Notes on Yum Changes

Rui Miguel Seabra rms at 1407.org
Tue Sep 7 07:21:23 UTC 2004


On Mon, 2004-09-06 at 19:45 -0400, seth vidal wrote:
> > Well, I thought there was a process for package submission, review and
> > what not. Of course accidents happen. But this is no excuse, or is a
> > "meta packager" like yum not better than plain rpm in itself?
>
> > As far as I can see, I have to solve about as many problems as I did
> > without a metapackager (--obsoletes, --exclude=..., etc...). The job of
> > a meta packager is making things easier.
> 
> it's not the job of the metapackagers to let potentially serious
> problems go unseen. That's ridiculous.

What do you think of a security update that doesn't update because
there's a bug in an unrelated package?

How "potentially serious" a problem would that be?

The job of a metapackager is making updating easier, not harder just as
hard as if it wasn't there.

> That's like saying: My kernel shouldn't scream at me about hard drive
> failures, it should quietly let them happen.

I fail to see a resemblance. I'm saying to not touch with the
problematic packages but try to keep on with the rest, if possible.

There would be a resemblance if I was saying to install even if it
screamed of a problem...

> if there is a potential conflict that can't really be resolved, prompt
> the user.

Yum doesn't prompt. Barfs. You could be right if it did, but
_it_doesn't_.

You can handle it. Fine.
I can handle it. Fine.
But is it satisfying the purpose?

> > Anyway, loops are not that hard to find, just mark where you've been
> > previously (prolog 101) and handle apropriately (ie, until repo fixes
> > it, I can't do anything that touches foo, bar or baz, may I proceed with
> > the rest?).
> 
> okay then define the procedure. Set up the standard for handling mutual
> obsoleting packages and updates.
> 
> you write up the process for what has to occur and get EVERYONE to agree
> on it and I'll write the code.

In a very highlevel way, I already did, wouldn't you say so?

You simple mark where you've been so far (and where you have had
problems), and check for those marks at each package.

A mutual obsolete is just another kind of loop.

There's no need to deny it from the start like you seem to be doing.

Rui

-- 
+ No matter how much you do, you never do enough -- unknown
+ Whatever you do will be insignificant,
| but it is very important that you do it -- Gandhi
+ So let's do it...?

Please AVOID sending me WORD, EXCEL or POWERPOINT attachments.
See http://www.fsf.org/philosophy/no-word-attachments.html
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
URL: <http://listman.redhat.com/archives/fedora-devel-list/attachments/20040907/aa7a307a/attachment.sig>


More information about the fedora-devel-list mailing list