Hey all,
I was catching up on scoop-checkins this evening, and was struck by
how many people were now able to commit to scoop. Not that having more
commiters is a bad thing, nor are any of the people with commit access
unqualified to do so; however, it does mean we have to watch out for the
'too many cooks' problem.
In the past, most patches were filtered through a patchmaster before
being committed. This ensured that at least one person other than the
developer had been able to get that patch to work properly, and that
there weren't any grossly obvious bugs. Now, it seems that the commiters
are checking in code that they, and only they (as near as I can tell,
anyways), have tested. This circumvents the safety-net that having a
patchmaster gave us.
My suggestion is to adopt the practice of never committing your own
patches. If you've done some development work you want checked in, you
create a patch and mail it to scoop-dev. Then someone else with commit
access can apply the patch, test it, and then commit it to CVS. This not
only allows for some more rigorous testing of new patches, but it also
gives everybody else in the scoop development community some insight
into what other people are working on, and a chance to voice any
opinions they have about it.
Anyways, that's just my 2c on the topic.
Mike
--
Michael Bain | One day I want to look through three hundred
mike at mostly-harmless.ca| thousand kilometres of space and say:
GPG-ID: 0xA30A5493 | "My isn't there a beautiful Earth out tonight!"
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
Url : http://lists.kuro5hin.org/pipermail/scoop-dev/attachments/20040102/d7ae9b1b/attachment.bin