]> matita.cs.unibo.it Git - helm.git/log
helm.git
18 years agoSerious bug fixed: without a Lazy.force the user obtained an error comparing
Enrico Zoli [Fri, 20 Oct 2006 08:35:00 +0000 (08:35 +0000)]
Serious bug fixed: without a Lazy.force the user obtained an error comparing
two functional values.

18 years agoDisambiguation errors are now compressed in a maybe better way.
Claudio Sacerdoti Coen [Thu, 19 Oct 2006 11:09:06 +0000 (11:09 +0000)]
Disambiguation errors are now compressed in a maybe better way.

18 years agoBug fixed: when trying to insert coercions after an Uncertain, a RefineFailure
Claudio Sacerdoti Coen [Thu, 19 Oct 2006 09:16:07 +0000 (09:16 +0000)]
Bug fixed: when trying to insert coercions after an Uncertain, a RefineFailure
was raised (instead of re-raising an Uncertain).

18 years agoPotential performance improvement + better disambiguation error messages:
Claudio Sacerdoti Coen [Thu, 19 Oct 2006 08:50:31 +0000 (08:50 +0000)]
Potential performance improvement + better disambiguation error messages:
the delift functions used to raise Uncertain every time it failed; now it
raises Failure when the local context contains no Metas (that means that
for every future subst it will not contain new Rels).
Note: if we are working up to conversion (I do not think so), we should check
that the local context is Meta-closed, that is a weaker check.

18 years agoMissing optimization implemented: before starting to analyze the disambiguation
Claudio Sacerdoti Coen [Wed, 18 Oct 2006 17:39:48 +0000 (17:39 +0000)]
Missing optimization implemented: before starting to analyze the disambiguation
domain we try first without interpreting any symbol. This way we can avoid
lot of refinements when the information in the uninterpreted term already
tells us that every instance will not be well typed.

18 years ago- Disambiguation error exception enriched with more information
Claudio Sacerdoti Coen [Wed, 18 Oct 2006 15:06:03 +0000 (15:06 +0000)]
- Disambiguation error exception enriched with more information
  (the same returned in case of correct disambiguation)
- The disambiguation error interface now shows the complete environment
  in which the error occurs. I am not sure if this is better or worst than
  showing only the diffs.

18 years agoBug fixed: the diff component of the exception raised when the term cannot
Claudio Sacerdoti Coen [Wed, 18 Oct 2006 14:13:25 +0000 (14:13 +0000)]
Bug fixed: the diff component of the exception raised when the term cannot
be disambiguated used to lack the last choice in case of the look-ahead
optimization.

18 years agoEXPERIMENTAL: new interface for disambiguation errors.
Claudio Sacerdoti Coen [Wed, 18 Oct 2006 13:18:37 +0000 (13:18 +0000)]
EXPERIMENTAL: new interface for disambiguation errors.

[ Moreover, a dead dialog window has been removed ]

18 years agoDead dialog window removed.
Claudio Sacerdoti Coen [Wed, 18 Oct 2006 13:17:10 +0000 (13:17 +0000)]
Dead dialog window removed.

18 years agoDead dialog window removed.
Claudio Sacerdoti Coen [Wed, 18 Oct 2006 13:16:06 +0000 (13:16 +0000)]
Dead dialog window removed.

18 years agoDisambiguation errors now carry more information (i.e. the patches to apply to
Claudio Sacerdoti Coen [Wed, 18 Oct 2006 12:11:48 +0000 (12:11 +0000)]
Disambiguation errors now carry more information (i.e. the patches to apply to
the current interpretation to re-obtain the refinement errors).

18 years agoDama is now in the night benchmarks.
Claudio Sacerdoti Coen [Wed, 18 Oct 2006 12:10:00 +0000 (12:10 +0000)]
Dama is now in the night benchmarks.

18 years agoToo verbose error message (probably activated by Enrico without really wanting
Claudio Sacerdoti Coen [Wed, 18 Oct 2006 09:42:40 +0000 (09:42 +0000)]
Too verbose error message (probably activated by Enrico without really wanting
it) removed.

18 years agonew objects for the LambdaDelta development (4th conjecture proved)
Ferruccio Guidi [Tue, 17 Oct 2006 20:45:12 +0000 (20:45 +0000)]
new objects for the LambdaDelta development (4th conjecture proved)

18 years agomore new objects for the LambdaDelta contribution
Ferruccio Guidi [Tue, 17 Oct 2006 10:52:58 +0000 (10:52 +0000)]
more new objects for the LambdaDelta contribution

18 years agoBeginning of the development of integration algebras.
Enrico Zoli [Mon, 16 Oct 2006 16:49:51 +0000 (16:49 +0000)]
Beginning of the development of integration algebras.

18 years agoContent level representation of LetRec changed.
Claudio Sacerdoti Coen [Fri, 13 Oct 2006 16:59:04 +0000 (16:59 +0000)]
Content level representation of LetRec changed.

18 years agoContent level representation of LetRec changed.
Claudio Sacerdoti Coen [Fri, 13 Oct 2006 16:57:10 +0000 (16:57 +0000)]
Content level representation of LetRec changed.

18 years agoNew content level representations for LetRec, Inductive and CoInductive.
Claudio Sacerdoti Coen [Fri, 13 Oct 2006 16:53:54 +0000 (16:53 +0000)]
New content level representations for LetRec, Inductive and CoInductive.
For LetRec the pretty-printer is now in sync with the parser.

18 years agofiles with newest objects (to be included in the respective developments)
Ferruccio Guidi [Thu, 12 Oct 2006 21:14:23 +0000 (21:14 +0000)]
files with newest objects (to be included in the respective developments)

18 years agocarr_of_term now returns Fun if a Prod is encountered
Enrico Tassi [Thu, 12 Oct 2006 16:10:53 +0000 (16:10 +0000)]
carr_of_term now returns Fun if a Prod is encountered

18 years agofixed defaultauto behaviour. not the cache is preserveed
Enrico Tassi [Thu, 12 Oct 2006 16:10:23 +0000 (16:10 +0000)]
fixed defaultauto behaviour. not the cache is preserveed

18 years agotimeout if unspecfied should be set to infinity, not 0, since the timeout inside...
Enrico Tassi [Thu, 12 Oct 2006 13:34:32 +0000 (13:34 +0000)]
timeout if unspecfied should be set to infinity, not 0, since the timeout inside the flagfs structure is a time in the future, not a decreasing quantity.
every loop of auto the check gettimeofday() > timeout raise Fail

18 years agoThe default for paramodulation is now back to false (I set it to true
Claudio Sacerdoti Coen [Thu, 12 Oct 2006 10:37:17 +0000 (10:37 +0000)]
The default for paramodulation is now back to false (I set it to true
by mistake).

18 years agoauto => auto new.
acciavat [Thu, 12 Oct 2006 09:53:08 +0000 (09:53 +0000)]
auto => auto new.
There is still one auto left that does not work with the new implementation
(because of implicative hypothesis for local lemmas?)

18 years agoInclusion "improved".
Claudio Sacerdoti Coen [Thu, 12 Oct 2006 09:46:37 +0000 (09:46 +0000)]
Inclusion "improved".

18 years agoCoRN integrated in the night benchmarks.
Claudio Sacerdoti Coen [Thu, 12 Oct 2006 09:44:03 +0000 (09:44 +0000)]
CoRN integrated in the night benchmarks.

18 years agoManual porting of CoRN to Matita.
acciavat [Thu, 12 Oct 2006 09:39:45 +0000 (09:39 +0000)]
Manual porting of CoRN to Matita.
The current version shows a bug in coercions to function classes.

18 years agoBug fixed: the conversion was done with the wront argument (a sort of typo).
Claudio Sacerdoti Coen [Thu, 12 Oct 2006 08:06:48 +0000 (08:06 +0000)]
Bug fixed: the conversion was done with the wront argument (a sort of typo).

18 years agoSome hocus-pocus to avoid a common race condition (Gtk/??? code not
Claudio Sacerdoti Coen [Wed, 11 Oct 2006 14:58:23 +0000 (14:58 +0000)]
Some hocus-pocus to avoid a common race condition (Gtk/??? code not
reentrant). I bet the "avoided" problem will surface again...

18 years agoUnlocking the interface was not performed as the last action of the callback.
Claudio Sacerdoti Coen [Wed, 11 Oct 2006 14:26:59 +0000 (14:26 +0000)]
Unlocking the interface was not performed as the last action of the callback.

18 years agoThe RELATIONAL contrib must be compiled before the others since the others
Claudio Sacerdoti Coen [Wed, 11 Oct 2006 12:25:27 +0000 (12:25 +0000)]
The RELATIONAL contrib must be compiled before the others since the others
depend on it.

18 years agoMy previous commit changed the regular timeout of paramodulation from
Claudio Sacerdoti Coen [Wed, 11 Oct 2006 09:49:27 +0000 (09:49 +0000)]
My previous commit changed the regular timeout of paramodulation from
infinity to 30s. Old behaviour restored.

18 years agomakefiles fixups
Ferruccio Guidi [Tue, 10 Oct 2006 19:51:39 +0000 (19:51 +0000)]
makefiles fixups

18 years agofguidi removed from RT in makefiles
Claudio Sacerdoti Coen [Tue, 10 Oct 2006 14:06:44 +0000 (14:06 +0000)]
fguidi removed from RT in makefiles

18 years agoImplemented:
Claudio Sacerdoti Coen [Tue, 10 Oct 2006 14:03:40 +0000 (14:03 +0000)]
Implemented:
 "by ... we proved ... that is equivalent to ... done"
and
 "we need to prove ... that is equivalent to ..."

18 years agoauto => auto new
Claudio Sacerdoti Coen [Tue, 10 Oct 2006 13:12:44 +0000 (13:12 +0000)]
auto => auto new

18 years agoauto => auto new
Claudio Sacerdoti Coen [Tue, 10 Oct 2006 13:10:08 +0000 (13:10 +0000)]
auto => auto new
added depth=4 to some file in order to make it pass

18 years agoI do not understand at all why Enrico removed the contribs from the bench!
Claudio Sacerdoti Coen [Tue, 10 Oct 2006 13:09:35 +0000 (13:09 +0000)]
I do not understand at all why Enrico removed the contribs from the bench!

18 years agoSorry, bug introduced by me yesterday now fixed.
Claudio Sacerdoti Coen [Tue, 10 Oct 2006 08:00:32 +0000 (08:00 +0000)]
Sorry, bug introduced by me yesterday now fixed.
auto should be the new one even when the paramodulation flag is given.

18 years agoBugs fixed in merging of composite coercions. In particular the imperative
Claudio Sacerdoti Coen [Mon, 9 Oct 2006 18:26:05 +0000 (18:26 +0000)]
Bugs fixed in merging of composite coercions. In particular the imperative
status was not handled properly and letins were handled inefficiently.

18 years agoOne auto modified in an apply since auto is no longer supposed to apply
Claudio Sacerdoti Coen [Mon, 9 Oct 2006 18:23:49 +0000 (18:23 +0000)]
One auto modified in an apply since auto is no longer supposed to apply
elimination principles automatically.

18 years agoadded to applyS in nat/gcd.ma a timeout large enough for compilation in bytecode
Claudio Sacerdoti Coen [Mon, 9 Oct 2006 18:18:24 +0000 (18:18 +0000)]
added to applyS in nat/gcd.ma a timeout large enough for compilation in bytecode
mode.

18 years ago1. applyS now uses its ~params
Claudio Sacerdoti Coen [Mon, 9 Oct 2006 18:17:50 +0000 (18:17 +0000)]
1. applyS now uses its ~params
2. autoTactic.ml* is going to be deprecated.
   auto.ml* has now the expected interface (that of a tactic)
   and it is almost ready to replace autoTactic.ml

18 years agoapplyS now receives the same parameters that auto receives.
Claudio Sacerdoti Coen [Mon, 9 Oct 2006 17:48:45 +0000 (17:48 +0000)]
applyS now receives the same parameters that auto receives.
The parameters are not used yet.

18 years agoTheorems from the library and from the context are indexed also in "unfolded"
Andrea Asperti [Mon, 9 Oct 2006 16:33:24 +0000 (16:33 +0000)]
Theorems from the library and from the context are indexed also in "unfolded"
version.

18 years agoauto => auto new and other minor changes to make it compile.
Claudio Sacerdoti Coen [Mon, 9 Oct 2006 16:28:22 +0000 (16:28 +0000)]
auto => auto new and other minor changes to make it compile.

18 years agoauto => auto new everywhere + minor updates to make more tests pass
Claudio Sacerdoti Coen [Mon, 9 Oct 2006 15:33:47 +0000 (15:33 +0000)]
auto => auto new everywhere + minor updates to make more tests pass

18 years agoComments updated with new reflections.
Claudio Sacerdoti Coen [Mon, 9 Oct 2006 14:04:31 +0000 (14:04 +0000)]
Comments updated with new reflections.

18 years agoThe two coercions sym_eq e eq_f gives BIG TROUBLES when checking composition during
Andrea Asperti [Mon, 9 Oct 2006 11:00:31 +0000 (11:00 +0000)]
The two coercions sym_eq e eq_f gives BIG TROUBLES when checking composition during
qed. In particular, checking composition calls unification, that calls reductions,
that is problematic when there are a lot of lein in the context (as is it the case
with proofs generated by TPTP: see eg GRP486-1.p.ma )

18 years agoMore work to handle -debug properly.
Claudio Sacerdoti Coen [Mon, 9 Oct 2006 09:23:09 +0000 (09:23 +0000)]
More work to handle -debug properly.

18 years agoFactorized "find_equalities" in demodulation_tac.
Andrea Asperti [Mon, 9 Oct 2006 06:53:04 +0000 (06:53 +0000)]
Factorized "find_equalities" in demodulation_tac.

18 years agoadded support for short name targets
Enrico Tassi [Fri, 6 Oct 2006 16:49:16 +0000 (16:49 +0000)]
added support for short name targets

18 years agoresumed ol auto
Enrico Tassi [Fri, 6 Oct 2006 16:41:53 +0000 (16:41 +0000)]
resumed ol auto

18 years agofixed all (that now uses long paths)
Enrico Tassi [Fri, 6 Oct 2006 15:56:52 +0000 (15:56 +0000)]
fixed all (that now uses long paths)

18 years agonow the makefile for developments requires the depend file
Enrico Tassi [Fri, 6 Oct 2006 15:30:37 +0000 (15:30 +0000)]
now the makefile for developments requires the depend file
that now has correct dependencies.

18 years ago1. some "try ... with _ " removed
Claudio Sacerdoti Coen [Fri, 6 Oct 2006 07:20:13 +0000 (07:20 +0000)]
1. some "try ... with _ " removed
2. type inference of LetIn terms is now closer than the kernel
3. LetIn code fixed to add the type to the Defs in the environment
4. some typos fixed

18 years agoreduced timeout to 100s
Enrico Tassi [Tue, 3 Oct 2006 16:40:35 +0000 (16:40 +0000)]
reduced timeout to 100s

18 years agocommented out are_convertible in is_identity
Enrico Tassi [Tue, 3 Oct 2006 16:08:31 +0000 (16:08 +0000)]
commented out are_convertible in is_identity
fixed metadataquery so that only are calculated without the DB (but are calculated)
removed is_meta_convertible left right in is_weak_identity
added extra dependencies between gcd < relevant_equations < ord to make
paramodulation happy. In the past this used to be the order used by make.
added an hack in applys.ma, since it used to work due to convertibility.

18 years agoSome declarative tactics did not allow the "done" option in place
maiorino [Tue, 3 Oct 2006 16:05:50 +0000 (16:05 +0000)]
Some declarative tactics did not allow the "done" option in place
of "(id)". Fixed.

18 years agoInline command implemented.
Claudio Sacerdoti Coen [Tue, 3 Oct 2006 15:59:34 +0000 (15:59 +0000)]
Inline command implemented.

18 years agoSyntax of paramodulation parameters changed.
Claudio Sacerdoti Coen [Tue, 3 Oct 2006 15:58:50 +0000 (15:58 +0000)]
Syntax of paramodulation parameters changed.

18 years agoFixed handling of exceptions by the worker threads: the exception is not
Claudio Sacerdoti Coen [Tue, 3 Oct 2006 14:18:34 +0000 (14:18 +0000)]
Fixed handling of exceptions by the worker threads: the exception is not
notified to the user and the thread exits gracefully.

18 years agoRemoved /home/tassi from the makefile!
Claudio Sacerdoti Coen [Tue, 3 Oct 2006 13:57:30 +0000 (13:57 +0000)]
Removed /home/tassi from the makefile!

18 years agoNew home page, with italian version.
Andrea Asperti [Tue, 3 Oct 2006 13:39:27 +0000 (13:39 +0000)]
New home page, with italian version.

18 years agoQuery fixed to handle the cases where 0 tests are failures or successfull.
Enrico Tassi [Tue, 3 Oct 2006 13:00:43 +0000 (13:00 +0000)]
Query fixed to handle the cases where 0 tests are failures or successfull.

18 years agoChanged auto from implicit to option and renamed a few functions.
Andrea Asperti [Tue, 3 Oct 2006 12:56:38 +0000 (12:56 +0000)]
Changed auto from implicit to option and renamed a few functions.

18 years agoRegular expression fixed to allow '-' into test names.
Claudio Sacerdoti Coen [Tue, 3 Oct 2006 12:15:00 +0000 (12:15 +0000)]
Regular expression fixed to allow '-' into test names.

18 years agosacerdot is now the user for the daily bench.
Claudio Sacerdoti Coen [Tue, 3 Oct 2006 12:02:55 +0000 (12:02 +0000)]
sacerdot is now the user for the daily bench.

18 years agoCrontab updated
Claudio Sacerdoti Coen [Tue, 3 Oct 2006 11:59:30 +0000 (11:59 +0000)]
Crontab updated

18 years agoNo more benches without gc.
Claudio Sacerdoti Coen [Tue, 3 Oct 2006 11:49:19 +0000 (11:49 +0000)]
No more benches without gc.

18 years agosequent_viewer.xml & co are now generated in /tmp if necessary!
Claudio Sacerdoti Coen [Tue, 3 Oct 2006 11:33:14 +0000 (11:33 +0000)]
sequent_viewer.xml & co are now generated in /tmp if necessary!

18 years agoupdated to use destruct instead of disciminate/injection
Ferruccio Guidi [Tue, 3 Oct 2006 11:18:21 +0000 (11:18 +0000)]
updated to use destruct instead of disciminate/injection

18 years agoBug fixed: the script Menu was not locked as expected.
Claudio Sacerdoti Coen [Tue, 3 Oct 2006 08:31:25 +0000 (08:31 +0000)]
Bug fixed: the script Menu was not locked as expected.

18 years agoThe Abort button is now working properly. Cool.
Claudio Sacerdoti Coen [Tue, 3 Oct 2006 07:52:16 +0000 (07:52 +0000)]
The Abort button is now working properly. Cool.

18 years agoXavier code for killing a thread commented out for a while since it interacts badly...
Claudio Sacerdoti Coen [Mon, 2 Oct 2006 17:55:36 +0000 (17:55 +0000)]
Xavier code for killing a thread commented out for a while since it interacts badly with (labl?)Gtk.

18 years agoThis commit implements the Abort button for the GUI using a clever trick by Xavier...
Claudio Sacerdoti Coen [Mon, 2 Oct 2006 17:44:01 +0000 (17:44 +0000)]
This commit implements the Abort button for the GUI using a clever trick by Xavier Leroy.

18 years agorestored is_identity instead of is_weak_identity
Enrico Tassi [Mon, 2 Oct 2006 16:30:28 +0000 (16:30 +0000)]
restored is_identity instead of is_weak_identity

18 years ago**** EXPERIMENTAL ****
Claudio Sacerdoti Coen [Mon, 2 Oct 2006 16:22:13 +0000 (16:22 +0000)]
**** EXPERIMENTAL ****
The GUI of Matita is now thread-based.
In principle Ctr-C works unreliably.
In practice it works reasonably well.

18 years agoadded a subtle List.rev that makes the order of the equalities taken from the library...
Enrico Tassi [Mon, 2 Oct 2006 15:33:24 +0000 (15:33 +0000)]
added a subtle List.rev that makes the order of the equalities taken from the library as it used to be. In this way the applys test works, but is not clear why the other order does not produce a solution in decent time

18 years ago...
Enrico Tassi [Mon, 2 Oct 2006 15:30:44 +0000 (15:30 +0000)]
...

18 years agoauto always uses the context (even if paramodulation) to try to close goals left...
Enrico Tassi [Mon, 2 Oct 2006 15:19:25 +0000 (15:19 +0000)]
auto always uses the context (even if paramodulation) to try to close goals left open by paramod

18 years agoreturns the right list of goals
Enrico Tassi [Mon, 2 Oct 2006 15:18:35 +0000 (15:18 +0000)]
returns the right list of goals

18 years ago50 steps on goal are fine for irrat2
Enrico Tassi [Mon, 2 Oct 2006 15:17:01 +0000 (15:17 +0000)]
50 steps on goal are fine for irrat2

18 years agoremoved only made with the DB
Enrico Tassi [Mon, 2 Oct 2006 15:16:02 +0000 (15:16 +0000)]
removed only made with the DB

18 years agoremoved a pointless call to auto
Enrico Tassi [Mon, 2 Oct 2006 15:15:31 +0000 (15:15 +0000)]
removed a pointless call to auto

18 years ago...
Enrico Tassi [Mon, 2 Oct 2006 15:15:03 +0000 (15:15 +0000)]
...

18 years agoadded tests for paramod
Enrico Tassi [Mon, 2 Oct 2006 15:14:46 +0000 (15:14 +0000)]
added tests for paramod

18 years ago...
Enrico Tassi [Mon, 2 Oct 2006 13:56:53 +0000 (13:56 +0000)]
...

18 years agoadded missing *)
Enrico Tassi [Mon, 2 Oct 2006 11:42:43 +0000 (11:42 +0000)]
added missing *)

18 years agorestored old (r6662) behaviour
Enrico Tassi [Mon, 2 Oct 2006 11:29:49 +0000 (11:29 +0000)]
restored old (r6662) behaviour

18 years agowhen use_only_paramod is true do not calculate the universe for applicative reasoning
Enrico Tassi [Sun, 1 Oct 2006 13:02:29 +0000 (13:02 +0000)]
when use_only_paramod is true do not calculate the universe for applicative reasoning

18 years ago2: is_identity -> is_weak_identity
Enrico Tassi [Fri, 29 Sep 2006 17:08:10 +0000 (17:08 +0000)]
2: is_identity -> is_weak_identity

18 years agofixed metaconvertibility that was completely wrong.
Enrico Tassi [Fri, 29 Sep 2006 17:07:53 +0000 (17:07 +0000)]
fixed metaconvertibility that was completely wrong.
is_identity and is_weak_identity fixed accordingly

18 years agoremoved bad guard that was always false (assert false in the line above)
Enrico Tassi [Fri, 29 Sep 2006 17:07:06 +0000 (17:07 +0000)]
removed bad guard that was always false (assert false in the line above)

18 years agofixed (dasabled paramod)
Enrico Tassi [Fri, 29 Sep 2006 14:49:31 +0000 (14:49 +0000)]
fixed (dasabled paramod)

18 years agorenamed inference in founif that is more appropriate
Enrico Tassi [Fri, 29 Sep 2006 14:41:27 +0000 (14:41 +0000)]
renamed inference in founif that is more appropriate

18 years agois_identity -> is_weak_identity
Enrico Tassi [Fri, 29 Sep 2006 14:27:20 +0000 (14:27 +0000)]
is_identity -> is_weak_identity

18 years agohack to make the Pp work (sometimes)
Enrico Tassi [Fri, 29 Sep 2006 14:22:25 +0000 (14:22 +0000)]
hack to make the Pp work (sometimes)

18 years agonew version of auto that is able to prove the irrationality of sqrt(2)
Enrico Tassi [Fri, 29 Sep 2006 12:49:11 +0000 (12:49 +0000)]
new version of auto that is able to prove the irrationality of sqrt(2)