Please see my responses to some of these points on a posting I've made
in a followup to "Improving the handling of PR:s", initially on
freebsd-current@ but now Cc:ed to freebsd-bugbusters@.

On Thu, Jan 10, 2008 at 09:16:54PM +0000, Robert Watson wrote:
quoted text
> I know that Mark Linimon has done quite a bit of analysis of the state of 
> the PRs, especially as to which ones stay open vs. which ones get closed, 
> and may be able to offer some insight.  I have a vague recollection that 
> last time around, he reported essentially linear growth in open kernel bug 
> reports, and essentially stable ports PRs, but I've not really seen stats 
> on how bug reports against the base system get closed.  For example, I'm 
> not sure we make a "fixed" vs "closed" distinction, which we'd need in 
> order to do a good analysis.

No, we don't make that distinction.  Also, we've lost the software that
was showing us the graphs of PR count per category over time; the
committer who was maintaining it had not had time to work on FreeBSD in
a long while and requested his commit bit be returned.  Unfortunately we
went ahead and cleared out his account, which is where the code that ran
that stuff lived.  (If I had known about it, I would have grabbed it.)

My recollection, last I looked, is there are large swings in the ports
PRs, which happen to coincide exactly with ports freezes :-)  The kern
and bin PRs increase linearly until someone hard-headed enough plows
through and knocks a couple of hundred out (hi Kip, Warner :-) )  The
curves have flattened out a bit in the past year but we're not close
to steady-state there.  kern is probably > 30% of the count; bin, > 20%,
ports, 20-30%, depending on how open the tree is for commits.

The other categories aren't as worrisome as the first 2, and the ports
stuff is affected by having an auto-assigner and to some extent portsmon
to hang off of them.  So really, over 50% of our problem is kern/bin
(kern includes drivers and libraries, fwiw.)

Again, as I say in that other post, I intend to task-switch onto
thinking about what we can do about these situations.  We really need
to translate "I'd like to help" into "here's what you can do".

mcl
_______________________________________________
freebsd-current@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscribe@freebsd.org"