Broken upgrade paths in F7

[Ralf Corsepius]

On Fri, 2007-05-18 at 09:48 -0400, Jesse Keating wrote:
> On Friday 18 May 2007 09:40:06 Ralf Corsepius wrote:
> > > There is a README in the package, not great, but a start.
> >
> > May-be helpful when being familiar with koji, but right now, what's
> > really urgently missing is a "Getting started with koji and Fedora"
> > and "Getting started with koji for plague users".
> 
> A couple wiki pages were posted for FAQ like things,
Yes, in recent days the situation has somewhat improved ...

> > Also, IMO, --nowait should be made default. Not using it is really
> > nagging (I only knew about it, because somebody else already had
> > complained and you had mentioned it some time last week).
> 
> The reason it isn't made default is that there is no converse of it, 
IMO: bug.

> IE --wait.  So by making it not default it gives users the opportunity to 
> either use it if they like, or add a --nowait flag to disable it.  One can't 
> have it disabled by default and force a way to have it actually wait.


> > > > perl-modules built flawlessly in a local fc7-mock ("make mock"), but
> > > > failed in koji ("make build").
> > >
> > > That's not helpful. You have logs,
> >
> > No, because koji doesn't email build results (another regression in
> > comparison to plague - These emails are very useful because they can be
> > archived.).
> 
> Hrm, Koji emails the owner of the package being built by default.

Does it? ... hmm, I have one koji mail in my local buildsys archive, but
I can't find any on perl-File-Copy-Recursive?

I recall me having dug through koji.fedoraproject.org (Even this isn't
mentioned anywhere inside of the koji package) to find out what might
have happened to the build.

> > This is the koji entry documenting the breakdown
> > http://koji.fedoraproject.org/koji/buildinfo?buildID=6550
> >
> > A "make mock" in "devel" on FC6 issued ca. 10-20 mins before the
> > breakdown (with the exact sources as above) did not exhibit anything.
> >
> > >  why don't they build right?
> >
> > The origin was a missing "BR: perl(..." (fixed in *-2), an oversight on
> > my part mine, I didn't notice in advance, because "make mock" did
> > exhibit the issue.
> >
> > The only explanation I have is an inconsistency between "koji's repo"
> > and "rawhide" probably related to the perl package.
> 
> Yes, that can happen.  Builds happen during the day which alter the buildroot 
> and drift from the previous night's rawhide.
:(

BTW: Invoking http://koji.fedoraproject.org/koji/buildinfo 
just returned this:

Mod_python error: "PythonHandler mod_python.publisher"

Traceback (most recent call last):

  File "/usr/lib/python2.4/site-packages/mod_python/apache.py", line 299, in HandlerDispatch
    result = object(req)

  File "/usr/lib/python2.4/site-packages/mod_python/publisher.py", line 213, in handler
    published = publish_object(req, object)

  File "/usr/lib/python2.4/site-packages/mod_python/publisher.py", line 412, in publish_object
    return publish_object(req,util.apply_fs_data(object, req.form, req=req))

  File "/usr/lib/python2.4/site-packages/mod_python/util.py", line 439, in apply_fs_data
    return object(**args)

TypeError: buildinfo() takes exactly 2 non-keyword arguments (1 given)


Ralf