Mon, 17 Dec 2007
HTML::Template in UTF-8
I use HTML::Template Perl module for some of my projects. There is a minor problem with it, that it does not allow the templates to be stored as UTF-8 encoded files. It treats data as bytes unconditionally, which is a problem with character-oriented processing in Perl.
If anybody is interested, I have created a patch which fixes this problem. The issue is tracked at CPAN and also at SourceForge.
Thu, 13 Dec 2007
How many HTML elements can you name in 5 minutes? Try it. I have scored 38, forgetting oh-so-many common ones...
Fri, 07 Dec 2007
Ekiga, PulseAudio, and D-Bus
It seems that more and more applications start using D-Bus for communication, so it is time to get ones feet wet with the D-Bus programming. Fedora 8 has PulseAudio enabled by default, which conflicts with Ekiga because of this ALSA bug (see also the related PulseAudio issue).
My solution is to suspend the PulseAudio server temporarily when the call is being made. This has an additional effect that you don't need to find where the music player window is hidden in order to mute it (so something similar would be usable even later, after the above two issues are fixed).
Anyway, here is my Ekiga PulseAudio suspender.
Written in Perl, needs
Net::DBus from CPAN,
Run it on background from your session startup script (or even from the
terminal inside your desktop session). It will mute your PulseAudio server when the call is being received (or when an outgoing
call is being made), and enables it back after the call is finished.
ekiga is not compiled with D-Bus support in
Fedora 8. New source
and x86_64 binary
packages are available, problem reported as
Comments are welcome.
Thu, 06 Dec 2007
Is Perl Being Maintained?
Another of the urgent works of the last month was rebuilding Perl and all of its modules for our production systems. The reason of it was that a security hole had been found in the Perl regular expression engine: CVE-2007-5116.
This hole has shown the sad state of Perl development: there is still not
a word about this hole at www.perl.org
nor CPAN, and new users are being informed
perl-5.8.8 is the latest and greatest.
However, from looking at the
spec file from
the Fedora Perl package,
it seems that in the last year they have applied even some patches
labeled as "fixes from the upstream". So there is some development
in the 5.8 branch, but it is apparently not public enough, and the Perl
developers do not even acknowledge serious security problems in their
Wed, 05 Dec 2007
Publish or Perish
Another deadline of the last month was a paper submission deadline to some conference. Well, I have decided that because I work for the company where academic titles matter, I need to restart my Ph.D. study.
I have submitted a paper about a (herein unspecified :-) problem, which I have tried to solve recently. The results are so far promising, so maybe they will accept the paper about these results as well.
When I have failed my previous attempt, it was mostly because my lack of time for doing a research in a topic that was not exactly related to my job. So I have decided that if I ever want to re-enter the Ph.D. study, the topic has to be something which I am directly involved in. My current plan is to have the sufficient number of publications first, and only then enroll to the Ph.D. studies, with the dissertation thesis topic already written. After defending it, submit the dissertation thesis itself.
Maybe I will have more luck this time.
Tue, 04 Dec 2007
I am sorry for the infrequent updates of this blog - I have been rather busy lately. One of my deadlines of the last month was the Japanese Language Proficiency Test exam, which was scheduled to the last Sunday.
There are only few exam sites in Europe, with the nearest one being in Budapest. Because the test was scheduled at 10:30 am, we have decided to go there a day before, and stay in a hostel over night. We have booked a room in Astoria City Hostel - it was relatively cheap with even breakfeast included. However, the beds were not very comfortable, and the hostel was quite noisy (esp. for those who want to sleep well before the exam).
The exam started an hour later than it was announced. The most annoying part of the whole process was the supervisor of our exam room: altought there were people from all over Central Europe there (even Serbia, Croatia, or Slovenia), he only spoke in Hungarian. The other supervisor was a woman which, when we asked "What that man said during his 10-minute speech?", said one or two English sentences, which we even couldn't understand, because the man was still speaking loudly :-(. Talk about rudeness.
As for the test itself, I think there is still a chance that I have passed JLPT 4. The Kanji and vocabulary part was quite easy (altough it would help if I could distinguish between the left and right side ;-). The listening was quite hard, as well as some grammar parts. An evening before I have tried to redo some exercises from our textbook, and I made the same errors like I did when we wrote those exercises before. Almost no improvement since then :-(. We will see some time in February or March, when the results will be announced. If I pass JLPT 4, I want definitely try JLPT 3 next year.