[XeTeX] sporadic shifted underbrace error

Daniel Greenhoe dgreenhoe at gmail.com
Tue Feb 12 04:11:37 CET 2013 

On Tue, Feb 12, 2013 at 11:05 AM, Daniel Greenhoe <dgreenhoe at gmail.com> wrote:
> Does this mean that the fix by Khaled did not make it into the latest
> win32 binary? Would anyone happen to know when a fixed one might
> possibly be available?

P.S. Sorry for all the whining.

Dan

On Tue, Feb 12, 2013 at 11:05 AM, Daniel Greenhoe <dgreenhoe at gmail.com> wrote:
>>> On Fri, Feb 8, 2013 at 9:41 PM, Khaled Hosny <khaledhosny at eglug.org> wrote:
>>> > The issue is now fixed in git repository.
> On Sat, Feb 9, 2013 at 6:53 AM, Khaled Hosny <khaledhosny at eglug.org> wrote:
>> http://sourceforge.net/p/xetex/code/
>
> This link references http://tug.org/texlive/build.html which lists
> Akira Kakuto as the builder for win32. Akira Kakuto apparently
> maintains http://w32tex.org/. As of 2013 February 11, the newest file
> in that repository is xetex-w32.tar.xz updated 2013 February 11
> 11:04am. However, in the change log for 2013 February 11 it says
>   "(04) xetex-w32.tar.xz
>      Update libpoppler (0.22.1).
>      A bug in the unicode-math is not fixed, because the fixed one
>      contains rather large differences and it is under discussion."
>
> Does this mean that the fix by Khaled did not make it into the latest
> win32 binary? Would anyone happen to know when a fixed one might
> possibly be available?
>
> Dan
>
>
>
>>
>> But there are lots of new stuff, so bugs are expected. Use on your own.
>>
>> Regards,
>> Khaled
>>
>> On Sat, Feb 09, 2013 at 05:15:16AM +0800, Daniel Greenhoe wrote:
>>> Is the git repository that you are using accessible only to the
>>> unicode-math developers, or to anyone with git software
>>> (http://git-scm.com/ or http://msysgit.github.com/ or ?) installed on
>>> their machine?
>>>
>>> On Fri, Feb 8, 2013 at 9:41 PM, Khaled Hosny <khaledhosny at eglug.org> wrote:
>>> > The issue is now fixed in git repository.
>>> >
>>> > Regards,
>>> > Khaled
>>> >
>>> > On Fri, Feb 08, 2013 at 03:48:51AM +0200, Khaled Hosny wrote:
>>> >> Thanks for reproducing the issue. This is clearly an engine bug (unlike
>>> >> old TeX, the over/under braces are handled solely by the engine not
>>> >> constructed by macros).
>>> >>
>>> >> I’ll try to debug this now it is reproducible, but whatever fix I can
>>> >> come up with will not be available before the next XeTeX release (TeX
>>> >> Live 2013), so It won’t solve your immediate problem, sorry.
>>> >>
>>> >> Regards,
>>> >> Khaled
>>> >>
>>> >> On Wed, Feb 06, 2013 at 12:50:17PM +0800, Daniel Greenhoe wrote:
>>> >> > Khaled, thanks for the great suggestion. I used an open type math font
>>> >> > (xits-math.otf) and the problem now shows up multiple times. In the
>>> >> > attached pdf, the shift occurs on pages 3, 8, 9, 10, 19, 21, 23, 28,
>>> >> > 34, ...
>>> >> >
>>> >> > My setup includes these:
>>> >> > \usepackage[no-config,no-math]{fontspec}  % font selection for xelatex
>>> >> > \usepackage{xunicode}                 % unicode character support
>>> >> >  \usepackage{amsmath}                 % AMS math package
>>> >> >  \usepackage{unicode-math}            % unicode math support
>>> >> >  \unimathsetup{math-style=ISO,bold-style=ISO,vargreek-shape=TeX}
>>> >> >  \setmathfont{xits-math.otf}
>>> >> >  \setmathfont[range={\mathcal,\mathbfcal},StylisticSet=1]{xits-math.otf}
>>> >> >
>>> >> > Attached is a tex file and pdf output. Can anyone else reproduce the
>>> >> > errors? Any suggestions?
>>> >> >
>>> >> >   command line:  xelatex underbrace
>>> >> >
>>> >> > Dan
>>> >> >
>>> >> > On Wed, Feb 6, 2013 at 11:43 AM, Khaled Hosny <khaledhosny at eglug.org> wrote:
>>> >> > > I’ve seen this once with OpenType math fonts, but I wasn’t able to
>>> >> > > reproduce it again. Trying using the same fonts in your minimal example,
>>> >> > > since it is probably and OpenType math-only issue.
>>> >> > >
>>> >> > > Regards,
>>> >> > > Khaled
>>> >> > >
>>> >> > > On Wed, Feb 06, 2013 at 11:13:51AM +0800, Daniel Greenhoe wrote:
>>> >> > >> I did make a quick minimal example (but not exactly the same as in the
>>> >> > >> previous email) that uses the underbrace 1000 (?) times. But so far
>>> >> > >> apparently no success in getting the problem to show itself again.
>>> >> > >>
>>> >> > >> If interested, please see attachments.
>>> >> > >>
>>> >> > >> Dan
>>> >> > >>
>>> >> > >>
>>> >> > >> On Wed, Feb 6, 2013 at 10:32 AM, Daniel Greenhoe <dgreenhoe at gmail.com> wrote:
>>> >> > >> > For some time now I have been having *sporadic* problems with
>>> >> > >> > underbraces being shifted to the right by maybe 15mm or so.
>>> >> > >> >
>>> >> > >> > The problem is sporadic---it appears sometimes, but on the next
>>> >> > >> > compile may well have disappeared. The document I am currently working
>>> >> > >> > on contains several hundred pages. I can try to put together a minimal
>>> >> > >> > example...but that may be of limited use since it may be difficult to
>>> >> > >> > get the problem to show itself when I want it to show itself.
>>> >> > >> >
>>> >> > >> > So, before I do that, might anyone have an idea of what might be the problem?
>>> >> > >> >
>>> >> > >> > There is a screen capture of a shift. The right underbrace has been
>>> >> > >> > shifted far to the right. It should be like the underbrace on the
>>> >> > >> > left.
>>> >> > >> >
>>> >> > >> > Here is my current underbrace macro:
>>> >> > >> >
>>> >> > >> > \newcommand{\mcom}[3][c]{\ensuremath{%
>>> >> > >> >   \begin{array}[t]{#1}%
>>> >> > >> >     \underbrace{#2}\\%
>>> >> > >> >     \text{\addfontfeatures{Color=mcom}\scriptsize{#3}}%
>>> >> > >> >   \end{array}%
>>> >> > >> >   }}%
>>> >> > >> >
>>> >> > >> > The color mcom is simply defined liked this:
>>> >> > >> >   \definecolor{mcom}           {cmyk}{0,0,0,1}
>>> >> > >> >
>>> >> > >> > The \mcom macro can be used something like this:
>>> >> > >> >   \[  \mcom{\sin^2(x)+\cos^2(x)=1}{this is a trig. identity} \]
>>> >> > >> >
>>> >> > >> > I am compiling with XeLaTeX from a TeXlive distribution.
>>> >> > >> > Any ideas?
>>> >> > >> > Many thanks in advance,
>>> >> > >> > Dan
>>> >> > >
>>> >> > >
>>> >> > >
>>> >> > >
>>> >> > >
>>> >> > >>
>>> >> > >>
>>> >> > >> --------------------------------------------------
>>> >> > >> Subscriptions, Archive, and List information, etc.:
>>> >> > >>   http://tug.org/mailman/listinfo/xetex
>>> >> > >
>>> >> > >
>>> >> > >
>>> >> > > --------------------------------------------------
>>> >> > > Subscriptions, Archive, and List information, etc.:
>>> >> > >   http://tug.org/mailman/listinfo/xetex
>>> >>
>>> >>
>>> >>
>>> >> >
>>> >> >
>>> >> > --------------------------------------------------
>>> >> > Subscriptions, Archive, and List information, etc.:
>>> >> >   http://tug.org/mailman/listinfo/xetex
>>> >>
>>> >
>>> >
>>> > --------------------------------------------------
>>> > Subscriptions, Archive, and List information, etc.:
>>> >   http://tug.org/mailman/listinfo/xetex
>>>
>>>
>>>
>>> --------------------------------------------------
>>> Subscriptions, Archive, and List information, etc.:
>>>   http://tug.org/mailman/listinfo/xetex
>>
>>
>> --------------------------------------------------
>> Subscriptions, Archive, and List information, etc.:
>>   http://tug.org/mailman/listinfo/xetex

More information about the XeTeX mailing list