Date: Fri, 11 Dec 1998 09:25:26 +1030 From: Greg Lehey <grog@lemis.com> To: Poul-Henning Kamp <phk@critter.freebsd.dk>, "Jordan K. Hubbard" <jkh@zippy.cdrom.com> Cc: Joseph Koshy <jkoshy@FreeBSD.ORG>, committers@hub.freebsd.org, vanmaren@fast.cs.utah.edu Subject: Re: Swat teams (was: problem reports) Message-ID: <19981211092526.B446@freebie.lemis.com> In-Reply-To: <1238.913286921@critter.freebsd.dk>; from Poul-Henning Kamp on Thu, Dec 10, 1998 at 11:48:41AM %2B0100 References: <29693.913281555@zippy.cdrom.com> <1238.913286921@critter.freebsd.dk>
next in thread | previous in thread | raw e-mail | index | archive | help
On Thursday, 10 December 1998 at 11:48:41 +0100, Poul-Henning Kamp wrote:
> In message <29693.913281555@zippy.cdrom.com>, "Jordan K. Hubbard" writes:
>>> The big thing that many people forget here, though (and it applies at
>>> least as much to commercial support organiziations) is that the real
>>> purpose of a PR is to draw attention to a problem. The fact that the
>>
>> The problem with this line of thinking is that when there are over
>> 1500 unclosed PRs, many of which are so old that they'll more than
>> likely never be looked at again, they're not drawing attention to much
>> more than the fact that there are over 1500 unclosed PRs.
>
> Let me chime in here, now that I've woken up.
>
> Spending about one hour a day, I was able to keep the numbers of PRs
> almost steady. This was possible by pretty ruthlessly closing the
> bogus new ones, and by sweeping through the old ones. I could on
> average close 50% of the old ones using a criteria of:
> If version is very old -> close
> If unreproducible -> close
> If suspected hardware -> close
> If mail bounces -> close.
You can do better than that! You could close nearly 100% of the PRs
using the following critieria:
If version is very old -> close
If unreproducible -> close
If suspected hardware -> close
If mail bounces -> close
If can't be bothered -> close.
The real problem is not the number of open PRs, it's the number of
real problems. If *you* can't reproduce the problem (after how much
effort?), it doesn't mean the problem doesn't exist. If mail bounces,
it makes it more difficult for you, but it doesn't have *any* affect
on how serious the problem is.
In a previous life I went through nearly a decade of this thinking:
people look at the queue instead of the problem. We have a number of
problems, particularly the kind that are difficult to transport ("if
unreproducible -> close"). I don't have too many good solutions for
how to transport them, but at least I'm not trying to pretend they
don't exist.
Greg
--
See complete headers for address, home page and phone numbers
finger grog@lemis.com for PGP public key
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe cvs-all" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19981211092526.B446>