Subject: Re: Versioning?
To: Trouble Free RecepPFFFFFFFFFFFFFFFF <firstname.lastname@example.org>
From: David Brownlee <email@example.com>
Date: 12/07/1998 17:00:27
Could those who are happily pasting in paragraphs of replies to
the question please read
and either make any updates themselves, or email any suggestions
to someone who can (even me).
I'd do it myself, but I'm busy fixing other stuff..
-=- "Give me one last video, just dressed in black" -=-
On Mon, 7 Dec 1998, Greg Hudson wrote:
> You're gonna get about 80 replies to that question message, but I bet
> mine will be the only one with ASCII art.
> Here is a little diagram of NetBSD development from a little while
> back up to the current point:
> | |
> 1.2 1.3
> | |
> 1.2.1 1.3.1
> The horizontal line across the top is called "the mainline". If at
> any point you snapshot the mainline, you'll get a NetBSD with a
> version number with a letter at the end, like 1.3H. The "1.3" part
> signifies the last major or minor release, and letter changes indicate
> when kernel interfaces have changed.
> The vertical are release branches. Prior to a major or minor release,
> we create a branch in the repository. Development continues unabated
> on the mainline--the VM system gets replaced, BIND takes a major
> version upgrade, whatever--but on the release branch we mostly just
> fix bugs, hopefully in well understood and tested ways.
> 1.3.3 is a release off the 1.3 branch. It doesn't have UVM or wscons
> or RAIDframe or any of the other cool but not really ready changes
> we've done since the 1.3 release branch was created. It just has bug
> fixes, and new drivers, and stuff like that. (And a partition ID
> change in the i386 port and a new XFree release. I'm not too fond of
> the partition ID change, but then I don't contribute much, so nobody
> consulted me.)
> 1.4 will be a release based on a new release branch which doesn't
> exist yet, but will have all the cool stuff currently on the mainline
> (which you currently know of as 1.3I).
> Nobody knows what will trigger the fabled major version bump, because
> nobody's made that decision. There's some speculation that a
> preemptible kernel with SMP support is the critical change.