2000-11-30  Martin Norbck  <d95mback@dtek.chalmers.se>

	* gnome-genius.desktop: added Swedish translation

2000-07-29  Hilaire Fernandes  <hilaire@ofset.org>

	* genius.c: Set the path to the real file (ie ncurse/).

2000-05-04  Valek Filippov  <frob@df.ru>

	* eval.c: typo fixed.

Tue Feb 08 21:25:07 2000  George Lebl <jirka@5z.com>

	* calc.[ch],genius.c,lexer.h,lexer.l,inter.[ch],funclib.c: use flex
	  buffers to handle nested files correctly and setup the correct
	  directory by passing directory prefixes to the parsing functions.

	* gnome-genius.c: small fixes

Tue Feb 08 20:01:34 2000  George Lebl <jirka@5z.com>

	* calc.c: some debugging output and a first (broken) pass at
	  changing directories correctly to make loading work correctly

Mon Feb 07 21:13:19 2000  George Lebl <jirka@5z.com>

	* eval.c: when expanding matricies make a private copy so that
	  we don't mangle constant matricies in the function library upon
	  execution

Fri Feb 04 17:44:16 2000  George Lebl <jirka@5z.com>

	* mpwrap.[ch],funclib.c: added numerator and denominator functions

Fri Feb 04 00:57:38 2000  George Lebl <jirka@5z.com>

	* geloutput.[ch],calc.[ch],eval.c,funclib.c,util.c,genius.c,
	  gnome-genius.c: Start using the new output stuff with somewhat
	  large hacks to make it work, but a step closer towards the new
	  interface.

Thu Feb  3 13:34:49 2000  George Lebl <jirka@5z.com>

	* geloutput.c: Finished implementing of the output stuff, now also
	  does tab expansion and hhandles the long line limiting mostly
	  correctly.

	* struct.h,geloutput.[ch]: use set/get_data style data on the output
	  as well and add dynamic line length getting function

Thu Feb 03 00:16:35 2000  George Lebl <jirka@5z.com>

	* matop.c,eval.c: Fixed very critical copy-on-write problems which
	  caused mangeled matrix variables

	* matop.[ch],funclib.c,eval.c: make the gauss function have a bit
	  nicer interface

Wed Feb 02 23:24:37 2000  George Lebl <jirka@5z.com>

	* gel.[ch],structs.h: fix some warnings and add accessors for
	  errorout and infoout

Wed Feb 02 23:19:16 2000  George Lebl  <jirka@5z.com>

	* gel.[ch],structs.h,geloutput.[ch]: implement a bunch of output
	  stuff

Tue Feb  1 13:38:48 2000  George Lebl <jirka@5z.com>

	* gel.[ch], mpwrap.[ch]: further implementation details are
	  being sorted out

Tue Jan 25 22:58:24 2000  George Lebl <jirka@5z.com>

	* mpwrap.[ch]: doh! fixed compile errors

Tue Jan 25 13:22:54 2000  George Lebl <jirka@5z.com>

	* mpwrap.[ch]: figure out some new interface to make this part
	  embeddable and modular

Wed Jan 19 01:52:56 2000  George Lebl <jirka@5z.com>

	* structs.h,gelcalc.[ch],geloutput.[ch]: just a tad of work to start
	  converting the actual calculator interface, the files are just
	  placeholders now

Mon Jan 17 22:59:17 2000  George Lebl <jirka@5z.com>

	* Makefile.am,gel.c,geldict.[ch]: fixup so that it actually
	  compiles and links, it's still not used however

Mon Jan 17 19:29:44 2000  George Lebl <jirka@5z.com>

	* gel.c,geldict.[ch]: implemented the full dictionary functionality

Mon Jan 17 01:10:53 2000  George Lebl <jirka@5z.com>

	* geldict.c: implement the global add and the global reference

Tue Jan 11 18:41:56 2000  George Lebl <jirka@5z.com>

	* geldict.[ch]: more dictionary implementation is being filled in

Mon Jan 10 20:35:40 2000  George Lebl <jirka@5z.com>

	* structs.h,calc.c,eval.[ch],compil.c,matrixw.c: fixup more namespace
	  stuff

Mon Jan 10 19:01:32 2000  George Lebl <jirka@5z.com>

	* structs.h,gel.[ch],geldict.[ch]: started implemented the actual
	  dictionary stuff.

Sun Jan 09 18:30:41 2000  George Lebl <jirka@5z.com>

	* gel.[ch],geldict.[ch],structs.h: a start on the next generation
	  calculator and dictionary interface to allow embedding and
	  modularization

Sun Jan 09 17:18:58 2000  George Lebl <jirka@5z.com>

	* mpwrap.c: sanity check of integer_output_base for the int output

	* calc.h: use gboolean instead of int for booleans

	* gnome-genius.c: correctly initialize the state structure

	* extra.h: never use the __weak__ attribute, for some reason this
	  causes weird segfaults in dr-genius (not genius nor gnome-genius
	  however)

Fri Jan 07 01:53:53 2000  George Lebl <jirka@5z.com>

	* structs.h: begin thinking about the new interface and start
	  on the structures

	* *.[chly]: to make transition to a lib easier start prepending
	  functions with gel_ and declaring some as weak symbols

Thu Jan 06 16:29:43 2000  George Lebl <jirka@5z.com>

	* mpwrap.c,geniustests.txt: fixup integer output in bases higher
	  then 10 (except 16) for small integers

Tue Nov 30 19:03:56 1999  George Lebl <jirka@5z.com>

	* eval.c: don't precalc MOD_CALC body

	* eval.c,structs.h,utype.[ch],Makefile.am: start on a user type,
	  but this is not yet finished, it only compiles now

Tue Nov 30 17:49:32 1999  George Lebl <jirka@5z.com>

	* eval.[ch],calc.c: do precalculation separately from parsing to
	  allow for restrictions on precalc

	* mpwrap.c: trim excess zeros off of roundend numbers

Sun Nov 28 23:45:24 1999  George Lebl <jirka@5z.com>

	* inter.c,drgenius-readline-helper.c: use a new readline system
	  with a separate forking process which talks to us with pipes (sue
	  me, corba would be WAAAY too much work for this)

	* genius_lists.c,eval.c,inter.c: move the constant lists of strings
	  to genius_lists.c and make them global.

Thu Nov 04 15:21:46 1999  George Lebl <jirka@5z.com>

	* lib.gel: sanitize triangular numbers

Sat Oct 23 15:54:00 1999  George Lebl  <jirka@5z.com>

	* lib.gel: add ll_test_mprime, it's actually not that slow even
	  though it's interpreted

Sat Oct 23 13:54:06 1999  George Lebl  <jirka@5z.com>

	* inter.c: make toplevels and operators const

Tue Oct 19 20:29:19 1999  George Lebl  <jirka@5z.com>

	* lib.gel: add "triangular" function from Matthew D Allen
	  <s2mdalle@titan.vcu.edu>

Mon Oct 18 21:18:20 1999  George Lebl  <jirka@5z.com>

	* dict.[ch],structs.h,calc.c,funclib.c,lib.gel: add protection
	  of certain global variables from overwriting and an interface
	  to modify the protections

	* mpwrap.c: check one more bit in exp calculation

	* lib.gel,funclib.c: more help strings added

Mon Oct 11 23:53:00 1999  George Lebl  <jirka@5z.com>

	* mpwrap.c: in power series functions, only check the precision we
	  will actually return, this avoids an infinite loop bug

Mon Sep 27 19:27:16 1999  George Lebl  <jirka@5z.com>

	* genius.c: support new parameters

Mon Sep 27 19:20:29 1999  George Lebl  <jirka@5z.com>

	* mpwrap.c,lexer.l: fixup the output base and mixed fraction support

Mon Sep 27 16:24:07 1999  George Lebl  <jirka@5z.com>

	* mpwrap.[ch],lexer.l,eval.c,funclib.c: add integer_output_base and
	  mixed_fractions parameters and implement them, however it most
	  likely doesn't even compile as I can't get the damn
	  autoconf/automake running for some weird reason

Sun Sep 26 21:14:28 1999  George Lebl  <jirka@5z.com>

	* inter.[ch]: add event based readline reading

Sun Sep 26 17:44:48 1999  George Lebl  <jirka@5z.com>

	* eval.[ch], funclib.c, structs.h, inter.c, calc.c: completely
	  go to the new interface and remove a lot of old cruft, and
	  make internal gel function calling work

Sun Sep 26 16:40:05 1999  George Lebl  <jirka@5z.com>

	* eval.c: fix special free from spurious freeing

	* calc.c: use the new interface from eval.c

Sun Sep 26 16:10:21 1999  George Lebl  <jirka@5z.com>

	* eval.c: fix copying of spacers, make a private copy of a matrix
	  before we evaluate any of it's arguments, never evaluate the
	  generic zero element of matrices

Fri Sep 24 19:14:12 1999  George Lebl  <jirka@5z.com>

	* eval.c: fixed matrix^1

	* geniustests.txt: added some tests

Fri Sep 24 18:33:49 1999  George Lebl  <jirka@5z.com>

	* funclib.c: fixed some problems with creating argument lists
	  for new nodes, down to 2 errors on the testsuite

Fri Sep 24 17:32:20 1999  George Lebl  <jirka@5z.com>

	* eval.c: fix matrix memory allocation, fix op_two_nodes for non
	  numeric values, and fix for loop semantics with respect to
	  iterator value after loop. We're down to 5 errors on the testsuite

Thu Sep 23 23:39:39 1999  George Lebl  <jirka@5z.com>

	* eval.c: support all operations except modular arithmetic, there
	  are still some problems, mostly with matricies, testsuite has
	  only, down to 12 errors for the test suite

	* geniustests.txt: update to take into account precalculation

	* geniustest.pl: support tests with null output

Thu Sep 23 18:56:28 1999  George Lebl  <jirka@5z.com>

	* eval.c: now do more work on the primitives and they almost all
	  work for simple numbers, as well as string additions and
	  matrix transpose

Thu Sep 23 18:02:07 1999  George Lebl  <jirka@5z.com>

	* eval.[ch]: cleanup eval_etree and make it possible to do more
	  evaluation with one context, also implement for in loops and
	  do minor cleanups

Thu Sep 23 15:57:25 1999  George Lebl  <jirka@5z.com>

	* eval.c: implement for and for/by loops and break/continue for
	  those, pop off the call for loops in free_special, remove
	  a bunch of old unused methods, and do precalculation during
	  tree building for operations that can't loose precision

Wed Sep 22 22:10:59 1999  George Lebl  <jirka@5z.com>

	* Start writing changelogs again as I come out of the completely
	  experimental stage of the engine into make-it-all-work stage

	* eval.c: implemented new stack and it's waay faster then the old
	  linked list one

1999-09-18  Jacob Berkman  <jberkman@andrew.cmu.edu>

	* Makefile.am (Utilitiesdir): put in the correct
	directory

Sun Sep 12 00:12:12 1999  George Lebl  <jirka@5z.com>

	* eval.c,eval-iter.c: move eval into eval-iter.c as I'm messing with
	  core functionality so this is a testing file for the new
	  evaluation engine

Fri Sep 10 18:07:56 1999  George Lebl  <jirka@5z.com>

	* eval.c: bit of work on new evaluation engine

Thu Sep 09 23:59:26 1999  George Lebl  <jirka@5z.com>

	* eval.c: a bit more work on the new evaluation engine

	* eval.h: start operators at 0

Thu Sep 09 16:09:17 1999  George Lebl  <jirka@5z.com>

	* eval.[ch]: fix some evalnode/evaltree brokenness, and start on
	  implementing a stack based evaluation engine instead of a recursive
	  one

Wed Sep 08 19:28:04 1999  George Lebl  <jirka@5z.com>

	* *.c: use GSList instead of GList everywhere

	* dict.[ch],eval.c,structs.h: store the current highest function
	  on the token itself to avoid another dereference and thus speed
	  things up by a very tiny fraction

Wed Sep 08 18:48:25 1999  George Lebl  <jirka@5z.com>

	* eval.c,calc.c,mpwrap.c: fixed bugs and quibbles reported by
	  Squeak <squeak@xirr.com> and applied his patch

