Rant and new stuff

It's a long time since the last post. In the meantime i have been on vacation for 10 days, and at least 1.5 weeks w/o a decent net connection.

I have continued to work on PKCS#11 support for GnuPG. I just have to say that I don't like how the code looks like. On the one hand, a set of mutually cooperating processes seems like a good idea. On the other, it's almost impossible to debug when writing own extensions, and the lack of documented error-handling protocol doesn't help either.

On the up-side, I have written a small C tutorial oriented towards OS development. It is aimed for students enrolled in the OS course taught at University of Oslo and University of Tromsø. You can find it here.



Prime sieves

Here I have described a relatively simple way to cut down the space requirements for computing the Sieve of Eratosthenes.



The plague of mailing lists

Open source is plagued by mailing lists. If you want to report a bug, how do you do it - well, in most cases it boils down to sending a mail to the project's mailing list. It gets worse: you have to subscribe before you're allowed to post anything. Subscribing/unsubscribing to a mailing list just to report a bug is a major hassle. Recently I've reported bugs for two different projects[1] directly to authors, and haven't received any reply. Maybe they received it, maybe got eaten by their spam filtering, maybe they just don't have time. Bottom line: I don't care. If they want to receive bug reports from their users, they'd better find some more convenient way. For example, take a look at how gentoo handles it. Reporting a bug to the gentoo project was very easy, painles and I didn't experience it as a hassle at all.

[1] One of the projects is Ingo Molnar's linux realtime preemption patch.



Switching to emacs

I've pulled emacs 22 from the CVS, compiled it and switched to it from Xemacs. Main reason is that I want to use the org-mode (included with emacs 22) to keep collection of notes. Although it is supposed to work also with xemacs, i haven't been able to byte-compile it; it fails with an obscure error message. I did report a bug to the author, and I'm waiting for an answer now. I have a feeling that xemacs is now lagging behind emacs development, and it has definetely less community support on freenode irc. Moreover, it seems that most packages are first developed for emacs, and only then (maybe) ported to xemacs.