Diary for Mar. 2002

plans

daily blurbs


31 Mar 2002

00:06:55 Checking many library packages in Debian. I have noticed that something is wrong with most packages that start with "lib" and end with "0".

16:06:07 I've put up the library packaging guide to my columns page.

23:33:37 bigloo looks really good. It looks like a pretty much well maintained program to me. It might be enjoyable to code. Emacs is a very nice environment, but I really needed something more than that to work with. Some weird tweaking exist in bigloo, slight difference exists in bigloo in Debian and bigloo from upstream, I believe.

24 Mar 2002

03:25:24 I've recovered the internal Debian Mirror server. It is up for few hours now. I hope it is working in a good shape. To my surprise so many packages in build-essential has changed since last mirror update. Too bad.

15:35:08 Found a possible flaw in mikilab-ssh. It md5 module for PAM doesn't seem to like the /etc/passwd file it creates. hmm.

17 Mar 2002

14:28:29 My girlfriend gains 24-hour full-time network connection. That sounds good.

14:35:44 www.eetimes.com seems like an interesting on-line resource. I never knew about it.

16 Mar 2002

14:56:00 I dreamt for a second if I were to be teaching in a Junior High School. It sounded surreal. I may be losing interest in many things that used to attract my attention. I don't want to do many things that are there. I am not sure. Maybe I am overloaded. I am overloading myself without planning, as always.

14 Mar 2002

20:29:16 I am testing pbuildler now. It seems to be working fine. I think it should be running great. I have tested "create", "update", and "build", and "pdebuild". I think it is apparently working. I will try and rebuild about 50 packages and see if any of them have apparent problem.

13 Mar 2002

10:29:06 Doing too much hacking on pbuilder, I wonder if it is not broken. I am doing too much... I wonder if it would run again. A software that does not run is useless. I am splitting it in parts, and adding possible dry-run options, to make it debuggable. I need to go through the code more thoroughly, and mark some parts as blackbox. I need to know some parts better. I have noticed that the parser is quite slow, and sluggish, very inefficient. Spawning awk/sed/tr and whatever else is not good. I need to make it optimized a bit more, using only awk, possibly. Actually dpkg-architecture command was taking a large amount of time to run..

12 Mar 2002

15:08:25 software page is updated. I need to make my works obvious.

18:22:06 I need to consider about how I might make my living in the future.

11 Mar 2002

01:29:57 Trying to write some kind of resume so that I can give some ideas to myself on how much competency I have, and what kind of things might appeal.

02:26:14 I suddently felt an urge to mirror Debian on my wave modem, but it is too slow. I think Debian archive for testing/i386 is at 7GB, and getting that much is impossible. I think it is possible to keep up the difference once the system is up and running, but bootstrapping it would be impossible, I guess. My calculation tells me it will take 20 to 50 days in my current bandwidth. Too bad. I should most probably take up some CD-Rom and get the archive fed from that.

03:01:47 I've discovered that many of the access comes from altavista search bot. I don't like having a search bot using up all the bandwidth.

8 Mar 2002

11:23:15 Complete break from most things, including internet connection. HOKKAIDO is a very cold place, but it is fun to be in a completely different location. It is nice to see the snow, at least. Yesterday I gave my presentation at a research meeting, and it was quite good. I think it was fine. Pretty difficult to define what "fine" is, but I would say it wasn't the worst case I had anticipated.

5 Mar 2002

14:20:11 I think the guys doing tetex for Debian really need some kind of help.

4 Mar 2002

22:25:16 If there are still some things that depend on that feature, that feature is not "obsolete". Free software does not have the responsibility, and it does not want to take the responsibility. There is no estimate of costs running in this field. Calling features obsolete, and behaviors broken does not help at all to improve the status quo of free software.

2 Mar 2002

00:35:48 Many things are going on. boot-floppies are progressing steadily, and the result of my hacking into slang and newt are somehow hopeful. Good.


Junichi Uekawa

$Id: 200203.html.en,v 1.14 2002/03/31 15:30:56 dancer Exp $