stefano franchi | 19 Sep 16:17 2014

Re: GSOC LyXToWordConversion

Hi Tariq,

sorry for not getting back to you---I thought Prannoy's message had reached you. My mistake.

On Fri, Sep 19, 2014 at 5:28 AM, tesmai4 <at> <tesmai4 <at>> wrote:
Did you manage to produce a single gsoc repository for testing purposes?

Prannoy created a single repository called "prannoy" that contains all the work he did over the summer. It is one one the main branches in the gsoc repository:;a=summary

Let me know how your tests go.



Stefano Franchi

stefano.franchi <at>
Richard Heck | 16 Sep 18:41 2014

2.1.2 Tarballs Now Available

You can find tarballs for 2.1.2 at
Please prepare binaries ASAP. I'd like to release at the end of the week, if possible.


José Matos | 15 Sep 18:39 2014

Torvalds says he has no strong opinions on systemd

While the above piece

I noticed that Linus refers to LyX in the context of the unix way:

"And systemd is in no way the piece that breaks with old UNIX legacy. Graphical applications seldom worked
that way (there are certainly _echoes_ of it in things like "LyX", but I think it's the exception rather
than the rule)..."

Funny and interesting to read at the same time. :-)


José Abílio

Scott Kostyshak | 12 Sep 14:53 2014

SystemcallPrivate::exitCode() does not return exit code if crash

If you have an updated TeX Live 2014 installation, exporting
examples/FeynmanDiagrams.lyx to PDF with LuaTeX will not work. But LyX
doesn't detect that something goes wrong. You need LuaTeX version
beta-0.79.1 to reproduce the following. The LuaTeX bug will not
trigger the LyX bug with LuaTeX version beta-0.76.0-2013061708, for
example. Also, I have no idea if this depends on the version of Qt. I
have confirmed that Kornel also sees the zero exit code but no
exported file, but we have similar systems.

In the GUI, LyX says that the preview was successful and tries to open
the PDF. On the command line, LyX gives a zero exit code:

If you run lualatex directly, you get a non-zero exit code:

lualatex: ../../../tex-live-2014-04-01/texk/web2c/luatexdir/tex/mlist.w:2542:
make_ord: Assertion `varmem[(p)].hh.v.RH!=0' failed.
Aborted (core dumped)
$ echo $?

This is LuaTeX, Version beta-0.79.1 (TeX Live 2014) (rev 4971)
 restricted \write18 enabled.
LaTeX2e <2014/05/01>
Babel <3.9k> and hyphenation patterns for 79 languages loaded.

If you run with LyX (I added some debug code for extra information):

$ lyx -e pdf5 FeynmanDiagrams.lyx
lualatex: ../../../tex-live-2014-04-01/texk/web2c/luatexdir/tex/mlist.w:2542:
make_ord: Assertion `varmem[(p)].hh.v.RH!=0' failed.
support/Systemcall.cpp (294): Systemcall: 'lualatex
"FeynmanDiagrams.tex"' exit code is: 0
d.errorMessage() is: The process crashed some time after starting successfully.
d.exitStatusMessage() is:  The process crashed.
$ echo $?

The strange thing going on here is that SystemcallPrivate::exitCode is
returning 0 because QProcess::exitCode returns 0. I'm not sure why
this happens. The documentation states that QProcess::exitCode
"Returns the exit code of the last process that finished." Is the
point that because lualatex crashed the process didn't "finish"? I
don't think so, because when I use the following patch:

 <at>  <at>  -610,7 +619,7  <at>  <at>  int SystemcallPrivate::exitCode()
        if (!process_)
                return -1;

-       return process_->exitCode();
+       return process_->exitStatus();

SystemcallPrivate::exitCode correctly returns non-zero (it returns 1)
and it is also described as "Returns the exit status of the last
process that finished." After replacing exitCode with exitStatus and
applying the patch I just sent, LyX behaves as expected. But that is
not quite correct because I think we want to return the exit code and
not just the exit status.

Any ideas?


Scott Kostyshak | 12 Sep 14:53 2014

[PATCH] Check exit code of LaTeX process in LaTeX::run

Systemcall::startscript returns the exit code of the LaTeX command
that is run, but the return value was not being checked by
LaTeX::run. It is now checked and if the exit code is non-zero,
it is returned immediately.

At a higher level, if the LaTeX command returns a non-zero
exit code, in the GUI a message such as "Error while exporting
format: PDF (LuaTeX)" is given instead of "Successful preview
of format: PDF (LuaTeX)" followed by the attempt to open an
incomplete/missing .pdf file.

When run on the commandline, lyx -e lualatex example.lyx now
gives "Error: LaTeX failed" and a non-zero exit code where before
it gave a zero exit code.

Any thoughts? See my other email for a use case.

Scott Kostyshak | 11 Sep 15:21 2014

conditional loading of packages from modules

I ran into a problem that might be best left alone, but wanted to check in here.

In my document I use the FiXme module. Recently I did some editing and
got an error because I happened to have deleted the last inset offered
by FiXme in my document. The reason I got the error is because LyX no
longer loaded the FiXme package (because there were no more insets) so
the command I use in the preamble caused an undefined control sequence

 % Makes it so FiXme notes always show up (e.g. without draft mode).

This system makes perfect sense to me and I understand that all rules
are off when I use the preamble. My guess is that this is expected,
but I just wanted to make sure? The only alternatives I can think of
would be to either always load the package specified by the module or
to have an option system like we have for Math Options (this seems
like overkill though).

I think I've narrowed this down to one question: Is there an advantage
to not loading the package fixme if there is no fixme inset used?

Note that I fixed the issue by instead using the following in my preamble:
I fixed this by instead using:
\ <at> ifpackageloaded{fixme}{\fxsetup{draft}}


Scott Kostyshak | 11 Sep 10:26 2014

Beamer "duplicate frame"

I often find it useful to duplicate a frame and go from there rather
than starting from new. Using the mouse to select a section and copy,
then position, then paste, often does not work for me. Somehow I mess
it up because I often position the cursor incorrectly.

I'm now using the following command sequence and it's working pretty
well (in limited testing). I click at the beginning of the "Frame"
layout (just before the frame title), and run it:

command-sequence section-select; copy; char-forward; paste

Would anyone find a "section-duplicate" LFUN useful? I have trouble
thinking of non-Beamer uses, so my current thought is that it should
not be included, but I wanted to check here to see what others


Mostafa | 9 Sep 23:15 2014

INSTALL.Win32 patch.

Hi all,

My first time using the development version of LyX on Win7 and I spotted a  
couple of minor typos and other issues in INSTALL.Win32 Attached is a  
patch. Hope it helps.
Attachment (INSTALL.Win32.diff): application/octet-stream, 1410 bytes
Richard Heck | 9 Sep 16:01 2014

GMO files?

I am all but ready to release 2.1.2. But I noticed that Kornel committed a GMO file after fixing up something to do with pl.po. Is there something I need to do to regenerate and then commit GMO files?


stefano franchi | 8 Sep 15:49 2014

Re: GSOC LyXToWordConversion

Hi Tariq,

thanks for asking about the project. I am afraid it is not quite ready yet for public consumption. Prannoy put together some scripts that tie everything together, but all the code is still parked in the various gsoc repositories.
I haven't had the time to look at it since the end of Google's summer of code because of a major professional deadline that's coming up this week. I hope to get back to the project next week and produce at the very least a single gsoc repo that could be easily cloned for testing purposes.



On Thu, Sep 4, 2014 at 4:29 AM, tesmai4 <at> <tesmai4 <at>> wrote:
Dear Stefano,

I did not have any update from you regarding LyxToWordConversion. Is it ready for use?


On Thu, Aug 7, 2014 at 11:07 PM, stefano franchi <stefano.franchi <at>> wrote:
Hi Tariq,

it is great to hear you are willing to test the work Prannoy did over the summer in the context of Google's summer of code.
As Richard pointed out, the code (it's all in .4ht files) is in the gsoc repository, and more precisely in the tests directory.

I would just ask you to wait a few more days, as Prannoy is now in the final week of the project and he is moving stuff around and testing a script that allows a the conversion to ODT to be launched directly from the LyX interface, thereby greatly simplifying the testing process.

Bug me again in a week or so if you don't hear from me.



On Thu, Aug 7, 2014 at 2:00 PM, Richard Heck <rgheck <at>> wrote:
On 08/06/2014 10:52 AM, tesmai4 <at> wrote:
Dear Richard,

Thanks for your reply.

I would like to participate in testing of the GSOC project to LyxToWordConversion [].

Would someone please add me to the list and provide the binaries/code for testing?

Communication about the project happens on this list. The code is in the LyX GSOC repository, which you can browse here:;a=summary
I'm not sure which branches (listed at the bottom) are part of this project, but presumably odt2lyx is one of them, as are the lyx2word branches.

I'm cc'ing the people mainly involved in this. They can give us more info.


Stefano Franchi
Associate Research Professor
Department of Hispanic Studies         Ph:   +1 (979) 845-2125
Texas A&M University                          Fax:  +1 (979) 845-6421
College Station, Texas, USA

stefano <at>

Stefano Franchi

stefano.franchi <at>
Richard Heck | 4 Sep 16:19 2014

Re: Is adding modules slow for anyone else?

On 09/04/2014 04:42 AM, Jürgen Spitzmüller wrote:
> Am Mittwoch 03 September 2014, 22:14:23 schrieb Scott Kostyshak:
>> Sometimes when I click on "Add" in the document settings module tab,
>> LyX lags just a tad (e.g. 1 second). It does not happen all the time
>> and the lag is not much but it makes me wonder what is taking so much
>> computation. My computer is not super fast and I am using a devel
>> build, but relative to other parts of LyX, it feels a little slow to
>> me.
>> One thing I've noticed is that every time I click "Add",
>> updateModuleInfo() is called three times. I have not investigated much
>> beyond that and thought I would check in here first to see if (1)
>> anyone else has noticed this or I am just being picky/ I have a slow
>> computer; and (2) if anyone has an idea why it is slow.
> Are these modules in the recent layout format? Maybe some layout2layout
> conversion takes time.

I'm not sure why updateModuleInfo() gets called three times, but I've seen
such redundancies before. The other thing that's happening, though, that
Jurgen is basically pointing out, is that we rebuild the DocumentClass each
time a module is added. We have to do this to reset various things in the
dialog that the module might change. (E.g., the module might make other
modules available, or exclude them.)

I think hunting down the source of the redundancies would be worthwhile.