octave-maintainers
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: building docs


From: Rik
Subject: Re: building docs
Date: Mon, 02 Dec 2013 22:49:24 -0800

On 12/02/2013 10:19 PM, Daniel J Sebald wrote:
> On 12/02/2013 08:56 AM, Rik wrote:
>> On 12/01/2013 08:59 AM, John W. Eaton wrote:
>>> On 12/01/2013 11:56 AM, Daniel J Sebald wrote:
>>>
>>>> The option --no-gui-libs is also needed along with --no-gui.
>>>
>>> If you use --no-gui-libs, then you don't need --no-gui.
>>>
>>> jwe
>>>
>> Back to the original question, is there really a problem building docs on
>> Linux with the current default branch?
>
> I say "yes", unless something has changed in the last couple days since I
> tried a fresh build.  The issue is that to generate the EPS files
> requires executing ./run-octave.  Currently it appears that running
> octave attempts to start the GUI even if in batch mode.
This works for me, as in no GUI is started and a correct  EPS file is
generated.  The test I used was:

cd doc/interpreter
rm convhull.eps
make convhull.eps
## which executes the following
./config-images.sh ../..
../../doc/interpreter/images.mk is unchanged
../../run-octave -f -q -H -p . --eval "geometryimages ('convhull', 'eps');"

>
> This works fine if the system is able to run the GUI, but if the system
> is unable to start Qt the attempts to generate the EPS figures fail. The
> scenario I'm thinking of is if someone logs onto a machine remotely and
> runs the build process.
I tried the above sequence after using ssh to connect to my local machine. 
I ran both with and without X11 forwarding and the eps file was generated
correctly.

--Rik
>
> In some sense I've no qualms with ./run-octave attempting to load the GUI
> libraries.  The issue is that the build process requires actively
> functioning Qt to complete.  If those commands generating EPS files using
> gnuplot graphics engine just used "--no-gui-libs", that wouldn't be the
> case.
>
> This wasn't an issue a few months back.  This weekend I'll search for the
> changeset where this might have changed.
>
> Dan
>
>
>



reply via email to

[Prev in Thread] Current Thread [Next in Thread]