Mike Frysinger posted on Tue, 14 Dec 2010 22:22:14 -0500 as excerpted:

> On Tuesday, December 14, 2010 20:54:45 Jeroen Roovers wrote:
>> On Tue, 14 Dec 2010 14:00:02 +0200 (EET) Alex Alexander wrote:
>> > Our bug queue has 118 bugs!
>> 
>> I am starting to wonder if this is helping. It looks like everyone now
>> attempts to keep it <100 on a daily basis, but not to far <100, which
>> means a lot of old, difficult, nasty bug reports are left unattended.
>> Still, I got it down to about two dozen now.
> 
> i think people will aim for whatever arbitrary limit is picked.  so
> raising it to say 200 wont help either.

Agreed.

Which begs the question[1], why not take the opportunity to lower it?

The notices have been demonstrated to be able to keep it to ~100 bugs, but 
IIRC that was a rather arbitrarily picked number, according to the 
previous discussion thread.

Now that the number is/was ~24, what about lowering that to say 50 before 
it hits that, and then by say one a day to some number deemed not to let 
bugs languish, probably not lower than 2-3 average days worth, however, 
given the warning period of once per day.  (I've no idea what the filings 
per day is, 50 obviously assumes <25.)

If the script could be improved to give the date and bug number (maybe 
with title/summary?) of the oldest unassigned one as well, and trigger a 
warning if it were more than, say, three days old, as well as by queue 
length, that might be nice, too.  However I recognize that's easy to say 
given I'm not coding it.

---
[1] Yeah, I know.  I'm using the term in the rhetorical personification 
sense, not the historical/legal sense.

-- 
Duncan - List replies preferred.   No HTML msgs.
"Every nonfree program has a lord, a master --
and if you use the program, he is your master."  Richard Stallman


Reply via email to