F7 Release Discussion (was Re: Slightly OT: bad rap for Fedora, and realistic effects)

[Kevin Fenzi]

On Sun, 25 Feb 2007 21:31:46 +0100
mschwendt.tmp0701.nospam at arcor.de (Michael Schwendt) wrote:

> On Sun, 25 Feb 2007 11:21:33 -0700, Kevin Fenzi wrote:  

> > > Now it's getting interesting. Being a sponsor has never before
> > > implied that you have to fix orphans or take over packages when a
> > > sponsored person leaves the project or is AWOL. With such a
> > > requirement, the sponsorship system would be too burdensome and
> > > too much of a risk.  
> > 
> > ok. So, who should handle those things? FESCo? 
> > No one?  
> 
> Quite obviously, those are not the only two options. And hence the
> answer to the latter two questions is "no".  

Perhaps I wasn't being clear... 

What are your thoughts/opinions on how this should work? 
I would love to hear your input. 
>   
> > Would you like to update that wiki page with your thoughts on it?  
> 
> Such thoughts ought to come from the governing body of the project,
> since 1) the CLA and the contributor sponsorship system is their
> sphere of responsibility, 2) a successful sponsorship system makes it
> possible to scale, and 3) it can also happen that sponsors leave the
> project.  

Sure, agreed. However, such thoughts should also allow for input from
the community to come up with a plan that tries to meet all the needs
involved. Since I just started a draft, I would love to hear ideas from
the community before presenting it to FESCo to agree on. 

> > For much of this with the merged tree, we will be using the
> > procedure that former Core used to use?  
> 
> Is that a question?  

Not really. 

> Apart from that, we don't have "the merged tree" yet, so it is more
> interesting how to proceed until test3 and beyond.  

Agreed. 

> > Some of the E-V-R and broken deps problems have been fixed, but not
> > pushed out in devel since we are in test2 freeze, it looks like?   
> 
> Well, Core is frozen, Extras doesn't have any schedule ;-), but right,
> after the topic had come up on maintainers-list, Extras devel is sort
> of in a freeze, too, and pushed to only on demand.  

Right. 

> > Also, it doesn't look like the broken EVR report mails anyone, just
> > goes to the list. Could you change it to mail owners? I think some
> > people might be missing the problem.  
> 
> The code that would do that [without creating too much spam] is
> missing.  

ok. Fair enough I will bug maintainers directly about it. 
Is there any easy way to run that script locally for me to see what
current EVR issues are? (Hopefully using the needsign repo so I can not
worry about packages that will be pushed after the freeze is over). 

> > I think we could also figure out
> > the core owners for the core packages that have broken EVR, and mail
> > them too... or I can do so manually.   
> 
> What is needed is a clear word that those issues are considered bugs
> that need addressing and will be fixed. It is an unthankful task for
> any contributor to spend time on filing such issues, if the tickets
> stay open for a very long time (e.g. until the affected dist will see
> EOL).  

Yes. They are bugs. They must be fixed. 

> > On the subject of broken deps, I can look at assisting people with
> > those. Will go file bugs and see if there are any that are easy to
> > fix. Does the script just run 'repoclosure' ?  
> 
> A modified version, but running the original repoclosure from
> yum-utils should work to check packages and local builds, too.  

I mailed every maintainer with a broken dep from the last report with a
email telling them they needed a simple rebuild or that they had a more
serious problem and where to look to fix it. 

Many of them have fixed things. 

I will go through probibly tomorrow and find any that didn't get fixed
yet. 

kevin
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
URL: <http://listman.redhat.com/archives/fedora-devel-list/attachments/20070226/97d5c1ed/attachment.sig>