Beginning last night, I reinstalled my laptop. Normally, if it were just to clean up some packages, I wouldn't do this - the aim was more to try out removing the disk encryption that I was using, and the effect has been quite dramatic. I worked out a while ago that the device-mapper encryption was slowing down disk access - with it gone, boot times are much shorter (and I don't have to type in a LUKS passphrase). GNOME loads a lot faster. Additionally, bash starts a lot quicker than it did. I suspect loading the bash completion routines takes quite a bit of disk access. Update: No, I just didn't have the bash-completion package installed. Bash is still slow. GNOME is still faster, though... for now. ...

So, a benchmark. tim@regulus:~$ time zsh -i -c exit real 0m0.064s user 0m0.048s sys 0m0.008s tim@regulus:~$ time bash -i -c exit exit real 0m0.540s user 0m0.436s sys 0m0.100s Both shells had their respective completion systems enabled. (Without them turned on, bash actually beats zsh... but the times are small enough that it doesn't matter.) These times are with a warm disk cache - the first time through both shells were slower. And the numbers stay roughly the same when repeating.

Yesterday evening, I finally found the patch for a bug in mono-addins that had been affecting f-spot extensions for a while - rebuilding the f-spot Debian package with no changes and reinstalling would cause the built-in extensions to disappear. In the end, the patch was just two lines long, and had been applied in mono-addins SVN (and in the copy of mono-addins that f-spot bundles). One less RC bug for lenny. ...

This weekend, I visited Derby, and updated the f-spot packaging in the Debian pkg-cli-apps repository. It's now at version 0.4.2, but this hasn't fixed the complicated extensions problems - they are Mono.Addins bugs, so we'll need to update libmono-addins0.2-cil to the version the f-spot devs claim fixes everything. Apart from that, I've been looking at my list of potential pet projects, and thinking about which of them I want to prioritize over the next year or so. I reckon I can organize them around the forthcoming Debian release - aiming to fix as many bugs as possible in time for lenny helps with rating the relative priority of things. It's probably not a coincidence that the more important bits of development I want to get done happen to tie in to Debian release goals. So I suppose now is not the time to start too many new projects. :) ...

FOSDEM was interesting, this year - I knew a lot more people than last time. Going to talks was a pain, because everywhere was so crowded; but the best bits are outside the talks, anyway. I eventually managed to sign my keys from the keysigning, and even caught up with some left over from last year. The next day I fell horribly ill - I was recovering for the whole of last weekend. ...

Counting off years at an arbitrary date on a calendar seems a bit meaningless - to many people I suppose it makes sense, because everyone gets together at New Year's Eve, and there's a big party to remember things by. With me it's different; for the last few New Years, things have been quiet. The big events in my life are not the times when I take a few days to relax, as pleasant as they are. Thinking about it, the biggest social events these days are software conferences. Sad, but true. I find myself reflecting on the year just gone. ...

It has been quite a while since I switched to using only Jabber (a.k.a. XMPP) for all my instant messaging. We use Jabber a lot at work, as well, and there are plans to set up an internal Jabber server eventually. (At the moment, everyone uses their personal Jabber IDs, which isn't ideal.) I've been having problems with jabber.org.uk lately - they seem unreliable at times, although they're going through a complicated migration to ejabberd which should eventually fix things. In the meantime, because I rely on it so heavily, I need it to work now. ...

Lately I've been having crazy ideas. Even more than usual, perhaps. I think the trick is to try and break the wildly unrealistic goals down into bite-size chunks that could actually work. Like, for instance, getting the project you want to hack on mirrored into your favourite repository format. In other news, I've finally got around to learning how to set up public git mirrors of cvs and svn repositories - there are plenty of tutorials about creating private repositories, but very few which deal with public ones in detail. ...

I haven't been writing so much recently - only one blog post for January, and I didn't even find the time to finish all the things I did in December. This is because I'm busy. Honest. So, before it gets too far into February, I'd better summarize things a little. Projects I worked on recently: f-spot - Near the end of December, I joined the Debian Mono packaging team, and helped clear up a few bugs in the photo management application I use. However, it turns out that f-spot could really use more help upstream - the number of forwarded Debian bugs is quite high, and there are some pretty nasty bugs outstanding. In fact, the upload we made has been synced to Ubuntu Hardy, so a bunch of users are going to run into horrible problems unless we fix them before the release. Oh dear. enscript - I finally started to write some code for GNU Enscript. Recently I fixed the build system so that it didn't fail `make distcheck', for the first time in over a decade. Win. I'm trying to work towards a maintenance release, rather like the new maintainer of a2ps has done. gnu-standards - the GNU maintainers' information document was under a license that didn't allow modifications, so it had to be packaged in Debian's non-free section. I asked about it, and it got relicensed. So now I'm trying to get the Debian package updated and moved back into `main'. postgresql-autodoc - this was a random package I picked up, adopted, and got uploaded sometime in December. I haven't heard any complaints yet, but then my fixes were mostly for typos. ...

I've had two cases recently where an old configuration file has been causing problems, and there's no clean upgrade strategy. My wireless card was made by Broadcom, and I use the newer b43 driver in Linux. However, the new mac80211 drivers introduce a second network interface (used internally by mac80211) but with the same MAC address as the normal interface. And because it gets created first, udev puts this master interface into the place where the real interface is meant to be, and the real one gets called 'wlan0_rename'. The fix was to look at /etc/udev/rules.d/z25_persistent-net.rules on Debian, and either add 'ATTRS{type}="1"' to the appropriate line, or delete the file, and let it get created on the next boot. ...