On Sat, 2009-10-24 at 10:09 +0200, Øyvind Harboe wrote:
> Should we try to set a 0.3 schedule now that we're all
> switched to git?

We had been talking about 0.3 from before the transition off SVN.
With GIT, pushing commits to the repository should happen only after
full review, so recent feature changes may push the schedule back.

All developers can commit on branches locally, so the community can be
given ample time to review patches before they get pushed.  We might
consider enforcing this review step on SF.net, by checking for multiple
Signed-off-by lines in the pre-commit hook (which could also prevent
pushing whitespace at ends of lines).  

I would even be happier knowing my patches will subjected to this extra
scrutiny, as any later blame can be shared.  ;)

> 2-3 weeks out?

That depends on how fast we can verify there are no regressions in the
documentation and scripts from recent commits. *cough*

I have a long branch full of changes that continue my crusade to improve
command argument parsing, but I am saving them for the 0.4.0-dev branch.
That should be branched on SF.net only after 0.3.0 is out the door, so
we should start an 0.3.0-rc-dev series almost immediately.

Posting release candidate packages should help attract more attention
for testing, and no patches should go in during the RC period without
full peer review -- even if we continue to allow ourselves to commit
directly to the repository during that time.

I'll start another thread to discuss other 0.3.0-related details that I
have been tracking, along with the current drafts of my release scripts.

Cheers,

Zach

_______________________________________________
Openocd-development mailing list
[email protected]
https://lists.berlios.de/mailman/listinfo/openocd-development

Reply via email to