Bug#584350: refcard: FTBFS: xmlroff segfaults on japanese PDF
On Sun, Sep 19, 2010 at 05:52:14PM +0200, W. Martin Borgert wrote:
> On 2010-09-19 17:44, Simon Paillard wrote:
> > On Sun, Sep 19, 2010 at 05:19:04PM +0200, W. Martin Borgert wrote:
> > > On 2010-09-18 23:04, Simon Paillard wrote:
> > > > I still wonder which were your build conditions that allowed you to
> > > > * successfully * build the package.
> > >
> > > Interesting: Refcard builds fine on my system (sid). Building in
> > > a clean sid chroot lead to xmlroff crashes. Disabling the
> > > deprecated gnomeprint backend entirely (GPLIST variable in the
> > > refcard Makefile), made it work.
> >
> > That's why I noticed too
> > (http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=584350#12) but the
> > build result for japanese is not correct
> >
> > > The reason to use gnomeprint was a bug in the default cairo backend.
> > > But the original bug is still there, I re-checked #492597 and I can
> > > reproduce the crash in my chroot. Anyway, I will upload refcard now
> > > and we will see whether xmlroff crashes on other machines...
> >
> > But the gnomeprint backend is deprecated #542564
>
> Yes, that's I wrote 14 lines before :~) Using deprecated stuff
> was necessary at that time, because the other backend used to
> crash (and still crashes in some cases).
Oops, I overlooked the word :)
> Btw. I committed my changes to SVN. Would you mind to build and
> upload? I'm busy - again - and if you to the build, it is one
> more check, that my changes are not completely bullshit :~) TIA!
Sure, once checked that the japanese and generally CJK build looks ok.
--
Simon Paillard
Reply to: