Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 03 Oct 2000 19:07:41 -0700
From:      Jordan Hubbard <jkh@winston.osd.bsdi.com>
To:        Kris Kennaway <kris@FreeBSD.ORG>
Cc:        Paul Richards <paul@originative.co.uk>, Christopher Masto <chris@netmonger.net>, Warner Losh <imp@village.org>, Joseph Scott <joseph.scott@owp.csus.edu>, Brian Somers <brian@FreeBSD.ORG>, cvs-committers@FreeBSD.ORG, cvs-all@FreeBSD.ORG, freebsd-security@FreeBSD.ORG
Subject:   Re: How long for -stable [ Re: cvs commit: src/usr.bin/finger finger.c ] 
Message-ID:  <85378.970625261@winston.osd.bsdi.com>
In-Reply-To: Message from Kris Kennaway <kris@FreeBSD.ORG>  of "Tue, 03 Oct 2000 17:34:14 PDT." <20001003173414.A58372@freefall.freebsd.org> 

next in thread | previous in thread | raw e-mail | index | archive | help
> I suggest that in practise it will be all but impossible for people to
> get active bugfix support for 3.x, and we'd probably be giving a false
> impression if we say otherwise.

Oh, I dunno, I think you may still be operating from the old "economic
assumptions" with these concerns.  We have a lot more people than we
used, after all, to and just the general buy-in to this thread so far
would tend to indicate that interest level in legacy support is higher
than it has been in recent memory.

We just need more volunteers who are committed to r.v-stable (-solid?
-stale?) and its upkeep, something which I'd expect to be somewhat
easier to recruit part-time help for than -current since the pace of
life is much more relaxed there and; you can get people to sign up
without having to make a truly major full-time commitment to
development.  Nobody also expects 50 commits a year on a -stale
branch, just 5 or 6 well-chosen ones, and even with part-time help,
cvs can easily provide the requisite diffs to someone with a very
specific point of focus (it's only you try to look at all of src
between a branch that life starts sucking rocks).

I also might not have said this 6 months ago when things like OpenSSL
and RSA were not integrated and enhancing security really often did
involve major infrastructural overhauls before you could address even
some very basic security issue.  We've gotten both far more functional
and more modular as of late there, however, and it would certainly be
my hope that future security fixes will generally be the easiest to
merge without requiring interface changes or tons of infrastructural
support to pull off (a buffer overflow or missing consistency check
being a fairly stand-alone change, neh?).

Or maybe I'm just smoking crack, but I honestly believe we have to at
least try this.  FreeBSD has really done some major growing up as OS
in the last 3 years alone and I think we simply have to make periodic
shifts in our MO to compensate or we're dogmeat.  I'm also fine with
your suggestion that we first try it for a few months to see how it
truly goes before announcing the policy change to the world, but we
should at least stop actively warning people away if we're going to
try this at all seriously.  We can be conservative and hold back on
the horn-blowing without dire effect to this "experiment", but we
can't have mixed messages going out without compromising it, either.

> Speaking for myself as part of the security team, I don't want to
> support 3.x for security fixes any more, since it's been just too damn

To me, this is only an argument that you need more help, not that the
fundamental idea of supporting security fixes for 3.x is somehow
unsound. :-) It seems like you essentially agree in your next two
paragraphs anyway, so can we now see a show of hands for "deputies"
who'd be willing to work on back-porting even just security
enhancements to 3.x (and, eventually, 4.x)?

- Jordan


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?85378.970625261>