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

Re: [users] Dropping the repotag



On Sun, 18 Mar 2007, Michael Schwendt wrote:

> On Sun, 18 Mar 2007 10:56:22 -0700, Greg Swallow wrote:
> 
> > Thorsten Leemhuis wrote:
> > > Thorsten Leemhuis schrieb:
> > >   
> > >> Please bring that up to the Fedora Packaging committee on
> > >> fedora-packaging redhat com, those guys are the ones that are
> > >> responsible for the Packaging Standard used in both EPEL and Fedora.
> > >>
> > >> The offer still stands: I can forward this discussion/your mail there if
> > >> you want. 
> > >>     
> > >
> > > Done now without waiting for an answer:
> > > https://www.redhat.com/archives/fedora-packaging/2007-March/msg00079.html
> > >
> > > People interested in this discussion please participate there!
> > >   
> > I already just joined this mailing list, now I need to join another?   
> > No thanks, I made my point clear here.  Maybe the packaging committee 
> > (5-10 people?) could have joined the EPEL list next time a problem like 
> > this comes up.
> > 
> > On the subject of getting a +1 or -1 from the community.  I would say a 
> > very, very small percentage of the community is here on this list, and 
> > even fewer on the packaging committee list, so if you have even 3 or 4 
> > people saying add the repotag, that's probably a few hundred thousand 
> > users.  Dag is trying to speak for most of them, as I bet that many use 
> > his repo, and I'm trying to speak for maybe 5,000 that use SME Server 
> > (an EL4 and soon to be EL5 deriitave).  So add 5000 +1's for me please 
> > on the other list ;-)
> 
> Why would a __user__ want a repotag? All the users want is that
> repositories "just work" when they are enabled and installed from, even if
> it is an unusual mix of repositories. A repotag does not contribute
> anything to achieving that. If the repotag is abused for RPM version
> comparison (as the least-significant part of %release), so packages from
> one repo upgrade packages from another repo, that would be really bad.
> Similarly bad is it when repositories compete with eachother in what they
> contain and when that leads to incompatibilities. A repotag only attempts
> at pushing some of the dirt under the carpet.

Hi Michael, welcome back, you haven't changed a bit.

If a user just wants RPM packages that work and doesn't care for the 
repotag, then there's no harm of having it, is there ? It doesn't break 
packages from 'just' working.

Secondly, there is currently *no* reason why the repotag would play a 
signifcant role in the version comparison. I mean come on, there is no 
relation between the release tags between repositories, so assuming that a 
repotag is there just to promote someones packages is STUPID. It's much 
easier to take a very big release-number to have certainty over version 
comparison.

Thirdly, if you don't like it, why deny someone else from having it ? It 
helps to debug certain situations _without_ asking for specifics. eg. if a 
package tagged .rf. clearly fails in a dependency resolution, you can 
simply forward the problem report to the correct project instead of first 
having to ask to give more information.

Yum doesn't where a package comes from and especially when the release-tag 
is identical (and this is not hypothetical) then you will have a hard time 
finding the cause.

You say that a repotag does not contribute to things just working. Is that 
because you prefer people just use one repository ? What then if CentOS 
does not join EPEL (which is likely) and has it's own extras repository ?
You know that CentOS has a much larger installbase than RHEL and may 
configure their repository by default. You understand that a single 
big repository in this light is less likely and that things may not 'just' 
work as you hope they may. In all those cases you want some information 
upfront to exclude possibilities.

Lastly, I will drop my repotag if the consensus is that a repotag is not 
desirable for EPEL. You probably do not care about that, but maybe you are 
not the audience I am packaging for. (You do not use my packages anyway 
and would stick with only EPEL)

Kind regards,
--   dag wieers,  dag wieers com,  http://dag.wieers.com/   --
[all I want is a warm bed and a kind word and unlimited power]


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