[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: issues using print on fltk/gnuplot
From: |
Ben Abbott |
Subject: |
Re: issues using print on fltk/gnuplot |
Date: |
Sat, 14 Jan 2012 10:25:02 -0500 |
On Jan 14, 2012, at 10:12 AM, HomeRun4711 wrote:
> Am 14.01.2012 15:11, schrieb Ben Abbott:
>
>> On Jan 14, 2012, at 3:23 AM, Walter White wrote:
>>
>>> Hello,
>>>
>>> I came across some issues while using fltk on Octave MinGW 3.4.3.
>>>
>>> 1.)
>>> Sometimes (randomly) figures are just not printed as png.
>>> I could not figure out how, since the conditions are always the same.
>>> My first suspicion was file name length, but some of the images are
>>> printed, some not, all having the same file name length.
>>> Neither is the hard disk full and there are no error messages.
>>>
>>> My command is
>>> print(print_filename, '-dpng','-FHelvetica:8','-S1024,768')
>>>
>>> I flipped the graphics toolkit to gnuplot and so far everything
>>> works fine there. But I am not sure if this is related to the
>>> toolkit since the problem occurs randomly...
>>>
>>> Do you know what might cause this problem?
>>>
>>> 2.)
>>> gnuplot and fltk are not using the correct plot size given by the command
>>>
>>> print(print_filename, '-dpng','-FHelvetica:8','-S1024,768'
>>>
>>> All files are 1023x767, 1 px less on each side...
>>>
>>> 3.)
>>> Some of the plots, though given -S1024,768, are in a completly
>>> different format (1333x917). I have no idea why...
>>>
>>> Has someone of you ever come across similar issues?
>>>
>>> Kind regards,
>>> Walter
>>
>> I've filed a bug report for 2)
>>
>> https://savannah.gnu.org/bugs/index.php?35307
>>
>> I'm unable to help with 1) or 3) as I can't duplicate the problem.
>>
>> If you can give us more details as to how the figures were produced (from
>> the command line, in a script, in a loop, in rapid succession, etc) we maybe
>> able to help isolate the conditions that produced the incorrect result. Also
>> for 3) was the problem with fltk, gnuplot, or both ?
>>
>> Ben
>
> Hello,
>
> thanks for your response.
>
> About 1.) and 3.), the figures are produced using a script, from within a
> loop, each turn of the loop produces three figures in rapid succession,
> before each image there is a clf and close all.
> Could this be a problem?
> So far I thought Octave takes its time to create the graphs.
> Is it possible to catch a return value from print() or something like this in
> case the figure was not produced?
> Or should Octave complain if there was a problem?
>
> I have never had this issue before on my Ubuntu installation
> though I was producing figures in the same rapid succession.
>
> About 3.), so far this only happened using fltk, but I have been using
> gnuplot only for a few hours now, so since this occured randomly using
> fltk, it could happen in the next days...
>
> Kind regards,
> Walter
*** please respond below so that those arrive late can follow along ***
Unfortunately, Octave doesn't have a way of knowing when gnuplot / gl2ps are
done generating their eps-output.
If slowing things down a bit isn't a problem, you can try adding "pause (1)"
after each plot. That should be sufficient time for gl2ps to finish its job.
Ben