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

Re: FESCo Meeting Summary for 2007-06-28

On Sun, 2007-07-01 at 18:19 +0200, Axel Thimm wrote:
> On Sun, Jul 01, 2007 at 12:04:34PM -0400, Brian Pepple wrote:
> > On Sun, 2007-07-01 at 17:47 +0200, Axel Thimm wrote:  
> > > So the top question is: Always have mass rebuilds, no matter what, or
> > > delegate this decision to a releng/tools team?
> > 
> > That question is actually the item I have on the schedule to discuss
> > (which like said I was planning on doing around the end of July).
> OK, if there really is enough time to discuss it in August, that would
> be fine with me. I just think that Jesse for example would not want
> the rebuilds to happen shortly before freeze time and freeze is
> scheduled to happen at Aug 27th.
> Past experience shows that it would have to be started to be discussed
> in fesco now, otherwise it will just not make it in time.

> Still fesco should decide on a general policy independent of
> toolschain matters. This time perhaps gcc 4.2 could save the
> discussion, but in F9 time people will be arguing that F8 was only
> rebuilt because of gcc 4.2.

I have to agree with Axel here.  The decision whether or not to rebuild
is one that is contentious because it is pitting engineering *necessity*
against end-user *expectation* (with a tiny bit of trying to mitigate
Murphy thrown in to further stir the pot.)  So discussing this early to
get all the issues on the table and to gather any appropriate data to
make a decision is important.

However, there are two factors that call for delaying: 1) the FESCo
elections.  If this is truly going to run to lengthy discussion, then
it's probably best for the new FESCo to pick this topic up and run with
it. 2) We should be trying to hash this out on the list and record it on
a wiki page so FESCo can look at the wiki, ask specific questions, and
then come to a decision otherwise the topic will be discussed in the
meeting to even more length than it already has the potential for.

So... here's a wiki page:

I think we should discuss on the list pros and cons and *record them* on
this page so we have a clear picture of where the argument stands, and
what questions might need to be answered.  If all the arguments have
been expressed and organized on that page then it really does boil down
to FESCo weighing the pros and cons and simply making a decision.


Attachment: signature.asc
Description: This is a digitally signed message part

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