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

xetex on armel - disable or not?



Hi Rene, hi all,

(adding the d-t-m list, this is of interest for everyone)

On Mi, 29 Mai 2013, Rene Engelhard wrote:
> Or do it like LO and HarfBuzz and either disable it for all archs (as HarfBuzz does,

Disable it for all archs is an absolute no-go. This would make
Debian not ship XeTeX, which would make us the most ridiculous 
distribution.

> imho bad) just for armel/sparc (as LO does). No internal graphite2 lib
> involved in either of them. Of course only possible if TeXlice supported

That is an option.

> Do they run the testsuite? graphite2 *builds*, but the *tests*
> fail miserably. And given right now test failure is fatal on all archs...

We have a test suite for xetex, passed. That doesn't mean anything, of course.
I will contact the builder of armel-linux in TeX Live anmd ask
him if he see problems with xe(la)tex runs.

It might well be - (we had this already several times!) - that the 
broken functionality of lib xyz is not actually used. xetex normally
uses a very small part of the libs.

Thus, although graphite2 might be broken in the sense tha not 100% of
the functionality works, but might still work in the restricted
case of usage via harfbuzz.


So that means, that I have two options:
* either re-include libgraphite2, too, into the source again and
  get a working xetex binary (hopefully) on armel/sparc, too

* or disable xetex for armel/sparc
  (current status, since build deps are not present)


I honestly tend to the former option, and only *after* there are confirmed
reports that xetex really does not work on armel/sparc, disable
it on these archs.

What do others think?

Norbert

------------------------------------------------------------------------
PREINING, Norbert                               http://www.preining.info
JAIST, Japan                                 TeX Live & Debian Developer
DSA: 0x09C5B094   fp: 14DF 2E6C 0307 BE6D AD76  A9C0 D2BF 4AA3 09C5 B094
------------------------------------------------------------------------


Reply to: