Re: bug in scatter() in 3.6.2 ?

[Ben Abbott]

On Jan 22, 2013, at 9:18 AM, Ben Abbott wrote:

> On Jan 22, 2013, at 7:33 AM, Andreas Weber wrote:
> 
>> On 22.01.2013 10:57, Eric Chassande-Mottin wrote:
>> 
>>> I think I've found a bug in scatter()
>>> N=101; a=rand(1,N); b=rand(1,N); c=ones(1,N); scatter(a,b,20,c,"filled");
>>> error: invalid value for color property "markerfacecolor"
>> 
>> Dear Eric,
>> I can confirm this with octave3.6.3 and graphics_toolkit gnuplot.
>> It works as expected with 3.6.3 and fltk. This also works with gnuplot and 
>> the current default branch 139f4b19a3ac.
>> 
>> Perhaps this patch http://hg.savannah.gnu.org/hgweb/octave/rev/566cf544d020 
>> tried to adress this issue.
>> 
>> You may try to change line 271 in __scatter__.m to
>>>   gnuplot_hack = (numel (x) > 1 && columns (c) == 3
>>> && strcmp (toolkit, "gnuplot"));
>> 
>> to see if my guess is right. Regards Andy
>> 
> 
> The default branch works correctly.   The changeset below fixed the problem.
> 
>       http://hg.savannah.gnu.org/hgweb/octave/rev/566cf544d020
> 
> I produced the attached changeset by back-porting/grafting to the stable 
> branch.
> 
>       hg update stable
>       hg graft 566cf544d020
>       hg tip
>       hg export tip > ~/Desktop/changeset.patch
>       hg rollback
>       hg revert --all
> 
> I'm doing a fresh build of the stable branch to check that nothing breaks.  
> If all looks ok, I'll push this change.  Any concerns/objections?
> 
> Ben

The changeset works for me.  None of the scatter3 demos broke.

Since we're close to a release, I'd appreciate it if someone else can take a 
look and confirm they don't see any regressions or new bugs after applying the 
changeset.

Ben