MOTU application for Michael Casadevall (NCommander/sonicmctails)

[Michael Casadevall]

To the MOTU council:

Wiki Page: http://wiki.ubuntu.com/MichaelCasadevall
LP Profile: http://launchpad.net/~sonicmctails
Primary Sponsors: Scott Kitterman, Steven Kowailk, Sebastien Bacher,
Jon Riddell, John Dong (Backports),  Luca Falavigna (SRU/MOTU)
Primary Team Memberships: Xubuntu (Developer), Kubuntu (Developer),
Backports (Tester), SRU (PowerPC tester), Ports (PowerPC)

I'm writing today apply for MOTU status in Ubuntu. I've been highly
active during the intrepid release cycle, working to help resolve and
fix many bugs, as well as build stronger ties with Debian by merging
patches from Ubuntu when appropriate. I have been urged by my sponsors
to apply for some time, and finally worked up the nerve to face the
possibility of rejection.

Within the intrepid cycle and the time I've been around, I've working
on doing a partial rewrite of REVU with RainCT to the point of vastly
increasing the usability, and improving the look and feel of the GUI
by making it easier to manage and navigate. Of all my contributions
this cycle, this one has the highest visibility, and greatest affect
on later Ubuntu releases as REVU is a critical tool to reviewing new
packages, and by improving it, helping all Ubuntu Developers. I've
also worked on creating ubuntu-cruftbusters, a ready made group of
volunteers t help work on transitioning packages and other related
tasks. I hope to also put this group to better use following UDS
Jaunty.

I applied for a MOTU advanced sponsor, and was assigned Sebastien
Bacher, who worked with me in packaging various GNOME releases, as
well as taught me more advanced (or evil) packaging techniques such as
multiple builds from a CDBS based package, and the specifics of
library building and transitioning. Through this mentorship, I was
able to fill in many gaps in my packaging knowledge I came from Debian
with.

My next major project beside fixing FTBFS in Universe was helping get
KDE3 and 4 fully portable across all architectures both in Hardy (via
SRU/Backports which lead to a rather amusing moment where the same
package was uploaded three times to proposed, backports, and intrepid
to clear FTBFS), making KDE now available on all architectures aside
from HPPA. I've also worked on debugging and helping with other issues
within KDE, including our recent language pack issues. By helping port
KDE to all architectures earned me Kubuntu membership.

Following this, I helped work on packaging Xfce 4.5.80 for Ubuntu,
which helped earn me Xubuntu Developer status; I continue to work on
merging Xfce beta packages from Debian, and have worked to reduce our
diffs from Debian greatly (pkg-xfce and Xubuntu now have an excellent
working relationship, and our diffs are now down to simply Xubuntu
specific changes, with all our bugfixes mutually shared). This work
has inspired me to work to help reduce the diffs in general by
submitting them back to Debian, and NMUing packages in cases where it
is needed so that the general workload of having to merge packages is
reduced.

During all of this, I've helped three transitions clamav,
libbluetooth3, and gnat-4.2 (via SRU). In both clamav, and
libbluetooth3, I've worked to resolve FTBFS via porting and
re-libtoolization and gnat-4.2 via my knowledge of Ada programming;
I'm pleased to announce due to the efforts of both myself and Luca,
this transition proceeded smoothly through updates and have fixed ada
packages in Hardy. I handle managing large changesets and helping to
coordinate and work a part of a transition, a valuable skill to have
in the everchanging world of ABI and API bumps.

Of all this, there are six uploads specifically which I feel deserve
special note. The first one is mono
(https://edge.launchpad.net/ubuntu/+source/mono/1.9.1+dfsg-2ubuntu2)
which solved a very nasty runtime fault that was the cause of quite a
few mono application freezes with F-Spot and a few FTBFS. I rewrote
the patch and got it uploaded and the issues vanished. Since this bug
in mono only manifested itself sporadically, and the backtraces were
useless, I feel quite proud I was able to isolate the Ubuntu patch
causing the failures.

Second is codeblocks. This is a new package upload, and was a rather
longstanding needs-packaging bug. It had a large list of dependencies,
a few previous attempts, and a very difficult problem in that the
package can not be forced via configure to comply with Linux FHS,
forcing the authoring of patches to resolve this behavior. I feel
quite proud I managed to bring this to intrepid, and I hope I can
bring it to hardy via backports relatively soon as well.

Third up courier
(https://edge.launchpad.net/ubuntu/+source/courier/0.60.0-1ubuntu1),
which was suffering from a FTBFS due to libtool. I helped incorporate
the necessary libtool fixes to get the package building again, and
have been from that point forward deemed an autotools wizard, for
better or worse. The skills though on manipulating autotools has been
extremely useful in resolving other FTBFS throughout the archive.

Fourth is subversion in backports
(https://edge.launchpad.net/ubuntu/hardy/+source/subversion/1.5.1dfsg1-1ubuntu2~hardy2),
which had an issue due to gnutls vs. openssl. By reading through
Debian bug reports, and the information posted in the bug, with Scott
Kitterman's assistance, was able to release a fix and resolve the
entire issue relatively straightforwardly.

Fifth is samba4, which was badly broken due to issues with the inotify
API in glibc and linux-headers not agreeing. Some careful debugging,
and a patch later, this FTBFS was cleared (at the cost of several
points of sanity). Once the samba4 FTBFS was cleared, this opened the
door for openchange, which in general has been considered a very good
thing .

Sixth and finally is linux-lpia 2.6.27-4.8
(https://edge.launchpad.net/ubuntu/+source/linux-lpia). This was my
first adventure into the world of kernel hacking, and boy was it fun.
The main issue here is that most of the metapackages were not being
built since they were at the time were arch all, and linux-lpia
obviously doesn't build on pure x86. Since linux-lpia is lpia only,
these were changed to lpia so they could be built on the lpia
authobuilders, but to make this work required major changes to the
rules file in git to build the packages sanely with the new package
names

I look forward to working in jaunty in helping get patches back in
Debian from Ubuntu, and attending at UDS for Jaunty to help map out
our future of Ubuntu. I personally would like to see better support
for ports, especially for studio. There are a great many PowerMac G4
and G5s that are used for studio work that Apple has decided to
support, and I would like to help bring studio to these platforms (the
main problem with studio is a lack of a real-time kernel for PowerPC,
the patches exist, they simply need to migrate into the linux-ports
kernel).

I also would like to greatly work towards removing cruft from the
archive, such as gtk1.2, and improving both Xfce and KDE. One of my
largest projects in terms of general usability with Xubuntu and Xfce
specifically is making Xfce fully syncable from Debian with the
expection of the few Ubuntu specific patches, as well as helping any
bug

With great regards,
Michael Casadevall