Michael Scherer wrote:
> Le Jeudi 09 Mars 2006 21:18, Anssi Hannula a écrit :
>
>>Anssi Hannula wrote:
>>
>>>Also, do we agree that *all* packages should be backported to latest
>>>stable (e.g. 2006.0)? As the chroots are now offline, many rpms do not
>>>get rebuilt for anything else than cooker-i586. I do have scripts that
>>>allow easy rebuilding of packages for other targets, but as I'm not able
>>>to thouroughly test every package, I'm hesitant to do uploads unless you
>>>confirm they should all be backported. If they do not, we also need a
>>>way to inform the bots on which targets the srpm should be build.
>>
>>Comments, please. This question is critical if you still want to get
>>2006.0 backports of PLF packages (or want not to).
>
>
> I am against version backport in the main repository :
>
> - some people prefer to have fixed and stable release, for deployment
> purpose,
> because thing backported have less testing ( even they have sufficient
> testing, but you can never be sure ), or for whatever reason they have.
>
> I prefer to not update anything on some workstations, especially if doing a
> roolback is not possible because the old package was removed.
True.
> - it introduces complexity in testing and building, and in the spec file.
.spec files should still have support for backporting, if user wants to
do so.
> - it is not consistant, because some people do not backport ( because they
> cannot test, do not want, cannot ), some backport for all release, some
> backport for the latest one, etc.
That is the current situation (at least before ryu chroot failure).
> However, I know that people want backports, usually gui stuff ( xchat, for
> whatever reason, kde, etc ).
> I will not discuss the reason as I do backport myself some applications when
> I
> need to ( server stuff usually )
> I do it for myself because I know
> 1) I can roll back
> 2) I can debug as I do it for stuff I know
> 3) it will be tested by me
> 4) I do not take responsability of breaking someone else system
>
> So yes, we should offer a way for people to have backport if the maintener
> want to offer, but I would prefer to also offer a way to not have backport.
>
> Mandriva suffer from the same problem, and we should try to have a common
> policy on this subject.
>
> I suggest a separate media, or a separate version, like ${VERSION}_updated/
> on
> the mirrors.
These points are all valid. IMHO the optimal situation (for both
Mandriva and PLF) would be that all pkgs (unless the maintainer doesn't
want to for some good reason) are backported, but by default no
backported packages are installed. Then the user could choose to e.g.
always install new xchat backports. (of course this applies only to pkgs
that no external pkgs require, e.g. firefox shouldn't be backported)
Unfortunately I have no idea how that could be implemented without
massive changes to urpmi.
--
Anssi Hannula
_______________________________________________
PLF-discuss mailing list
[email protected]
https://www.zarb.org/mailman/listinfo/plf-discuss
