------------------------------------------------------------------
07-08-14 JoeB: Binom doesn't check for positive characteristic,
which means errors when trying to (HOMOGENISEINPUT) in the
non-commmutative mode, like in RABBIT.
Actually, the characteristic modification in PolAlgRead is 
rather ad hoc:
	(COND ((AND Modulus (FIXP Modulus) (LESSP 0 Modulus))
	       (SETQ kk (REMAINDER kk Modulus))
	       (COND ((MINUSP kk)
		      (SETQ kk (PLUS2 kk Modulus))))))
which of course could be copied to Binom, but ought to be
replaced by uniform coefficient converter calls to
InCoeff2RedandCoeff, in all places. (In char 0 (and no SF), this
just replaces 0 with NIL.)
------------------------------------------------------------------



------------------------------------------------------------------
04-04-30 Kae,JoeB: The MODULEHSERIES from the test sequence now
gives an erroneous Hilbert series. (The error appeared between
versions 0.97 and 0.975.)
------------------------------------------------------------------


------------------------------------------------------------------
00-07-01 JoeB: First setting weights, then changing order doesn't
work. Probably similarly for setting variables before order.
 Should it work? Else, we should warn!
------------------------------------------------------------------



------------------------------------------------------------------
99-07-02 SV, JoeB: Erroneous RETURN from PROGN in the anick
t_inout procedures anAssignSDPMode and anAssignTMstrings.
------------------------------------------------------------------



------------------------------------------------------------------
99-02-09 JoeB: Make mode changer alternatives for the pbseries
display procedure in pbseries; and check why turning it off by
redefinition in Roos's script for multi-input anick wasn't
successfull.
------------------------------------------------------------------



------------------------------------------------------------------
99-02-09 JoeB: primaux.sl doesn't compile nicely.
------------------------------------------------------------------



------------------------------------------------------------------
98-10-26 JoeB: Don't forget to fix an appropriate case conversion
update check in the make-bergman scripts!
------------------------------------------------------------------



------------------------------------------------------------------
98-10-23 SK: Some bugs related to case conversion were 
found in mkbergman.

1. $src should be replaced by $bmsrc
2. In the "if"  statement && should be replaced by ||
------------------------------------------------------------------



------------------------------------------------------------------
98-10-23 SK: After the loading of image file in version 0.926 the
message "bergman0.925.img" is displayed.
------------------------------------------------------------------



------------------------------------------------------------------
98-10-22: oldded is very badly used in NCPBHGROEBNER (topproc.sl).
------------------------------------------------------------------



------------------------------------------------------------------
98-10-21 JoeB: SETSETUP doesn't work well with
 (STRATEGY (INTERRUPTION ORDINARY)).
------------------------------------------------------------------



------------------------------------------------------------------
98-02-05: $bmdomains/mkprimefile should be changed to $bmshells/mkprimefile

 --- Done (JoeB) 1998-08-24 ---
------------------------------------------------------------------



------------------------------------------------------------------
97-10-24 JoeB: Staggering problems; e.g.:
pavidus[19]./bin/sun4/bergman
Bergman 0.922, 12-Oct-97
1 lisp> (stagsimple) vars x,y; x^2,x*y+y^2;
*** Function `STAGSIMPLE' has been redefined
*** Function `DEGREEENDDISPLAY' has been redefined
Now input in-variables and ideal generators in mapleform, thus:
        vars v1, ..., vn;
        r1, ..., rm;
where v1, ..., vn are the variables, and r1, ..., rm the generators.
SetupGlobals
 ... done
% 2
x*y+y^2,
   x^2,
   
*** Garbage collection starting
*** GC 1: 24-Oct-97 17:54:02 (~ 272 ms cpu time, gc : 18 %)
*** time 51 ms, 17438 occupied, 10929 recovered, 1682562 free
***** Segmentation Violation in commMonTimes
Break loop
2 Bergman (with stag) break (1) > I
Backtrace, including interpreter functions, from top of stack:
INTERPBACKTRACE C!h!a!r0C!o!n!cP!o!lNM!o!nM!u!l!t 
D!e!n!s!eC!o!n!t!e!n!tC!h!a!r0P!o!lL!i!nC!o!m!b R!e!d!u!c!eP!o!lS!t!e!p 
R!e!d!u!c!eS!u!b!s!t!a!n!t!i!a!lP!o!l STAGGERKERNEL PROG STAGSIMPLE 
STAGSIMPLE STANDARDLISP

 --- Done (JoeB) 1998-08-24 ---
------------------------------------------------------------------



------------------------------------------------------------------
97-09-16 JoeB: Compiling yields the following warnings:
*** in C!h!e!c!kI!n!p!u!t: InRingFile declared fluid
C!h!e!c!kI!n!p!u!t
*** in C!h!e!c!kO!u!t!p!u!tGB: GbOutFile declared fluid
C!h!e!c!kO!u!t!p!u!tGB
*** in C!h!e!c!kO!u!t!p!u!tH!s: HsOutFile declared fluid
C!h!e!c!kO!u!t!p!u!tH!s
 and
*** in ANICKDISPLAY: OldChan declared fluid
ANICKDISPLAY
------------------------------------------------------------------



------------------------------------------------------------------
97-08-27 JoeB: pbseries do not work with setweights. E. g.:
(setweights 2 1)
NIL
...(simple) vars y,x; x*x-y, y^10;
Now input in-variables and ideal generators in algebraic form, thus:
        vars v1, ..., vn;
        r1, ..., rm;
where v1, ..., vn are the variables, and r1, ..., rm the generators.
+t^2*(z^2)
% 2
x*x-y,
   
*** Garbage collection starting
*** GC 5: 27-Aug-97 17:29:01 (~ 442 ms cpu time, gc : 34 %)
*** time 34 ms, 16898 occupied, 265 recovered, 1683102 free
+t^3*(z^2+z^3)
% 3
x*y-y*x,
   
*** Garbage collection starting
*** GC 6: 27-Aug-97 17:29:01 (~ 476 ms cpu time, gc : 39 %)
*** time 34 ms, 17012 occupied, 144 recovered, 1682988 free
+t^4*(z^3+z^4)
*** Garbage collection starting
*** GC 7: 27-Aug-97 17:29:01 (~ 493 ms cpu time, gc : 41 %)
*** time 17 ms, 17074 occupied, 144 recovered, 1682926 free
***** Index `8' out of range for NIL in PNTH
Break loop
13 lisp break (1) > I
Backtrace, including interpreter functions, from top of stack:
INTERPBACKTRACE PNTH F!i!n!dN!e!wRMF!s2 F!i!n!dN!e!wRMF!s< F!i!n!dN!e!wRMF!s 
F!i!xNG!b!e GROEBNERKERNEL PROG SIMPLE STANDARDLISP 
NIL
------------------------------------------------------------------


------------------------------------------------------------------
97-08-17 A seemingly variable assignment depending bug in
bergman-under-Reduce:
Red/sun4/bergman

4: in "$bmsrc/redif.red";
COMMENT Reduce interface procedures. May be loaded from Reduce (with
normal Reduce syntax, no 'lisp-hook*'. Provides a rudimentary interface
for using bergman from Reduce.
 All mode-changers - including setiovars - must be run symbolically.
Then bmgroebner may be called FROM ALGEBRAIC MODE, with a list of
polynomials as (single) argument. Each polynomial should be homogeneous,
and ONLY contain set bergman input variables. If all is well, bmgroebner
returns a reduced groebner basis (as a list of polynomials):
 Right now, only integer or standard form coefficients are permitted;


SYMBOLIC;

	.
	.
	.

!o!r!dbmgroebnerexpr





ON RAISE;


nil


ALGEBRAIC;



;

END;

5: algebraic;

6: set := {(x+y)*(x-y),x*y};

         2    2
set := {x  - y ,x*y}

7: lisp (setinvars '(x y));

8: lisp getiovars();

((x y))

9: bmgroebner(set);

{1,0,0}

 It seems to work if setiovars is used. Perhaps a feature rather than
a bug.



Another bug:
pavidus[30]Red/sun4/bergman

4: noncommify(); setiovars '(x y);simple();

nil


nil

Now input in-variables and ideal generators in algebraic form, thus:
        setinvars v1, ..., vn;
        r1, ..., rm;
where v1, ..., vn are the variables, and r1, ..., rm the generators.
algebraic form (noncomm.) input> x^2-y^2,x*y;
***** Segmentation Violation in algforminput


^$$$2-y^2,x*y;

***** Redundant operator 

 The same bug appears in commutative mode. Might be related to
the next bug:
------------------------------------------------------------_____



-----------------------------------------------------------------
97-08-17 It appears a bug in examples without relations 
a) in R-LISP syntax under Unix;
b) both in ordinary (lisp) and R-Lisp mode under Ms Dos.

This example was running under Unix:

43: simple();
Now input in-variables and ideal generators in algebraic form, thus:
        setinvars v1, ..., vn;
        r1, ..., rm;
where v1, ..., vn are the variables, and r1, ..., rm the generators.
algebraic form (comm.) input> setinvars x,y;
algebraic form (comm.) input> ;
***** Segmentation Violation in listcopy

44: clearall();

nil

45: simple();
Now input in-variables and ideal generators in algebraic form, thus:
        setinvars v1, ..., vn;
        r1, ..., rm;
where v1, ..., vn are the variables, and r1, ..., rm the generators.
algebraic form (comm.) input> setinvars x,y;;
***** Segmentation Violation in listcopy

46: lisp getiovars();

((!x !y) !x !y)

------------------------------------------------------------------_____



-----------------------------------------------------------------
96-12-05: Remove double surcing psl-names from shellenv files.

 --- Done 1997-08-22 ---
-----------------------------------------------------------------


---------------------------------------------------
1997-04-02: I may get erroneous print-out of (e.g.) the P-B series
of the associated monomial ring in other domain moods than char0,
since the print-out uses domain mode dependent procedures. (This
indeed happens in char2 mode.)
1997-04-04: PRELIMINARY FIXED, by re-defining appropriate
Char0Alg(First)TermPrint in inout.sl. However, these are not
declared exported; and should anyhow rather be renamed to or
replaced by general series handling printing auxiliaries.
---------------------------------------------------


---------------------------------------------------
1997-03-03: CfNegP not defined but used in adDP!_PRETTYPRINTSDP
(in ant_inout.sl).
---------------------------------------------------


---------------------------------------------------
1997-03-03: The large numbers in $bmauxil/primaux.sl seem not to
be legible under sun4 compilation.
 Furthermore, the (LOAD "nbig30") therein should be replaced by
something else.
 Besides, a variant NOT accepting bignums might be sensible, in
order to be able to create in situ tables without loading bignums.
---------------------------------------------------


---------------------------------------------------
1996-11-29: Warning *** in ANICKDISPLAY: OldChan declared fluid
during compilation.
---------------------------------------------------


---------------------------------------------------
1996-11-10: (OFF REDUCTIVITY) seems to give error in AmputeMons in
the non-commutative case.

1998-08-24: Seemingly, the reason is that we assume all interned
monomials to have lengths at least cDeg. PBSERIES on clearly is
in compatible with this; but even with PBSERIES off, the interned
MONONE (with PMon = (0)) causes a collaps already at the first
run of AmputeMons.
 Suggested fix: Write a noncommAmputeMons, which only amputes
mons of exactly degree cDeg, skipping the preceeding degrees./JoeB
---------------------------------------------------


---------------------------------------------------
1996-08-29: Modify references to "MAPLE form" in the manual.
 --- Done; but the tex files should be changed too. /JoeB 1997-08-22 ---
---------------------------------------------------


---------------------------------------------------
From sveta@maths.lth.se Thu Aug 29 11:42:53 1996

DEAR JOERGEN,
Here is bug report in  trying setweights.
OUR comments starts from %

4: end;
Entering LISP ... 
Bergman 0.9, 26-Aug-96

%attempt to use setweights

1 lisp> (SETWEIGHTS 1 2)
***** Undefined function `copylist' called from compiled code
------------------------------------------------------------------_


%attempt to redefine copylist and use setweights

2 lisp>  (COPYD 'COPYLIST 'LISTCOPY)   
copylist
3 lisp> (SETWEIGHTS 1 2)
nil

%attempt to use clearall to save the variables. 

4 lisp> (simple)
Now input in-variables and ideal generators in algebraic form, thus:
     setinvars v1, ..., vn;
      r1, ..., rm;
where v1, ..., vn are the variables, and r1, ..., rm the generators.
algebraic form (comm.) input> setinvars x,y; x*y;
***** Undefined function `!*' called from compiled code
5 lisp> (clearall)
nil
6 lisp> (SETWEIGHTS 1 2)
(1 2)

%now weights exists! 

7 lisp>  (simple)
Now input in-variables and ideal generators in algebraic form, thus:
     setinvars v1, ..., vn;
      r1, ..., rm;
where v1, ..., vn are the variables, and r1, ..., rm the generators.
algebraic form (comm.) input> setinvars x,y; x*y;
***** Undefined function `!*' called from compiled code

% now try noncommutative version

8 lisp> (clearall)
nil
9 lisp> (noncommify)
nil
10 lisp> (SETWEIGHTS 1 2)
(1 2)
11 lisp> (simple)
Now input in-variables and ideal generators in algebraic form, thus:
     setinvars v1, ..., vn;
      r1, ..., rm;
where v1, ..., vn are the variables, and r1, ..., rm the generators.
algebraic form (noncomm.) input> setinvars x,y; x*y;
+t^2*(0)
+t^3*(z^2)
% 3
x*y,
   
Done
 - All is OK (I hope). Now you may (e. g.):
   - kill bergman with (QUIT); or
   - interrupt bergman with ^Z; or
   - clear the memory with (CLEARALL), and run a new (SIMPLE).
nil

%looks better, but ...

12 lisp> (clearall)
nil
13 lisp> (noncommify)
nil
14 lisp> (SETWEIGHTS 1 2)
(1 2)
15 lisp> (simple)
Now input in-variables and ideal generators in algebraic form, thus:
     setinvars v1, ..., vn;
      r1, ..., rm;
where v1, ..., vn are the variables, and r1, ..., rm the generators.
algebraic form (noncomm.) input> setinvars x,y; x*x-y;
***** Segmentation Violation in noncommtotaldegree

% maybe problem in the pure variables in the relations?

16 lisp> (clearall)
nil
17 lisp> (noncommify)
nil
18 lisp> (SETWEIGHTS 1 2)
(1 2)
19 lisp> (simple)
Now input in-variables and ideal generators in algebraic form, thus:
     setinvars v1, ..., vn;
      r1, ..., rm;
where v1, ..., vn are the variables, and r1, ..., rm the generators.
algebraic form (noncomm.) input> setinvars x,y; x*x*x*x-y*y;
+t^2*(0)
+t^3*(z^2)
+t^4*(0)
% 4
-y*y+x*x*x*x,
   
+t^5*(z^2+z^3)
+t^6*(0)
% 6
-y*x*x*x*x+x*x*x*x*y,
   
+t^7*(z^3+z^4)
+t^8*(0)
Done
 - All is OK (I hope). Now you may (e. g.):
   - kill bergman with (QUIT); or
   - interrupt bergman with ^Z; or
   - clear the memory with (CLEARALL), and run a new (SIMPLE).
nil
20 lisp> 

%Here GB is OK. But what about Poincare series?

%Another note. Do You know that Bergman does not likes empty list of relations?

Entering LISP ... 
Bergman 0.9, 26-Aug-96
1 lisp> (simple)
Now input in-variables and ideal generators in algebraic form, thus:
     setinvars v1, ..., vn;
      r1, ..., rm;
where v1, ..., vn are the variables, and r1, ..., rm the generators.
algebraic form (comm.) input> setinvars x,y; ;
***** Segmentation Violation in listcopy
2 lisp>


1998-08-24, JoeB: Most of the SETWEIGHTS troubles mentioned above
seem to be fixed now, except the incopatibility with series.
The empty list of relations bug remains, but ONLY in commutative
mode, and ONLY in bergman-under-reduce (!?).
-----------------------------------------------------------------


-----------------------------------------------------------------
96-07-13: Procedures used which are NOT documented in the Standard
Lisp  document, nor defined in slext: LOAD.
-----------------------------------------------------------------


-----------------------------------------------------------------
96-06-24: EMDIM and inner embdim handling is bad and inconsistent.
Probably the best solution is to remove the user available variable
-----------------------------------------------------------------



-----------------------------------------------------------------
96-06-20: The treatment of the monomial 1 is bad commutatively
and catastrophal in the non-commutative case. Already
noncommMONINTERN assumes its argument not to represent 1.
 In any case, SETINVARS sets a global variable InPMonOne to a
list of (# input variables) 0's. This gives a correct lispform
representation of 1 IN THE COMMUTATIVE CASE.
 Perhaps I should intern an augmon representation of 1 in
some initialiser? But, it might be monreclaimed!?

98-08-24, JoeB: The confusion seems larger now. There is InPMonOne;
MONONE (in the non-commutative case); MonOne (in the s.a.w.s. case).
Unify!
-----------------------------------------------------------------


-----------------------------------------------------------------
96-06-17: The new standard forms coefficients mode seems not to
be made fully compatible with SETMODULUS mode changing. Actually,
I haven't decided on the appropriate way to handle modularity
together with Reduce coefficients. Thus SETMODULUS and
SETREDUCESFCOEFFS cancel the effects of each others.
-----------------------------------------------------------------


-----------------------------------------------------------------
94-11-21: I have done modularising too smart (?): I get complaints
about the application of exponent taking on a PB series coefficient
(running with modlogarithms and modulus 7):
-----------------------------------------------------------------
*** Function `DEGREEENDDISPLAY' has been redefined
SetupGlobals
 ... done

% No. of Spolynomials calculated until degree 2: 0
% No. of ReducePol(0) demanded until degree 2: 0
% Time: 2669
*** Garbage collection starting
*** GC 3: 21-Nov-94 14:06:37 (~ 2720 ms cpu time, gc : 3 %)
*** time 51 ms, 36572 occupied, 501909 recovered, 9963428 free
***** Subscript `17' in GetV is out of range
Break loop
6 lisp break (1) > I
Backtrace, including interpreter functions, from top of stack:
INTERPBACKTRACE R!e!d!o!rC!o!e!f!f2O!u!tC!o!e!f!f C!h!a!r0A!l!gR!e!d!o!rT!e!r!mP!r!i!n!t
ALGOUTLIST DEGREEPBSERIESDISPLAY TDEGREECALCULATEPBSERIES1
TDEGREECALCULATEPBSERIES PROGN DEGREEENDDISPLAY GROEBNERKERNEL PROG
NCPBHGROEBNER STANDARDLISP
-----------------------------------------------------------------
