Fwd: Re: regression with draw3d ?



El 27/08/12 06:50, Jean-Claude Arbaut escribi?:
>> Still same problem.
>>
>> Here is an experiment : I change "draw .lisp" so that there is
>>
>> (setq *autoconf-win32* "true")
>> (defvar *windows-OS* (string= *autoconf-win32* "true"))
>> ($print *autoconf-win32*)
>> ($print *windows-OS*)
>>
>> Then when I load, it print
>> true
>> false
>>
>> Next, I change "draw.lisp" again (notice the "setq" instead of "setvar")
>> (setq *autoconf-win32* "true")
>> (setq *windows-OS* (string= *autoconf-win32* "true"))
>> ($print *autoconf-win32*)
>> ($print *windows-OS*)
>>
>> And I get
>> true
>> true
>>
>> And of xourse I can then plot with
>> draw2d(explicit(x,x,0,1))$

This is an interference with wxMaxima. Global variable *windows-OS* is 
also declared with defvar in file wxmathml.lisp. When you call package 
draw, *windows-OS* was already declared, and its value is not changed by 
the new defvar. With other Maxima interfaces, your first experiment 
should work.

Not sure if this declaration can be safely removed from wxmathml.lisp.

> Also, the only difference between the two maxout.gnuplot files (with 
> Maxima 5.27 and 5.28) is two lines :
> set obj 1 rectangle behind from screen 0.0,0.0 to screen 1.0,1.0
> set style rectangle fillcolor rgb '#ffffff' fs solid 1.0 noborder
>
> They are inserted just after
> set terminal windows enhanced  size 600, 500
>
> set size 1.0, 1.0
> set origin 0.0, 0.0
>
> I don't know if it has something to do with the problem above, seems 
> unlikely though...

This was a fix to make draw compatible with Gnuplot 4.6. The new version 
of Gnuplot is not 100% back compatible. I don't think this is 
responsible for your problems.

--
Mario