Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 1999-02-01 16:38:23 -0800 (Mon, 01 Feb 1999)
Revision: 2581
Log message:

      We do not have to ignore any *_verb* or *_simp* files anymore
      

Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 1999-02-01 17:54:59 -0800 (Mon, 01 Feb 1999)
Revision: 2582
Log message:

      - Now all user configuration options are in mk/config file
      - The old mk/config file (the one with all the make rules)
        is now called mk/rules, the new mk/config is _not_ on CVS
      - When mk/config.init is newer than mk/config (e.g. mk/config
        does not exist), mk/config.init is appended to mk/config
      - Toplevel Makefile checks if mk/config definies everything
        it should define and prints an error message if not
      
      - Fixed a bug in refiner/refsig/refine_error.h that prevented
        SIMPLE refiner from being built
      
      - I was not able to completely check if everything works because
        of the bug introduced by the previous check-in
      

Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 1999-02-01 18:05:03 -0800 (Mon, 01 Feb 1999)
Revision: 2583
Log message:

      Jason says we do not need -g gcc option anymore
      Also, I've decided it's better if make clean checks config file too
      

Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 1999-02-01 18:24:12 -0800 (Mon, 01 Feb 1999)
Revision: 2584
Log message:

      Fixed the native code profiling build - added back the filter goal
      

Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 1999-02-01 18:57:15 -0800 (Mon, 01 Feb 1999)
Revision: 2585
Log message:

      Ouch!  I accidentally removed PFILES from mk/rules, breaking the
      construction of prlc.  This should fix your "prlc.cmiz not found"
      problems.
      

Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 1999-02-04 19:44:12 -0800 (Thu, 04 Feb 1999)
Revision: 2586
Log message:

      There is no need for separate .mlz file
      

Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 1999-02-05 17:47:36 -0800 (Fri, 05 Feb 1999)
Revision: 2587
Log message:

      This module gives an ro-array-like interface based on splay trees
      It may be useful for sequent representation since it allows
      lazy operations on subtrees and more sharing than arrays
      
      I have not yet made the changes in the term modules to make use
      of this new module
      

Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 1999-02-09 18:44:05 -0800 (Tue, 09 Feb 1999)
Revision: 2589
Log message:

      Fixed some bugs, added some debugging code
      
      I am not going commit my Term module changes until I create
      the compile option to use usual ROArrays instead of LinearSets.
      
      From what I saw, LinearSets can be much slower.
      

Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 1999-02-12 15:50:37 -0800 (Fri, 12 Feb 1999)
Revision: 2590
Log message:

      Added the Linear_set implementation based on arrays (the same as was used
      before). Somehow it still made things a little bit faster, I have no idea why.
      
      Added the new compile-time option into mk/config file that allows to choose
      which Linear_set implementation to use. All relevant comments will be added
      to your mk/config file automatically when you ran make for the first time
      after "cvs update"
      

Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 1999-02-12 16:08:50 -0800 (Fri, 12 Feb 1999)
Revision: 2591
Log message:

      Rebuild Makefile.dep when something important changes
      

Changes by: Yegor N. Bryukhov (ybryukhov at gc.cuny.edu)
Date: 1999-02-15 00:09:13 -0800 (Mon, 15 Feb 1999)
Revision: 2592
Log message:

      Local _ 'a option _ type in Simplehashtbl replaced with OCAML _ 'a option _ type
      

Changes by: Yegor N. Bryukhov (ybryukhov at gc.cuny.edu)
Date: 1999-02-16 00:27:19 -0800 (Tue, 16 Feb 1999)
Revision: 2593
Log message:

      Term_copy_weak, Term_copy2_weak, Term_norm - new modules for term coping and
      term normalization. Them faster than old ones: O(n) instead of O(n*n).
      These modules were tested on terms supplied to Term_copy during Metaprl
      compilation and now difference between old Term_copy and Term_copy_weak found.
      

Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 1999-02-16 08:53:19 -0800 (Tue, 16 Feb 1999)
Revision: 2594
Log message:

      Added the new weak term_copy files into the build (they are being built,
      but not being used yet).
      

Changes by: Eli Barzilay (eli at cs.cornell.edu)
Date: 1999-02-17 08:34:24 -0800 (Wed, 17 Feb 1999)
Revision: 2595
Log message:

      Re-added compilation of all fol files.
      

Changes by: Eli Barzilay (eli at cs.cornell.edu)
Date: 1999-02-17 14:25:11 -0800 (Wed, 17 Feb 1999)
Revision: 2596
Log message:

      The OCaml preprocessor.  Compiled to byte code for now.
      

Changes by: Yegor N. Bryukhov (ybryukhov at gc.cuny.edu)
Date: 1999-02-23 04:38:58 -0800 (Tue, 23 Feb 1999)
Revision: 2597
Log message:

      TermHash become a parameter for TermHeaderConstr, TermNorm,
      TermCopyWeak and TermCopy2Weak.
      
      It was made to support global hashing structures for each TermModule.
      

Changes by: Yegor N. Bryukhov (ybryukhov at gc.cuny.edu)
Date: 1999-02-23 04:44:23 -0800 (Tue, 23 Feb 1999)
Revision: 2598
Log message:

      These files were intermediate version of TermCopy between old one
      and TermCopyWeak. So there is no necessity in it now.
      

Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 1999-02-23 18:08:29 -0800 (Tue, 23 Feb 1999)
Revision: 2599
Log message:

      Added profile_mem goal that allows one to profile memory allocation
      

Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 1999-02-24 15:26:23 -0800 (Wed, 24 Feb 1999)
Revision: 2600
Log message:

      Added print_gc_stats: unit -> unit command to the toploop.
      

Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 1999-02-25 09:59:51 -0800 (Thu, 25 Feb 1999)
Revision: 2601
Log message:

      1) Functorized the Term_copy[2]_weak/Term_hash/Termm_header/etc modules
      correctly, now it should be easy make the Term_hash module
      a part of the refiner
      
      2) Created a new Term_io module that provides Refiner_std <-> Refiner
      convertion based on the term_copy2_weak, these functions are now used
      instead of the old Term_copy functions
      
      3) Removed old Term_copy - related modules
      

Changes by: Aleksey Nogin (nogin at cs.caltech.edu)
Date: 1999-02-25 16:37:49 -0800 (Thu, 25 Feb 1999)
Revision: 2602
Log message:

      Made TermHeader, TermHash and TermNorm a part of the term module