RCBW - week #11Here is this week squashes, by yours truly:
This week I've been both lazy and lucky, with a single NMU of db4.2 3 RC bugs were gone (well, truth to be said, I've discovered later on that db4.2 is not meant to be released with Squeeze, but well, stuff happens ...) and with some triaging a few more were gone too. The rest is nothing special. Also, during past week, Tim got back to my welcome message, adding some interesting considerations. Beside the issue of personal sponsorship, I fully agree that we used to lack pointers to NMU activity in NM. Now it is a bit better, though. As an AM, I pose the "please fix at least 2 RC bugs" question, and the usual outcome is the sponsoring of NMUs prepared by applicants. But surely we lack a more general "culture" of NMU, which is de facto more and more needed to actually deliver our releases. In fact, my RCBW initiative is not meant to make other people
feel bad (hello
Gunnar, we all love you and your work, and you should know that
|
). Rather, it is meant to convince people that fixing
RC bugs (at least some of them which are nevertheless a
huge slide of the total) is easier than what we might
think and help fighting the boredom. Not
sure it is being successful in its "marketing" facet, but it has
surely been fun for me thus far!
AFAIK there is no "bug fixing guide" for Debian, but it is not really something which is really Debian-specific either.
Regarding how to test a package, I usually don't care and install it on my "real machine" (unless the bug menaces to be really disruptive of couse) and then remove it a posteriori (yes, I keep a list of the package I will want to remove at the end of a debugging session). Alternatively, which is something you would need anyhow if you are a DD that eventually wants to NMU, you can install and test packages in a chroot. Helper tools in that respect are pbuilder/cowbuilder/schroot (look for packages with those names in the archive).
Thanks for your help!