Message-Id: <199512191839.NAA24430@jekyll.piermont.com>
To: Jason Thorpe <thorpej@nas.nasa.gov>
Cc: netbsd-developers@netbsd.org
Subject: Re: the state of PRs. 
In-Reply-To: Your message of "Tue, 19 Dec 1995 10:13:58 PST."
             <199512191813.KAA00742@lestat.nas.nasa.gov> 
Reply-To: perry@piermont.com
Date: Tue, 19 Dec 1995 13:39:56 -0500
From: "Perry E. Metzger" <perry@piermont.com>
Sender: owner-netbsd-developers@NetBSD.ORG
Precedence: first-class


Jason Thorpe writes:
> I don't really see how the added pressure of actively advertising
> the bug list as an incentive to work harder is really going to
> accomplish anything.  That's just not the way one motivates
> volunteers.

I think the problem is that many of us forget about things we've left
in the PR database. This is a way to jog the memory.

>  > The fact that we have stuff outstanding for years, though, is a bit
>  > embarassing...
> 
> Yup, it is.  Though, it sort of begs the question: "How many were fixed 
> and just not closed?"

Dunno. I suspect a lot of them. Lots of them are also for single bugs.

For enhancements and the like, we should decide, very quickly, either
to accept the enhancement and put it in, or to close the PR having
declined the enhancement. I don't think we generally do that, though.

BTW, I'll point out that we have only one pr left from 1993, and it
would be neat if it didn't survive until 1996. Its comments indicate
that it should now be very easy to fix if it hasn't, in fact, already
become closeable.

> Maybe a good start would be to run though the "how to reproduce"
> sections and see how many of the Really Old ones still have any
> meaning?

Probably a good idea, but who is to do the work? *Someone* has to be
responsible if the things are actually going to get done.

Here is a strawman to get rid of a whole year's worth of PRs: maybe we
ought to parcel out the very oldest pr's (say, the 100 oldest) to the
members of the developers list. That ends up being two (or three at
most) PRs each -- hardly an ominous burden or more than a single
individual can track. By coincidence, if we cleared the 100 oldest
PRs, we would have brought ourselves up to the PRs from January, 1995
-- and thus would only have outstanding problems one year old to deal
with.

If people think this is a good idea I'll happily write a script to
give everyone on developers a couple of PR numbers to examine in
detail.

Perry