Date: Tue, 10 Jun 2003 15:39:11 -0700 (PDT) From: Doug Barton <DougB@FreeBSD.org> To: Mark Valentine <mark@valentine.me.uk> Cc: freebsd-chat@freebsd.org Subject: Re: FreeBSD Version Release numbers Message-ID: <20030610152545.S16934@12-234-22-23.pyvrag.nggov.pbz> In-Reply-To: <200306102031.h5AKVIxF060081@dotar.thuvia.org> References: <200306102031.h5AKVIxF060081@dotar.thuvia.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Tue, 10 Jun 2003, Mark Valentine wrote:
> > You're not only weird, you're seriously out of touch with general
> > software development models, and FreeBSD history.
>
> I'm very sorry to say you couldn't be more wrong on both counts. I only
> admit to the weird bit; I have this old-fashioned view that released software
> should work, and a long history of making it so.
Well, I don't disagree with that definition of "released software," but
I think what we're discussing is actually the definition of "works." The
problem we have with each new branch is that while we don't ever
purposely ship something that we KNOW is broken, because the developer
base is only a small percentage of actual FreeBSD users we can't
certify that we KNOW it's NOT broken for sufficiently large combinations
of hardware and user environments.
So, we choose to release stuff that's in the best shape we can make it
at the time, and (this is the novel bit), we actually TELL people,
"Beyond this point there be dragons," so that when one comes up and
flames their perky bottom, hopefully they'll be prepared.
I was one of the people who protested most loudly about the previous 5.0
release schedule, since I felt that there were too many dragons, with
too many sharp teeth. We delayed the release by over a year so that we
could file as many of those suckers down as we could. We did "Developer
Preview" releases, and tried everything we could figure out how to do in
order to widen our field of victi^W I mean beta testers, without drawing
in people who would be adversely affected by their own enthusiasm. But
at some point, you have to ship the thing, or you'll never make it
beyond the beta point.
Now that we have a 5.0 in the field, we've used the feedback we received
to greatly improve 5.1, and 5.2 is going to be even better than that.
Meanwhile, we're still telling people to be careful, and we will be
doing so for a while still.
As for the rest of the world, I'm glad to hear that you're working hard
to buck the trend, but you're well in the minority. The "real world"
ships software that they know is broken, and worry about fixing it
later. Everyone in the business knows that .0 releases suck rocks, and
no one adopts them unless they have a lot of time on their hands, or a
desperate need for new features. Saying, "I work hard to make the world
other than it is" doesn't change the way it is, however laudable your
pursuit.
Doug
--
This .signature sanitized for your protection
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20030610152545.S16934>