10:10:48 # Life Open Source Conference Tokyo 2006 Fall, Debian booth. Yesterday Tokyo Area Debian Meeting held a Debian Booth. Gotom, Iwamatsu, Yamane, Komuro, Eto, Kobayashi and I attended. We had Yoshinoya beef bowl for lunch. We had several Debian/SuperH evaluation boards and a Zaurus and several Debian/MacBook on display. Yamane, Kobayashi, and I presented in a BOF session, where we talked about Translation, IceWeasel update, and about what we do in our Monthly Meeting.
13:37:09 # Life Summarising this month's Tokyo Debian meeting. Yesterday we had another Tokyo Debian meeting. It is a monthly meeting held in Tokyo, Japan. approx. 23 people attended. Kobayashi, kashiwagi, kurasawa, Maeda, Sutou, Muto, Yamane, Nokubi, Komuro, Takasugi, Eto, Yoshida@Itabashi, mhatta, Kitahara, gotom, nakano, ysjj. Matsuyama, Kawachi, David Smith, Charles Plessy, et moi. Report follows:
08:38:18 # Life chroot and proc filesystem. Although I am also tempted to consider taking up ballroom dancing in that case, on a very unrelated note, is this just my kernel, or is it a general knowledge that I can use /proc/1/root/ to access what is outside the chroot?
root@dancer64:/proc/self# ls /proc/self/root/. -id 693684 /proc/self/root/. root@dancer64:/proc/self# ls /proc/1/root/. -id 2 /proc/1/root/.
The reson why I've wanted to check was that initscripts package checks for whether it is inside a chroot with [ -r /proc/1/root ] and that's not a valid check.
06:50:46 # Life Curry party in Ginza. There were a few Debian Developers from the U. S., canada and Italy, and I did a meeting with them at a vegetarian curry restaurant in Ginza. 11 people in total attended, with 6 people from Japan (Knok, Horms, Nori1, Iwamatsu, Daigo, dancerj). Mattia Dongili is the Debian maintainer for user-mode-linux, and I did receive patches for pbuilder user-mode-linux support; only that I didn't remember until he mentioned. Matthew Wilcox and I and few others discussed git/mercurial, and how it improved our lives. I commented that mercurial is not well-accepted because HG has a subtly different connotation in Japan. I discussed a bit about ruby with Daigo, and it soon became midnight. Thanks to all who attended for attending.
11:06:52 # Life After 1 month of playing with ruby. Since I've picked up apt-listbugs, I'm intensively working with ruby, and gained some experience. There are few things that I like about ruby, and few things that I hate. First, the good points. Exception processing is good. Block format is good, looks better than the lisp counterparts. irb is a very useful tool. Then, the bad points. There are too many undocumented when I look up with ri, and there seems to be multiple documents, and none of them covers all; and I need to look into the source-code to obtain basic information like what does this method return, and what exceptions this method gives. net/http, http-access2, and soap4r have incomplete implementation for http (such as proxy authorization and http redirects). It's not completely unusable, but not an utopia. At least it has some promising future.
As for block structures, I like this syntax.
method.something { |var1,var2|
XXXX
}
in method you'll do:
yield val1, val2
And for exceptions, it's good and simple.
begin XXX rescue =>exception XXX retry if XXXX end
10:35:31 # Life Playing still with apt-listbugs. After the hacking sessions, I have versioning support into apt-listbugs. There are two problems; amount of memory consumed, and speed. Soap4r memory consumption is pretty bad; it consumes like 500MB when it receives bug information via SOAP when I do apt-get dist-upgrade (which is around 1000 bug reports). Also, the speed of comparing versions stems from having to fork dpkg every time version comparison happens.
$Id: 200610.html.en,v 1.9 2006/10/29 01:14:56 dancer Exp $