Bug#359909: subversion: FTBFS: Ruby test fails

[Daniel Schepler]

Le Mercredi 29 Mars 2006 19:46, vous avez écrit :
> Right, I'm not sure what to make of this.  You'll notice that 1.3.0-4
> includes an upstream patch expressly intended to make this one go away,
> and on my i386 machine, it seemed to work.  It is said to be is related
> to swig 1.3.28 - and indeed, it didn't occur when unstable had swig
> 1.3.25 and 1.3.27.
>
> I noticed several days ago (after the 1.3.0-4 upload) that powerpc was
> still hitting this case, but none of the other buildds were.  I was
> also told that ruby itself had some problems on powerpc, so I was
> unsure what to think of that.  (I haven't checked the 1.3.0-5 buildd
> logs yet, but if what you say is true, it disproves the theory about
> general ruby problems on powerpc.)
>
> Thanks for the report, I'll discuss it further with upstream,

Reading the report again, it seemed to me that it could be read as a 
chastisement; I apologize for this.  I was a bit exasperated after spending 
an hour or so yet again trying to build subversion and just getting another 
build failure, and then when I looked at the changelogs and saw something 
about an "intermittent build failure", I just snapped and lost it.  Again, 
I'm sorry for the tone of the bug report.

I'm willing to help in discovering what the issue is and in testing fixes.  
Something I forgot to mention, though: I was doing the builds in a "pbuildd" 
chroot with all packages being locally built versions, and I haven't yet 
checked whether the build failure happens with a vanilla pbuilder chroot 
using official packages from the archive.  I would be happy to do so if you 
want.
-- 
Daniel Schepler