Subject: Re: A release-users list
To: John Hawkinson <jhawk@MIT.EDU>
From: Alistair Crooks <email@example.com>
Date: 05/18/2001 15:00:35
On Thu, May 17, 2001 at 10:07:14PM -0400, John Hawkinson wrote:
> | > > How would this be substantially different from the general purpose
> | > > netbsd-users list that we already have?
> | For people who want to discuss problems related to the release branch,
> | the general purpose netbsd-users list is far too much of a scatter-gun
> | approach.
> I don't understand why that is the case. Could you please be a little
> more clear?
There are a number of people who run -release code in production
environments, myself included. There can be all manner of questions
relating to the release branch itself - e.g. are the fixes for PCI
tlp driver going to be pulled up to the branch? What about support
for DP83815? Or, on 1.4, does anyone else see periodic hangs?
netbsd-users was, I understood, a general purpose "why are things
done this way?" kind of list, or "is there an LVM manager for NetBSD?"
kind of general-purpose list.
> Are you saying that people on netbsd-users have a reasonable
> expectation of not getting mail about the problem-of-the-week when the
> netbsd-1-5 branch fails to compile or work properly on their hardware,
> and they only want to hear about issues with 1.5 or 1.5.1 but never
> netbsd-1-5 of 15 May 2001?
> If that's the case, I do think that naming the list "release-users"
> would be poor. What you really want is "release-branch-users"
> or "-current-release-users" or somesuch, but those names are unwieldy.
I'm not advocating renaming the netbsd-users mailing list, that was your
What I am saying that would be useful, if perhaps low volume, is if
there were to be a list called release-users, analogous to
current-users, on which matters pertaining to the release branch of
NetBSD were discussed.
> Initially I agreed with Nathan that netbsd-users was adequate for
> release branch issues. But now I'm inclined to think that is untrue,
> and it's not appropriate to use that list for problem-of-the-moment
> issues with the release branch.
> I concur that having only one list is fine, though.
> I will also put forth the possibly irrelevent datum that such a list
> might be a disincentive for people to file PRs (as already happens
> with current-users).
I would imagine that people reporting problems with -current would be
told to file a PR. Likewise with -release. Often, however, people
are just trying to ascertain if it is a NetBSD problem, a problem
related to hardware, configuration error, or the kids have stuck
chocolate in the CDROM drive.
I find it mildly amusing that this whole discussion is taking place on
current-users - if netbsd-users was the appropriate forum for all this,
why hasn't the discussion been redirected there?