MS RFC 7: MapServer CVS Commit Management

[Frank Warmerdam]

On 9/22/05, Paul Spencer <pspencer at dmsolutions.ca> wrote:
> would that mean that you (as in Frank) would have to actually file
> bugs before committing changes (like bug fixes on the stable branch ;) ?

Paul,

Yes, it would mean that.  Till now my fallback has been "I don't know of
any such rule."

> But seriously, should all commits be associated with a bug number?  I
> would assume trivial changes would not, but perhaps there should be
> some guidelines for trivial changes vs non-trivial?

I am not in favor of requiring bug numbers in the main dev branch.
If I come upon something that needs to be fixed up a bit (perhaps just
a spelling mistake in an error message) I don't want to have to file a
bug about it.  Generally I don't want to make the barrier-to-contribution
to high, at least in the dev branch.

Actually, we might want to apply the same rules about bug #'s in
the dev branch between release code freeze and the creation of the
branch.  A high threshold for change is appropriate then as well.

Come to think of it, I think another requirement of being a committer should
be that the person has to stay subscribed to mapserver-dev.

Best regards,
--
---------------------------------------+--------------------------------------
I set the clouds in motion - turn up   | Frank Warmerdam, warmerdam at pobox.com
light and sound - activate the windows | http://pobox.com/~warmerdam
and watch the world go round - Rush    | Geospatial Programmer for Rent