Re: How to add files that upstream does not install?
Norbert Preining <preining@logic.at> wrote:
> On Die, 10 Apr 2007, Frank Küster wrote:
>> >> Why do we need this indirection step? Wouldn't it be possible to let
>> >
>> > We don't need it for libkpathsea, only for the tpms.
>>
>> But only to save us from passing an ugly
>>
>> echo $(dpkg-architecture -qDEB_HOST_ARCH_CPU)-$(dpkg-architecture -qDEB_HOST_ARCH_OS)
>
> No. Also because the tpms list the files in bin/$specialarch/... so we
> *have* to move them there! Or we have to make extra code for *all*
> binaries in tpm2deb-bin.pl.
Ah, I see.
>> about uploading? If we upload to unstable without building a kpathsea
>> package, we get our packages to be used and tested sooner. On the other
>> hand, if any urgent bugs show up, we cannot solve them by uploading
>> texlive-bin when it has already a version in NEW - this means that we
>> will always stall our development by including libkpathsea. So why not
>> do it now?
>
> Hmm, I would say upload to unstable now. I am running 2007 since quite
> some time without any big problems. And some(?) others, too.
I run it since a week, and I didn't have any problems (except one quite
annoying one, but that was related to a change in layout in koma-script
and has nothing to do with Debian packaging).
> So I guess
> that an upload to unstable NOW would not create *emergency* problems.
Probably not (or rather, the problems would be solved by uploading other
packages, namely those that build-depend on tetex and fail with the new
dummy packages).
> Furthermore, we *COULD* do the following (I believe!):
> - upload 2007-2 to unstable
> - upload 2007-3~exp1 with libkpathsea to experimental
>
> In case there are problems with -2, we upload -3 without libkpathsea to
> unstable.
>
> A soon as 3~exp1 is through NEW we upload 3 to unstable.
>
> Comments?
Yes, that's a good idea. One thing more should be done: I think d-d-a
should get a mail, it's quite an important change, although behind the
scenes.
Regards, Frank
--
Dr. Frank Küster
Single Molecule Spectroscopy, Protein Folding @ Inst. f. Biochemie, Univ. Zürich
Debian Developer (teTeX/TeXLive)
Reply to: