Re: PCI DMA lockups in 3.2 (3.3 maybe?)

[Eivind Eklund]

On Mon, Dec 06, 1999 at 01:43:33PM -0800, Matthew Dillon wrote:
>     If we enforce a stabilizing period between .0 and .1 and branch at .1
>     rather then at .0, this combined with the 12 month schedule should result
>     in pretty damn good releases.
> 
>     If we just do the 12 month schedule, I don't think it will produce as
>     good a result.

I'd just like to point out how I've understood what NetBSD is doing here:

1. Put down the branch
2. Ask all developers to switch to that branch, and drop using
   -current for stabilizing changes and other changes that should go
   into the release branch
3. After a suitable period of this (when the branch is considered
   ready to 'go golden'), ask all the developers to switch back to
   -current.
4. Merge the changes from the branch back to -current.

This seems like a good way to kick-start a branch; you get a while
when there is focus on stabilizing among developers that are actually
running the branch, while there still is somewhere to stick the
'dangerous' changes.

Eivind.


To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-hackers" in the body of the message