2001-12-09  Vadim V. Zhytnikov <vvzhy@mail.ru>
        
        * New online info browsing code for clisp and cmulisp
	by Raymond Toy. To make describe work with clisp you have
	to compile clisp with REGEXP module and later invoke it
	during maxima compilation and at run time with full
	linking kit
	  clisp -K full
	To make it work with cmulisp you need regex.o object
	file. It's source regex.c/h is part of glibc and also
	can be found in clisp sources. 	Before compiling maxima 
	with cmulisp edit compile-cmulisp.lisp, cmulisp-regex.lisp 
	and replace
	  /apps/.../regex-0.12/regex.o
	by he path where regex.o resides on your system (value of 
	the *regex-lib* variable).

2001-11-19  Vadim V. Zhytnikov <vvzhy@mail.ru>

	*  patches to fix problems with new clisp releases.
	
2001-04-20  Bill Schelter  <wfs@sonia.ma.utexas.edu>

	* Now clisp and gcl work ok on all tests No errors found,
	and apparently cmulisp does too.

2001-04-18  Bill Schelter  <wfs@sonia.ma.utexas.edu>

	* trying to correct all the remaining bugs for clisp.
	there is still one lurking, but have eliminated 7 more in
	trprop.lisp, rat3e.lisp,	mstuff.lisp
	* fixed the source level debugging code and debugger code
	for both clisp and gcl versions.  

	* Multiple changes to make maxima under cmulisp perform all the
	tests.
	* changes to plot.lisp to remove gcl dependencies.
	* move some system specific stuff to the compile-{clisp,cmulisp}
	files.  

2001-04-15  Bill Schelter  <wfs@sonia.ma.utexas.edu>
	* added the file specfun.mc from Barton Willis, and the
	documentation to the manual under Orthogonal Polynomials,
	just after Special Functions.

	*  added the special declarations, to the maxima to lisp
	compiler, so that the correct dynamic binding scenario is
	respected.

	* changed the default optimize level for compilation
	of the maxima files to -O4, by changing this in GCL. 
	
	*  added changes to integrate to handle the irreducible
	denominators if given rational function f/g, using the
	rootsof method.   To enable that use INTEGRATE_USE_ROOTSOF:true;

2001-04-06  Bill Schelter  <wfs@sonia.ma.utexas.edu>

	* Fix the install to install all share/*.lisp
	* Fix auto-mexpr which had a bug, but only seen
	with 'trigrat'
	* Fix file_search to also look for .lisp in share/*
	* have the install on MS Windows add the .mc and .lisp files in
	share
	

2001-03-30  Raymond Toy <toy@rtp.ericsson.se>
	
	compile-cmulisp.lisp:
	o Add some text to explain how to build it.
	o Add support for maxima's describe.
	o Changed the name of the core file to maxima.core.
	
	macdes.lisp:
	o Add $describe for CMUCL.
	
	init_max1.lisp:
	o Changed SYSTEM:: to SI::
	o Set up file extensions appropriately for searching.
	o Update user::run to load up the foreign code so describe works.
	
	mload.lisp:
	o In $file_search, symbols are passed to probe-file and friends.
	  Convert to strings.
	o In new-file-search, pass a string instead of NIL to
	    new-file-search1.
	    
	mactex.lisp:
	o Make tex(1.2e20) work on CMUCL, and change the output to 1.2 \times
	      10^{20} instead of \cdot.
	      
       clmacs.lisp:
        o most-positive-fixnum is too large a count for CMUCL in delq and
	 zl-remove.

	
2001-03-30  Bill Schelter  <wfs@sonia.ma.utexas.edu>

	* fix to nparse.lisp for parsing infix and prefix user
	defined operators
	* fix in sin.lisp to fix a bug involving definite integrals
	eg: integrate((sec(t))^2,t,-%pi/4,0);  was incorrect.

Raymond Toy <toy@rtp.ericsson.se>

	SYS-PROCLAIM.lisp:
	o Not sure about these, but some proclaims are definitely wrong.  (I
	  got this from the 5.4 patches on the CMUCL web site.)
	o Put #+gcl around the COMPILER:: symbols
	
	displa.lisp:
	o linearray should be initialized to NIL.
	
	init_max1.lisp:
	o Add CMUCL "version" of getenv.
	
	nparse.lisp:
	o Change occurrences of string-char to #.(array-element-type "a").
	  Wouldn't it have been better to leave string-char everywhere and
	  just do a deftype string-char to base-char for those systems without
	  string-char?
	
	simp.lisp:
	o Don't define the onep macro. It's already a function.  Or maybe get
	  rid of the function and leave the onep macro in?  CMUCL has inline
	  functions so the onep function can be as fast as the macro.
	
	sysdef.lisp:
	o Get the right extensions for :maxima :object-path for CMUCL.
	o I turned off the (safety 0) for CMUCL, but perhaps that's not
	  necessary anymore.

2000-05-23  Bill Schelter  <wfs@linux34.ma.utexas.edu>

	* Fix to $save
	* I can see this changelog is a bit out of date!!!
	Only 10 years of changes left out of it ...

Wed Dec 13 22:38:05 1989  Bill Schelter  (wfs at nicolas.ma.utexas.edu)

	* in psmonsubst check that the arg p is not the constant series.
	eg. taylor(x*(1-%e^(2*x)),x,0,0); yields a call where the
	arg is a rational (0 . 1), and so no substitution is necessary.

Sat Nov 11 07:24:45 1989  Bill Schelter  (wfs at nicolas.ma.utexas.edu)

	* There seem to be other problems with resultant:mod;
	We have removed its setting, so that an error will be
	signalled in this mode.
	* misc changes to help file_search
	* describe fixed [it was missing a number of doc items]
	* nutplot added (see nutplot.lisp) for 2d plots with nutmeg


Tue Jul  4 08:11:59 1989  Bill Schelter  (wfs at rascal)

	* fix resultant:mod; option in ufact.lisp.
	Boole was given more than the 3 args mandated by
	common lisp.

Fri Mar 17 18:27:40 1989  Bill Schelter  (wfs at rascal)

	* Remove a #'ncons in trans3.lisp, change defmtrfun
	to refer to defmfun, so that the nargs construct
	will be ok.
	* Fix bug in File_search so that lisp pathname objects
	permitted in the file_search path.
	

Tue Oct 25 02:51:50 1988  Bill Schelter  (wfs at rascal)

	* Added Lazard's trigrat package.
	Added autoloading capability for maxima functions.
	(see max_ext.lisp and autol.lisp)
        Fixed the #$ reader array in commac.lisp to be adjustable.
	Changed primcyclo because cyclotomic can return a fixnum
	(and I think this can happen there). (file factor.lisp).
	

Tue Oct  4 15:54:59 1988  Bill Schelter  (wfs at rascal)

	* nparse.lisp:
	make the parse error swallow further input, and
	print a short message of what was wrong, then
	return to the prompt.   This is for non lispms,
	there we interface with the rubout handler.
	add the tensor/* files to the distribution.
	ctensor.mac has been tested a bit, itensr has not.
	src/maxima-mode.el the gnu emacs maxima mode has been
	improved.

Mon Oct  3 10:07:14 1988  Bill Schelter  (wfs at rascal)

	* spgcd.lisp: determine-lifting-factor had an illegal let*

Sun Oct  2 22:40:19 1988  Bill Schelter  (wfs at rascal)

	* Altered the maxima hash table referencing and setting
	under use_fast_arrays, to use all indicies not just the
	first.   In the end it is probably worth the check
	for more indicies, and to do the consing. files mlisp, acall,ar,array
	* copy-rest-arg a macro (does nothing except on lispm)file:clmacs
	* fixed up the maxima-mode.el somewhat
	* altered let in letmac to be closer to the common lisp let,
	and to allow declares.  Note the order of evaluation
	in a let had been wrong.   It would appear (amazingly) that
	not too much code depends on the order of evaluation there.
	* added special declarations in make-tramp$ (numer.lisp)
	This bug showed up in trying to do romberg(sin(x),x,0,2);
	* changed the mread-synerr syntax error message to not go into the
	debugger.

Thu Sep 29 11:17:35 1988  Bill Schelter  (wfs at rascal)

        Some changes done fairly recently:
	* Numerous changes to bessel to make it work.
	* changes to hyp and hpgeo.lisp to make specint work
	* addition of specint to tests.
	* DOC file changed to allow use of find-doc.el
	a gnu emacs utility (source  for find-doc.el is inside the akcl
	distribution).   Allows completion on doc strings.
	* maxima-mode.el file included for gnu emacs,
	which supports a maxima mode for buffers with
	maxima code, and an inferior-maxima-mode, for
	running maxima inside emacs in a shell or shell
	like window.   Appropriate syntax matching is supplied.
	* makefiles changed to allow them to work on
	systems like HP which don't support the include syntax.
	

Thu Sep 15 16:46:49 1988  Bill Schelter  (wfs at rascal)

	* Fix bug in matcom.lisp (to do with specials and
	rule definitions).
	Misc fixes to mload.
	Fix bug involving complaint about format args
	when trying to print version.  (the file
	src/version.lisp was not always getting loaded)
	
	

Thu Jun 16 12:07:08 1988  Bill Schelter  (wfs at rascal)
	* Fix defn of $status

	* Added autoloading capabilities.  If you load
	autol.lisp and max_ext.lisp, just before dumping
	AND set (get :maxima :object-path) to a full
	pathname then autoloading of things like nusum,..
	should take place.  It may be possible to delete
	a number of sources from the default load.
	

Wed Jun 15 19:46:40 1988  Bill Schelter  (wfs at rascal)

        * Fix example to work.  Add manual.demo to the doc directory.
	
	* Extend alphalessp to make
	numbers<strings<symbols<lists< any other,
	and if x and y aren't in the above types we compare
	the printed representations.

	* ../src/version.lisp was not being loaded in akcl makefile,
	version, causing error in macsyma-top-level, when
	trying to print the version.
	

Tue May 24 09:42:41 1988  Bill Schelter  (wfs at rascal)

	* Changed the $*, $+,.. to use double-float.
	shadow float to be such that (float x) --> double.
	Reader reads doubles by default.  

	* Change the max,min  constants in cpoly.  They needed
	to be doubles.  I left them approximately the same size
	as before (namely used the short-float max min), to avoid
	overflow problems.

	* Verified the test suite. All ok.
	
	
	

