers flag superseded by "HOL" print mode, which is
  3584 disabled by default; run isabelle with option -m HOL to get back to
  3585 the original Gordon/HOL-style output;
  3586 
  3587 * HOL/Ord.thy: new bounded quantifier syntax (input only): ALL x<y. P,
  3588 ALL x<=y. P, EX x<y. P, EX x<=y. P;
  3589 
  3590 * HOL basic syntax simplified (more orthogonal): all variants of
  3591 All/Ex now support plain / symbolic / HOL notation; plain syntax for
  3592 Eps operator is provided as well: "SOME x. P[x]";
  3593 
  3594 * HOL/Sum.thy: sum_case has been moved to HOL/Datatype;
  3595 
  3596 * HOL/Univ.thy: infix syntax <*>, <+>, <**>, <+> eliminated and made
  3597 thus available for user theories;
  3598 
  3599 * HOLCF/IOA/Sequents: renamed 'Cons' to 'Consq' to avoid clash with
  3600 HOL/List; hardly an INCOMPATIBILITY since '>>' syntax is used all the
  3601 time;
  3602 
  3603 * HOL: new tactic smp_tac: int -> int -> tactic, which applies spec
  3604 several times and then mp;
  3605 
  3606 
  3607 *** LK ***
  3608 
  3609 * the notation <<...>> is now available as a notation for sequences of
  3610 formulas;
  3611 
  3612 * the simplifier is now installed
  3613 
  3614 * the axiom system has been generalized (thanks to Soren Heilmann)
  3615 
  3616 * the classical reasoner now has a default rule database
  3617 
  3618 
  3619 *** ZF ***
  3620 
  3621 * new primrec section allows primitive recursive functions to be given
  3622 directly (as in HOL) over datatypes and the natural numbers;
  3623 
  3624 * new tactics induct_tac and exhaust_tac for induction (or case
  3625 analysis) over datatypes and the natural numbers;
  3626 
  3627 * the datatype declaration of type T now defines the recursor T_rec;
  3628 
  3629 * simplification automatically does freeness reasoning for datatype
  3630 constructors;
  3631 
  3632 * automatic type-inference, with AddTCs command to insert new
  3633 type-checking rules;
  3634 
  3635 * datatype introduction rules are now added as Safe Introduction rules
  3636 to the claset;
  3637 
  3638 * the syntax "if P then x else y" is now available in addition to
  3639 if(P,x,y);
  3640 
  3641 
  3642 *** Internal programming interfaces ***
  3643 
  3644 * tuned simplifier trace output; new flag debug_simp;
  3645 
  3646 * structures Vartab / Termtab (instances of TableFun) offer efficient
  3647 tables indexed by indexname_ord / term_ord (compatible with aconv);
  3648 
  3649 * AxClass.axclass_tac lost the theory argument;
  3650 
  3651 * tuned current_goals_markers semantics: begin / end goal avoids
  3652 printing empty lines;
  3653 
  3654 * removed prs and prs_fn hook, which was broken because it did not
  3655 include \n in its semantics, forcing writeln to add one
  3656 uncoditionally; replaced prs_fn by writeln_fn; consider std_output:
  3657 string -> unit if you really want to output text without newline;
  3658 
  3659 * Symbol.output subject to print mode; INCOMPATIBILITY: defaults to
  3660 plain output, interface builders may have to enable 'isabelle_font'
  3661 mode to get Isabelle font glyphs as before;
  3662 
  3663 * refined token_translation interface; INCOMPATIBILITY: output length
  3664 now of type real instead of int;
  3665 
  3666 * theory loader actions may be traced via new ThyInfo.add_hook
  3667 interface (see src/Pure/Thy/thy_info.ML); example application: keep
  3668 your own database of information attached to *whole* theories -- as
  3669 opposed to intra-theory data slots offered via TheoryDataFun;
  3670 
  3671 * proper handling of dangling sort hypotheses (at last!);
  3672 Thm.strip_shyps and Drule.strip_shyps_warning take care of removing
  3673 extra sort hypotheses that can be witnessed from the type signature;
  3674 the force_strip_shyps flag is gone, any remaining shyps are simply
  3675 left in the theorem (with a warning issued by strip_shyps_warning);
  3676 
  3677 
  3678 
  3679 New in Isabelle98-1 (October 1998)
  3680 ----------------------------------
  3681 
  3682 *** Overview of INCOMPATIBILITIES (see below for more details) ***
  3683 
  3684 * several changes of automated proof tools;
  3685 
  3686 * HOL: major changes to the inductive and datatype packages, including
  3687 some minor incompatibilities of theory syntax;
  3688 
  3689 * HOL: renamed r^-1 to 'converse' from 'inverse'; 'inj_onto' is now
  3690 called `inj_on';
  3691 
  3692 * HOL: removed duplicate thms in Arith:
  3693   less_imp_add_less  should be replaced by  trans_less_add1
  3694   le_imp_add_le      should be replaced by  trans_le_add1
  3695 
  3696 * HOL: unary minus is now overloaded (new type constraints may be
  3697 required);
  3698 
  3699 * HOL and ZF: unary minus for integers is now #- instead of #~.  In
  3700 ZF, expressions such as n#-1 must be changed to n#- 1, since #-1 is
  3701 now taken as an integer constant.
  3702 
  3703 * Pure: ML function 'theory_of' renamed to 'theory';
  3704 
  3705 
  3706 *** Proof tools ***
  3707 
  3708 * Simplifier:
  3709   1. Asm_full_simp_tac is now more aggressive.
  3710      1. It will sometimes reorient premises if that increases their power to
  3711         simplify.
  3712      2. It does no longer proceed strictly from left to right but may also
  3713         rotate premises to achieve further simplification.
  3714      For compatibility reasons there is now Asm_lr_simp_tac which is like the
  3715      old Asm_full_simp_tac in that it does not rotate premises.
  3716   2. The simplifier now knows a little bit about nat-arithmetic.
  3717 
  3718 * Classical reasoner: wrapper mechanism for the classical reasoner now
  3719 allows for selected deletion of wrappers, by introduction of names for
  3720 wrapper functionals.  This implies that addbefore, addSbefore,
  3721 addaltern, and addSaltern now take a pair (name, tactic) as argument,
  3722 and that adding two tactics with the same name overwrites the first
  3723 one (emitting a warning).
  3724   type wrapper = (int -> tactic) -> (int -> tactic)
  3725   setWrapper, setSWrapper, compWrapper and compSWrapper are replaced by
  3726   addWrapper, addSWrapper: claset * (string * wrapper) -> claset
  3727   delWrapper, delSWrapper: claset *  string            -> claset
  3728   getWrapper is renamed to appWrappers, getSWrapper to appSWrappers;
  3729 
  3730 * Classical reasoner: addbefore/addSbefore now have APPEND/ORELSE
  3731 semantics; addbefore now affects only the unsafe part of step_tac
  3732 etc.; this affects addss/auto_tac/force_tac, so EXISTING PROOFS MAY
  3733 FAIL, but proofs should be fixable easily, e.g. by replacing Auto_tac
  3734 by Force_tac;
  3735 
  3736 * Classical reasoner: setwrapper to setWrapper and compwrapper to
  3737 compWrapper; added safe wrapper (and access functions for it);
  3738 
  3739 * HOL/split_all_tac is now much faster and fails if there is nothing
  3740 to split.  Some EXISTING PROOFS MAY REQUIRE ADAPTION because the order
  3741 and the names of the automatically generated variables have changed.
  3742 split_all_tac has moved within claset() from unsafe wrappers to safe
  3743 wrappers, which means that !!-bound variables are split much more
  3744 aggressively, and safe_tac and clarify_tac now split such variables.
  3745 If this splitting is not appropriate, use delSWrapper "split_all_tac".
  3746 Note: the same holds for record_split_tac, which does the job of
  3747 split_all_tac for record fields.
  3748 
  3749 * HOL/Simplifier: Rewrite rules for case distinctions can now be added
  3750 permanently to the default simpset using Addsplits just like
  3751 Addsimps. They can be removed via Delsplits just like
  3752 Delsimps. Lower-case versions are also available.
  3753 
  3754 * HOL/Simplifier: The rule split_if is now part of the default
  3755 simpset. This means that the simplifier will eliminate all occurrences
  3756 of if-then-else in the conclusion of a goal. To prevent this, you can
  3757 either remove split_if completely from the default simpset by
  3758 `Delsplits [split_if]' or remove it in a specific call of the
  3759 simplifier using `... delsplits [split_if]'.  You can also add/delete
  3760 other case splitting rules to/from the default simpset: every datatype
  3761 generates suitable rules `split_t_case' and `split_t_case_asm' (where
  3762 t is the name of the datatype).
  3763 
  3764 * Classical reasoner / Simplifier combination: new force_tac (and
  3765 derivatives Force_tac, force) combines rewriting and classical
  3766 reasoning (and whatever other tools) similarly to auto_tac, but is
  3767 aimed to solve the given subgoal completely.
  3768 
  3769 
  3770 *** General ***
  3771 
  3772 * new top-level commands `Goal' and `Goalw' that improve upon `goal'
  3773 and `goalw': the theory is no longer needed as an explicit argument -
  3774 the current theory context is used; assumptions are no longer returned
  3775 at the ML-level unless one of them starts with ==> or !!; it is
  3776 recommended to convert to these new commands using isatool fixgoal
  3777 (backup your sources first!);
  3778 
  3779 * new top-level commands 'thm' and 'thms' for retrieving theorems from
  3780 the current theory context, and 'theory' to lookup stored theories;
  3781 
  3782 * new theory section 'locale' for declaring constants, assumptions and
  3783 definitions that have local scope;
  3784 
  3785 * new theory section 'nonterminals' for purely syntactic types;
  3786 
  3787 * new theory section 'setup' for generic ML setup functions
  3788 (e.g. package initialization);
  3789 
  3790 * the distribution now includes Isabelle icons: see
  3791 lib/logo/isabelle-{small,tiny}.xpm;
  3792 
  3793 * isatool install - install binaries with absolute references to
  3794 ISABELLE_HOME/bin;
  3795 
  3796 * isatool logo -- create instances of the Isabelle logo (as EPS);
  3797 
  3798 * print mode 'emacs' reserved for Isamode;
  3799 
  3800 * support multiple print (ast) translations per constant name;
  3801 
  3802 * theorems involving oracles are now printed with a suffixed [!];
  3803 
  3804 
  3805 *** HOL ***
  3806 
  3807 * there is now a tutorial on Isabelle/HOL (do 'isatool doc tutorial');
  3808 
  3809 * HOL/inductive package reorganized and improved: now supports mutual
  3810 definitions such as
  3811 
  3812   inductive EVEN ODD
  3813     intrs
  3814       null "0 : EVEN"
  3815       oddI "n : EVEN ==> Suc n : ODD"
  3816       evenI "n : ODD ==> Suc n : EVEN"
  3817 
  3818 new theorem list "elims" contains an elimination rule for each of the
  3819 recursive sets; inductive definitions now handle disjunctive premises
  3820 correctly (also ZF);
  3821 
  3822 INCOMPATIBILITIES: requires Inductive as an ancestor; component
  3823 "mutual_induct" no longer exists - the induction rule is always
  3824 contained in "induct";
  3825 
  3826 
  3827 * HOL/datatype package re-implemented and greatly improved: now
  3828 supports mutually recursive datatypes such as
  3829 
  3830   datatype
  3831     'a aexp = IF_THEN_ELSE ('a bexp) ('a aexp) ('a aexp)
  3832             | SUM ('a aexp) ('a aexp)
  3833             | DIFF ('a aexp) ('a aexp)
  3834             | NUM 'a
  3835   and
  3836     'a bexp = LESS ('a aexp) ('a aexp)
  3837             | AND ('a bexp) ('a bexp)
  3838             | OR ('a bexp) ('a bexp)
  3839 
  3840 as well as indirectly recursive datatypes such as
  3841 
  3842   datatype
  3843     ('a, 'b) term = Var 'a
  3844                   | App 'b ((('a, 'b) term) list)
  3845 
  3846 The new tactic  mutual_induct_tac [<var_1>, ..., <var_n>] i  performs
  3847 induction on mutually / indirectly recursive datatypes.
  3848 
  3849 Primrec equations are now stored in theory and can be accessed via
  3850 <function_name>.simps.
  3851 
  3852 INCOMPATIBILITIES:
  3853 
  3854   - Theories using datatypes must now have theory Datatype as an
  3855     ancestor.
  3856   - The specific <typename>.induct_tac no longer exists - use the
  3857     generic induct_tac instead.
  3858   - natE has been renamed to nat.exhaust - use exhaust_tac
  3859     instead of res_inst_tac ... natE. Note that the variable
  3860     names in nat.exhaust differ from the names in natE, this
  3861     may cause some "fragile" proofs to fail.
  3862   - The theorems split_<typename>_case and split_<typename>_case_asm
  3863     have been renamed to <typename>.split and <typename>.split_asm.
  3864   - Since default sorts of type variables are now handled correctly,
  3865     some datatype definitions may have to be annotated with explicit
  3866     sort constraints.
  3867   - Primrec definitions no longer require function name and type
  3868     of recursive argument.
  3869 
  3870 Consider using isatool fixdatatype to adapt your theories and proof
  3871 scripts to the new package (backup your sources first!).
  3872 
  3873 
  3874 * HOL/record package: considerably improved implementation; now
  3875 includes concrete syntax for record types, terms, updates; theorems
  3876 for surjective pairing and splitting !!-bound record variables; proof
  3877 support is as follows:
  3878 
  3879   1) standard conversions (selectors or updates applied to record
  3880 constructor terms) are part of the standard simpset;
  3881 
  3882   2) inject equations of the form ((x, y) = (x', y')) == x=x' & y=y' are
  3883 made part of standard simpset and claset via addIffs;
  3884 
  3885   3) a tactic for record field splitting (record_split_tac) is part of
  3886 the standard claset (addSWrapper);
  3887 
  3888 To get a better idea about these rules you may retrieve them via
  3889 something like 'thms "foo.simps"' or 'thms "foo.iffs"', where "foo" is
  3890 the name of your record type.
  3891 
  3892 The split tactic 3) conceptually simplifies by the following rule:
  3893 
  3894   "(!!x. PROP ?P x) == (!!a b. PROP ?P (a, b))"
  3895 
  3896 Thus any record variable that is bound by meta-all will automatically
  3897 blow up into some record constructor term, consequently the
  3898 simplifications of 1), 2) apply.  Thus force_tac, auto_tac etc. shall
  3899 solve record problems automatically.
  3900 
  3901 
  3902 * reorganized the main HOL image: HOL/Integ and String loaded by
  3903 default; theory Main includes everything;
  3904 
  3905 * automatic simplification of integer sums and comparisons, using cancellation;
  3906 
  3907 * added option_map_eq_Some and not_Some_eq to the default simpset and claset;
  3908 
  3909 * added disj_not1 = "(~P | Q) = (P --> Q)" to the default simpset;
  3910 
  3911 * many new identities for unions, intersections, set difference, etc.;
  3912 
  3913 * expand_if, expand_split, expand_sum_case and expand_nat_case are now
  3914 called split_if, split_split, split_sum_case and split_nat_case (to go
  3915 with add/delsplits);
  3916 
  3917 * HOL/Prod introduces simplification procedure unit_eq_proc rewriting
  3918 (?x::unit) = (); this is made part of the default simpset, which COULD
  3919 MAKE EXISTING PROOFS FAIL under rare circumstances (consider
  3920 'Delsimprocs [unit_eq_proc];' as last resort); also note that
  3921 unit_abs_eta_conv is added in order to counter the effect of
  3922 unit_eq_proc on (%u::unit. f u), replacing it by f rather than by
  3923 %u.f();
  3924 
  3925 * HOL/Fun INCOMPATIBILITY: `inj_onto' is now called `inj_on' (which
  3926 makes more sense);
  3927 
  3928 * HOL/Set INCOMPATIBILITY: rule `equals0D' is now a well-formed destruct rule;
  3929   It and 'sym RS equals0D' are now in the default  claset, giving automatic
  3930   disjointness reasoning but breaking a few old proofs.
  3931 
  3932 * HOL/Relation INCOMPATIBILITY: renamed the relational operator r^-1
  3933 to 'converse' from 'inverse' (for compatibility with ZF and some
  3934 literature);
  3935 
  3936 * HOL/recdef can now declare non-recursive functions, with {} supplied as
  3937 the well-founded relation;
  3938 
  3939 * HOL/Set INCOMPATIBILITY: the complement of set A is now written -A instead of
  3940     Compl A.  The "Compl" syntax remains available as input syntax for this
  3941     release ONLY.
  3942 
  3943 * HOL/Update: new theory of function updates:
  3944     f(a:=b) == %x. if x=a then b else f x
  3945 may also be iterated as in f(a:=b,c:=d,...);
  3946 
  3947 * HOL/Vimage: new theory for inverse image of a function, syntax f-``B;
  3948 
  3949 * HOL/List:
  3950   - new function list_update written xs[i:=v] that updates the i-th
  3951     list position. May also be iterated as in xs[i:=a,j:=b,...].
  3952   - new function `upt' written [i..j(] which generates the list
  3953     [i,i+1,...,j-1], i.e. the upper bound is excluded. To include the upper
  3954     bound write [i..j], which is a shorthand for [i..j+1(].
  3955   - new lexicographic orderings and corresponding wellfoundedness theorems.
  3956 
  3957 * HOL/Arith:
  3958   - removed 'pred' (predecessor) function;
  3959   - generalized some theorems about n-1;
  3960   - many new laws about "div" and "mod";
  3961   - new laws about greatest common divisors (see theory ex/Primes);
  3962 
  3963 * HOL/Relation: renamed the relational operator r^-1 "converse"
  3964 instead of "inverse";
  3965 
  3966 * HOL/Induct/Multiset: a theory of multisets, including the wellfoundedness
  3967   of the multiset ordering;
  3968 
  3969 * directory HOL/Real: a construction of the reals using Dedekind cuts
  3970   (not included by default);
  3971 
  3972 * directory HOL/UNITY: Chandy and Misra's UNITY formalism;
  3973 
  3974 * directory HOL/Hoare: a new version of Hoare logic which permits many-sorted
  3975   programs, i.e. different program variables may have different types.
  3976 
  3977 * calling (stac rew i) now fails if "rew" has no effect on the goal
  3978   [previously, this check worked only if the rewrite rule was unconditional]
  3979   Now rew can involve either definitions or equalities (either == or =).
  3980 
  3981 
  3982 *** ZF ***
  3983 
  3984 * theory Main includes everything; INCOMPATIBILITY: theory ZF.thy contains
  3985   only the theorems proved on ZF.ML;
  3986 
  3987 * ZF INCOMPATIBILITY: rule `equals0D' is now a well-formed destruct rule;
  3988   It and 'sym RS equals0D' are now in the default  claset, giving automatic
  3989   disjointness reasoning but breaking a few old proofs.
  3990 
  3991 * ZF/Update: new theory of function updates
  3992     with default rewrite rule  f(x:=y) ` z = if(z=x, y, f`z)
  3993   may also be iterated as in f(a:=b,c:=d,...);
  3994 
  3995 * in  let x=t in u(x), neither t nor u(x) has to be an FOL term.
  3996 
  3997 * calling (stac rew i) now fails if "rew" has no effect on the goal
  3998   [previously, this check worked only if the rewrite rule was unconditional]
  3999   Now rew can involve either definitions or equalities (either == or =).
  4000 
  4001 * case_tac provided for compatibility with HOL
  4002     (like the old excluded_middle_tac, but with subgoals swapped)
  4003 
  4004 
  4005 *** Internal programming interfaces ***
  4006 
  4007 * Pure: several new basic modules made available for general use, see
  4008 also src/Pure/README;
  4009 
  4010 * improved the theory data mechanism to support encapsulation (data
  4011 kind name replaced by private Object.kind, acting as authorization
  4012 key); new type-safe user interface via functor TheoryDataFun; generic
  4013 print_data function becomes basically useless;
  4014 
  4015 * removed global_names compatibility flag -- all theory declarations
  4016 are qualified by default;
  4017 
  4018 * module Pure/Syntax now offers quote / antiquote translation
  4019 functions (useful for Hoare logic etc. with implicit dependencies);
  4020 see HOL/ex/Antiquote for an example use;
  4021 
  4022 * Simplifier now offers conversions (asm_)(full_)rewrite: simpset ->
  4023 cterm -> thm;
  4024 
  4025 * new tactical CHANGED_GOAL for checking that a tactic modifies a
  4026 subgoal;
  4027 
  4028 * Display.print_goals function moved to Locale.print_goals;
  4029 
  4030 * standard print function for goals supports current_goals_markers
  4031 variable for marking begin of proof, end of proof, start of goal; the
  4032 default is ("", "", ""); setting current_goals_markers := ("<proof>",
  4033 "</proof>", "<goal>") causes SGML like tagged proof state printing,
  4034 for example;
  4035 
  4036 
  4037 
  4038 New in Isabelle98 (January 1998)
  4039 --------------------------------
  4040 
  4041 *** Overview of INCOMPATIBILITIES (see below for more details) ***
  4042 
  4043 * changed lexical syntax of terms / types: dots made part of long
  4044 identifiers, e.g. "%x.x" no longer possible, should be "%x. x";
  4045 
  4046 * simpset (and claset) reference variable replaced by functions
  4047 simpset / simpset_ref;
  4048 
  4049 * no longer supports theory aliases (via merge) and non-trivial
  4050 implicit merge of thms' signatures;
  4051 
  4052 * most internal names of constants changed due to qualified names;
  4053 
  4054 * changed Pure/Sequence interface (see Pure/seq.ML);
  4055 
  4056 
  4057 *** General Changes ***
  4058 
  4059 * hierachically structured name spaces (for consts, types, axms, thms
  4060 etc.); new lexical class 'longid' (e.g. Foo.bar.x) may render much of
  4061 old input syntactically incorrect (e.g. "%x.x"); COMPATIBILITY:
  4062 isatool fixdots ensures space after dots (e.g. "%x. x"); set
  4063 long_names for fully qualified output names; NOTE: ML programs
  4064 (special tactics, packages etc.) referring to internal names may have
  4065 to be adapted to cope with fully qualified names; in case of severe
  4066 backward campatibility problems try setting 'global_names' at compile
  4067 time to have enrything declared within a flat name space; one may also
  4068 fine tune name declarations in theories via the 'global' and 'local'
  4069 section;
  4070 
  4071 * reimplemented the implicit simpset and claset using the new anytype
  4072 data filed in signatures; references simpset:simpset ref etc. are
  4073 replaced by functions simpset:unit->simpset and
  4074 simpset_ref:unit->simpset ref; COMPATIBILITY: use isatool fixclasimp
  4075 to patch your ML files accordingly;
  4076 
  4077 * HTML output now includes theory graph data for display with Java
  4078 applet or isatool browser; data generated automatically via isatool
  4079 usedir (see -i option, ISABELLE_USEDIR_OPTIONS);
  4080 
  4081 * defs may now be conditional; improved rewrite_goals_tac to handle
  4082 conditional equations;
  4083 
  4084 * defs now admits additional type arguments, using TYPE('a) syntax;
  4085 
  4086 * theory aliases via merge (e.g. M=A+B+C) no longer supported, always
  4087 creates a new theory node; implicit merge of thms' signatures is
  4088 restricted to 'trivial' ones; COMPATIBILITY: one may have to use
  4089 transfer:theory->thm->thm in (rare) cases;
  4090 
  4091 * improved handling of draft signatures / theories; draft thms (and
  4092 ctyps, cterms) are automatically promoted to real ones;
  4093 
  4094 * slightly changed interfaces for oracles: admit many per theory, named
  4095 (e.g. oracle foo = mlfun), additional name argument for invoke_oracle;
  4096 
  4097 * print_goals: optional output of const types (set show_consts and
  4098 show_types);
  4099 
  4100 * improved output of warnings (###) and errors (***);
  4101 
  4102 * subgoal_tac displays a warning if the new subgoal has type variables;
  4103 
  4104 * removed old README and Makefiles;
  4105 
  4106 * replaced print_goals_ref hook by print_current_goals_fn and result_error_fn;
  4107 
  4108 * removed obsolete init_pps and init_database;
  4109 
  4110 * deleted the obsolete tactical STATE, which was declared by
  4111     fun STATE tacfun st = tacfun st st;
  4112 
  4113 * cd and use now support path variables, e.g. $ISABELLE_HOME, or ~
  4114 (which abbreviates $HOME);
  4115 
  4116 * changed Pure/Sequence interface (see Pure/seq.ML); COMPATIBILITY:
  4117 use isatool fixseq to adapt your ML programs (this works for fully
  4118 qualified references to the Sequence structure only!);
  4119 
  4120 * use_thy no longer requires writable current directory; it always
  4121 reloads .ML *and* .thy file, if either one is out of date;
  4122 
  4123 
  4124 *** Classical Reasoner ***
  4125 
  4126 * Clarify_tac, clarify_tac, clarify_step_tac, Clarify_step_tac: new
  4127 tactics that use classical reasoning to simplify a subgoal without
  4128 splitting it into several subgoals;
  4129 
  4130 * Safe_tac: like safe_tac but uses the default claset;
  4131 
  4132 
  4133 *** Simplifier ***
  4134 
  4135 * added simplification meta rules:
  4136     (asm_)(full_)simplify: simpset -> thm -> thm;
  4137 
  4138 * simplifier.ML no longer part of Pure -- has to be loaded by object
  4139 logics (again);
  4140 
  4141 * added prems argument to simplification procedures;
  4142 
  4143 * HOL, FOL, ZF: added infix function `addsplits':
  4144   instead of `<simpset> setloop (split_tac <thms>)'
  4145   you can simply write `<simpset> addsplits <thms>'
  4146 
  4147 
  4148 *** Syntax ***
  4149 
  4150 * TYPE('a) syntax for type reflection terms;
  4151 
  4152 * no longer handles consts with name "" -- declare as 'syntax' instead;
  4153 
  4154 * pretty printer: changed order of mixfix annotation preference (again!);
  4155 
  4156 * Pure: fixed idt/idts vs. pttrn/pttrns syntactic categories;
  4157 
  4158 
  4159 *** HOL ***
  4160 
  4161 * HOL: there is a new splitter `split_asm_tac' that can be used e.g.
  4162   with `addloop' of the simplifier to faciliate case splitting in premises.
  4163 
  4164 * HOL/TLA: Stephan Merz's formalization of Lamport's Temporal Logic of Actions;
  4165 
  4166 * HOL/Auth: new protocol proofs including some for the Internet
  4167   protocol TLS;
  4168 
  4169 * HOL/Map: new theory of `maps' a la VDM;
  4170 
  4171 * HOL/simplifier: simplification procedures nat_cancel_sums for
  4172 cancelling out common nat summands from =, <, <= (in)equalities, or
  4173 differences; simplification procedures nat_cancel_factor for
  4174 cancelling common factor from =, <, <= (in)equalities over natural
  4175 sums; nat_cancel contains both kinds of procedures, it is installed by
  4176 default in Arith.thy -- this COULD MAKE EXISTING PROOFS FAIL;
  4177 
  4178 * HOL/simplifier: terms of the form
  4179   `? x. P1(x) & ... & Pn(x) & x=t & Q1(x) & ... Qn(x)'  (or t=x)
  4180   are rewritten to
  4181   `P1(t) & ... & Pn(t) & Q1(t) & ... Qn(t)',
  4182   and those of the form
  4183   `! x. P1(x) & ... & Pn(x) & x=t & Q1(x) & ... Qn(x) --> R(x)'  (or t=x)
  4184   are rewritten to
  4185   `P1(t) & ... & Pn(t) & Q1(t) & ... Qn(t) --> R(t)',
  4186 
  4187 * HOL/datatype
  4188   Each datatype `t' now comes with a theorem `split_t_case' of the form
  4189 
  4190   P(t_case f1 ... fn x) =
  4191      ( (!y1 ... ym1. x = C1 y1 ... ym1 --> P(f1 y1 ... ym1)) &
  4192         ...
  4193        (!y1 ... ymn. x = Cn y1 ... ymn --> P(f1 y1 ... ymn))
  4194      )
  4195 
  4196   and a theorem `split_t_case_asm' of the form
  4197 
  4198   P(t_case f1 ... fn x) =
  4199     ~( (? y1 ... ym1. x = C1 y1 ... ym1 & ~P(f1 y1 ... ym1)) |
  4200         ...
  4201        (? y1 ... ymn. x = Cn y1 ... ymn & ~P(f1 y1 ... ymn))
  4202      )
  4203   which can be added to a simpset via `addsplits'. The existing theorems
  4204   expand_list_case and expand_option_case have been renamed to
  4205   split_list_case and split_option_case.
  4206 
  4207 * HOL/Arithmetic:
  4208   - `pred n' is automatically converted to `n-1'.
  4209     Users are strongly encouraged not to use `pred' any longer,
  4210     because it will disappear altogether at some point.
  4211   - Users are strongly encouraged to write "0 < n" rather than
  4212     "n ~= 0". Theorems and proof tools have been modified towards this
  4213     `standard'.
  4214 
  4215 * HOL/Lists:
  4216   the function "set_of_list" has been renamed "set" (and its theorems too);
  4217   the function "nth" now takes its arguments in the reverse order and
  4218   has acquired the infix notation "!" as in "xs!n".
  4219 
  4220 * HOL/Set: UNIV is now a constant and is no longer translated to Compl{};
  4221 
  4222 * HOL/Set: The operator (UN x.B x) now abbreviates (UN x:UNIV. B x) and its
  4223   specialist theorems (like UN1_I) are gone.  Similarly for (INT x.B x);
  4224 
  4225 * HOL/record: extensible records with schematic structural subtyping
  4226 (single inheritance); EXPERIMENTAL version demonstrating the encoding,
  4227 still lacks various theorems and concrete record syntax;
  4228 
  4229 
  4230 *** HOLCF ***
  4231 
  4232 * removed "axioms" and "generated by" sections;
  4233 
  4234 * replaced "ops" section by extended "consts" section, which is capable of
  4235   handling the continuous function space "->" directly;
  4236 
  4237 * domain package:
  4238   . proves theorems immediately and stores them in the theory,
  4239   . creates hierachical name space,
  4240   . now uses normal mixfix annotations (instead of cinfix...),
  4241   . minor changes to some names and values (for consistency),
  4242   . e.g. cases -> casedist, dists_eq -> dist_eqs, [take_lemma] -> take_lemmas,
  4243   . separator between mutual domain defs: changed "," to "and",
  4244   . improved handling of sort constraints;  now they have to
  4245     appear on the left-hand side of the equations only;
  4246 
  4247 * fixed LAM <x,y,zs>.b syntax;
  4248 
  4249 * added extended adm_tac to simplifier in HOLCF -- can now discharge
  4250 adm (%x. P (t x)), where P is chainfinite and t continuous;
  4251 
  4252 
  4253 *** FOL and ZF ***
  4254 
  4255 * FOL: there is a new splitter `split_asm_tac' that can be used e.g.
  4256   with `addloop' of the simplifier to faciliate case splitting in premises.
  4257 
  4258 * qed_spec_mp, qed_goal_spec_mp, qed_goalw_spec_mp are available, as
  4259 in HOL, they strip ALL and --> from proved theorems;
  4260 
  4261 
  4262 
  4263 New in Isabelle94-8 (May 1997)
  4264 ------------------------------
  4265 
  4266 *** General Changes ***
  4267 
  4268 * new utilities to build / run / maintain Isabelle etc. (in parts
  4269 still somewhat experimental); old Makefiles etc. still functional;
  4270 
  4271 * new 'Isabelle System Manual';
  4272 
  4273 * INSTALL text, together with ./configure and ./build scripts;
  4274 
  4275 * reimplemented type inference for greater efficiency, better error
  4276 messages and clean internal interface;
  4277 
  4278 * prlim command for dealing with lots of subgoals (an easier way of
  4279 setting goals_limit);
  4280 
  4281 
  4282 *** Syntax ***
  4283 
  4284 * supports alternative (named) syntax tables (parser and pretty
  4285 printer); internal interface is provided by add_modesyntax(_i);
  4286 
  4287 * Pure, FOL, ZF, HOL, HOLCF now support symbolic input and output; to
  4288 be used in conjunction with the Isabelle symbol font; uses the
  4289 "symbols" syntax table;
  4290 
  4291 * added token_translation interface (may translate name tokens in
  4292 arbitrary ways, dependent on their type (free, bound, tfree, ...) and
  4293 the current print_mode); IMPORTANT: user print translation functions
  4294 are responsible for marking newly introduced bounds
  4295 (Syntax.mark_boundT);
  4296 
  4297 * token translations for modes "xterm" and "xterm_color" that display
  4298 names in bold, underline etc. or colors (which requires a color
  4299 version of xterm);
  4300 
  4301 * infixes may now be declared with names independent of their syntax;
  4302 
  4303 * added typed_print_translation (like print_translation, but may
  4304 access type of constant);
  4305 
  4306 
  4307 *** Classical Reasoner ***
  4308 
  4309 Blast_tac: a new tactic!  It is often more powerful than fast_tac, but has
  4310 some limitations.  Blast_tac...
  4311   + ignores addss, addbefore, addafter; this restriction is intrinsic
  4312   + ignores elimination rules that don't have the correct format
  4313         (the conclusion MUST be a formula variable)
  4314   + ignores types, which can make HOL proofs fail
  4315   + rules must not require higher-order unification, e.g. apply_type in ZF
  4316     [message "Function Var's argument not a bound variable" relates to this]
  4317   + its proof strategy is more general but can actually be slower
  4318 
  4319 * substitution with equality assumptions no longer permutes other
  4320 assumptions;
  4321 
  4322 * minor changes in semantics of addafter (now called addaltern); renamed
  4323 setwrapper to setWrapper and compwrapper to compWrapper; added safe wrapper
  4324 (and access functions for it);
  4325 
  4326 * improved combination of classical reasoner and simplifier:
  4327   + functions for handling clasimpsets
  4328   + improvement of addss: now the simplifier is called _after_ the
  4329     safe steps.
  4330   + safe variant of addss called addSss: uses safe simplifications
  4331     _during_ the safe steps. It is more complete as it allows multiple
  4332     instantiations of unknowns (e.g. with slow_tac).
  4333 
  4334 *** Simplifier ***
  4335 
  4336 * added interface for simplification procedures (functions that
  4337 produce *proven* rewrite rules on the fly, depending on current
  4338 redex);
  4339 
  4340 * ordering on terms as parameter (used for ordered rewriting);
  4341 
  4342 * new functions delcongs, deleqcongs, and Delcongs. richer rep_ss;
  4343 
  4344 * the solver is now split into a safe and an unsafe part.
  4345 This should be invisible for the normal user, except that the
  4346 functions setsolver and addsolver have been renamed to setSolver and
  4347 addSolver; added safe_asm_full_simp_tac;
  4348 
  4349 
  4350 *** HOL ***
  4351 
  4352 * a generic induction tactic `induct_tac' which works for all datatypes and
  4353 also for type `nat';
  4354 
  4355 * a generic case distinction tactic `exhaust_tac' which works for all
  4356 datatypes and also for type `nat';
  4357 
  4358 * each datatype comes with a function `size';
  4359 
  4360 * patterns in case expressions allow tuple patterns as arguments to
  4361 constructors, for example `case x of [] => ... | (x,y,z)#ps => ...';
  4362 
  4363 * primrec now also works with type nat;
  4364 
  4365 * recdef: a new declaration form, allows general recursive functions to be
  4366 defined in theory files.  See HOL/ex/Fib, HOL/ex/Primes, HOL/Subst/Unify.
  4367 
  4368 * the constant for negation has been renamed from "not" to "Not" to
  4369 harmonize with FOL, ZF, LK, etc.;
  4370 
  4371 * HOL/ex/LFilter theory of a corecursive "filter" functional for
  4372 infinite lists;
  4373 
  4374 * HOL/Modelcheck demonstrates invocation of model checker oracle;
  4375 
  4376 * HOL/ex/Ring.thy declares cring_simp, which solves equational
  4377 problems in commutative rings, using axiomatic type classes for + and *;
  4378 
  4379 * more examples in HOL/MiniML and HOL/Auth;
  4380 
  4381 * more default rewrite rules for quantifiers, union/intersection;
  4382 
  4383 * a new constant `arbitrary == @x.False';
  4384 
  4385 * HOLCF/IOA replaces old HOL/IOA;
  4386 
  4387 * HOLCF changes: derived all rules and arities
  4388   + axiomatic type classes instead of classes
  4389   + typedef instead of faking type definitions
  4390   + eliminated the internal constants less_fun, less_cfun, UU_fun, UU_cfun etc.
  4391   + new axclasses cpo, chfin, flat with flat < chfin < pcpo < cpo < po
  4392   + eliminated the types void, one, tr
  4393   + use unit lift and bool lift (with translations) instead of one and tr
  4394   + eliminated blift from Lift3.thy (use Def instead of blift)
  4395   all eliminated rules are derived as theorems --> no visible changes ;
  4396 
  4397 
  4398 *** ZF ***
  4399 
  4400 * ZF now has Fast_tac, Simp_tac and Auto_tac.  Union_iff is a now a default
  4401 rewrite rule; this may affect some proofs.  eq_cs is gone but can be put back
  4402 as ZF_cs addSIs [equalityI];
  4403 
  4404 
  4405 
  4406 New in Isabelle94-7 (November 96)
  4407 ---------------------------------
  4408 
  4409 * allowing negative levels (as offsets) in prlev and choplev;
  4410 
  4411 * super-linear speedup for large simplifications;
  4412 
  4413 * FOL, ZF and HOL now use miniscoping: rewriting pushes
  4414 quantifications in as far as possible (COULD MAKE EXISTING PROOFS
  4415 FAIL); can suppress it using the command Delsimps (ex_simps @
  4416 all_simps); De Morgan laws are also now included, by default;
  4417 
  4418 * improved printing of ==>  :  ~:
  4419 
  4420 * new object-logic "Sequents" adds linear logic, while replacing LK
  4421 and Modal (thanks to Sara Kalvala);
  4422 
  4423 * HOL/Auth: correctness proofs for authentication protocols;
  4424 
  4425 * HOL: new auto_tac combines rewriting and classical reasoning (many
  4426 examples on HOL/Auth);
  4427 
  4428 * HOL: new command AddIffs for declaring theorems of the form P=Q to
  4429 the rewriter and classical reasoner simultaneously;
  4430 
  4431 * function uresult no longer returns theorems in "standard" format;
  4432 regain previous version by: val uresult = standard o uresult;
  4433 
  4434 
  4435 
  4436 New in Isabelle94-6
  4437 -------------------
  4438 
  4439 * oracles -- these establish an interface between Isabelle and trusted
  4440 external reasoners, which may deliver results as theorems;
  4441 
  4442 * proof objects (in particular record all uses of oracles);
  4443 
  4444 * Simp_tac, Fast_tac, etc. that refer to implicit simpset / claset;
  4445 
  4446 * "constdefs" section in theory files;
  4447 
  4448 * "primrec" section (HOL) no longer requires names;
  4449 
  4450 * internal type "tactic" now simply "thm -> thm Sequence.seq";
  4451 
  4452 
  4453 
  4454 New in Isabelle94-5
  4455 -------------------
  4456 
  4457 * reduced space requirements;
  4458 
  4459 * automatic HTML generation from theories;
  4460 
  4461 * theory files no longer require "..." (quotes) around most types;
  4462 
  4463 * new examples, including two proofs of the Church-Rosser theorem;
  4464 
  4465 * non-curried (1994) version of HOL is no longer distributed;
  4466 
  4467 
  4468 
  4469 New in Isabelle94-4
  4470 -------------------
  4471 
  4472 * greatly reduced space requirements;
  4473 
  4474 * theory files (.thy) no longer require \...\ escapes at line breaks;
  4475 
  4476 * searchable theorem database (see the section "Retrieving theorems" on
  4477 page 8 of the Reference Manual);
  4478 
  4479 * new examples, including Grabczewski's monumental case study of the
  4480 Axiom of Choice;
  4481 
  4482 * The previous version of HOL renamed to Old_HOL;
  4483 
  4484 * The new version of HOL (previously called CHOL) uses a curried syntax
  4485 for functions.  Application looks like f a b instead of f(a,b);
  4486 
  4487 * Mutually recursive inductive definitions finally work in HOL;
  4488 
  4489 * In ZF, pattern-matching on tuples is now available in all abstractions and
  4490 translates to the operator "split";
  4491 
  4492 
  4493 
  4494 New in Isabelle94-3
  4495 -------------------
  4496 
  4497 * new infix operator, addss, allowing the classical reasoner to
  4498 perform simplification at each step of its search.  Example:
  4499         fast_tac (cs addss ss)
  4500 
  4501 * a new logic, CHOL, the same as HOL, but with a curried syntax
  4502 for functions.  Application looks like f a b instead of f(a,b).  Also pairs
  4503 look like (a,b) instead of <a,b>;
  4504 
  4505 * PLEASE NOTE: CHOL will eventually replace HOL!
  4506 
  4507 * In CHOL, pattern-matching on tuples is now available in all abstractions.
  4508 It translates to the operator "split".  A new theory of integers is available;
  4509 
  4510 * In ZF, integer numerals now denote two's-complement binary integers.
  4511 Arithmetic operations can be performed by rewriting.  See ZF/ex/Bin.ML;
  4512 
  4513 * Many new examples: I/O automata, Church-Rosser theorem, equivalents
  4514 of the Axiom of Choice;
  4515 
  4516 
  4517 
  4518 New in Isabelle94-2
  4519 -------------------
  4520 
  4521 * Significantly faster resolution;
  4522 
  4523 * the different sections in a .thy file can now be mixed and repeated
  4524 freely;
  4525 
  4526 * Database of theorems for FOL, HOL and ZF.  New
  4527 commands including qed, qed_goal and bind_thm store theorems in the database.
  4528 
  4529 * Simple database queries: return a named theorem (get_thm) or all theorems of
  4530 a given theory (thms_of), or find out what theory a theorem was proved in
  4531 (theory_of_thm);
  4532 
  4533 * Bugs fixed in the inductive definition and datatype packages;
  4534 
  4535 * The classical reasoner provides deepen_tac and depth_tac, making FOL_dup_cs
  4536 and HOL_dup_cs obsolete;
  4537 
  4538 * Syntactic ambiguities caused by the new treatment of syntax in Isabelle94-1
  4539 have been removed;
  4540 
  4541 * Simpler definition of function space in ZF;
  4542 
  4543 * new results about cardinal and ordinal arithmetic in ZF;
  4544 
  4545 * 'subtype' facility in HOL for introducing new types as subsets of existing
  4546 types;
  4547 
  4548 
  4549 $Id$