Older blog entries for roozbeh (starting at number 122)

31 May 2006 (updated 11 Jun 2006 at 16:45 UTC) »
Sharif Linux: Mehr News has covered our Sharif Linux event fine enough (Persian, Urdu, with more photos). The report could have been much better, of course…

The Persian title is somehow funny, and somehow resembles my rants here in this diary: “Despite United States economic sanctions, Sharif Linux was produced according to the requirements of the Persian language”. Of course I only mentioned economic sanctions as the reason for why several western software companies were not interested in the Iranian market. I’m really glad I didn’t talk about US government regulations and how they affect Iranians, or the article/title might have become a propaganda piece.

This is a photo showing me show an anaconda screenshot using gthumb. Unfortunately the laptop sticker that says “EFF, Proud Member” cannot be seen:

Me showing a screenshot of anaconda

31 May 2006 (updated 11 Jun 2006 at 16:49 UTC) »

Following is some Sharif Linux 2 Desktop Edition statistics, as promised.

  • Number of binary CDs: 2
  • Number of RPMs shipped: 612
  • Number of SRPMs: 518
  • Number of changed SRPMs: 101 (19.5%)
  • Number of SRPMs used intact: 417 (80.5%)
  • Number of patches over upstream: 137
  • Number of subversion commits for non-translation updates: 865
  • Number of subversion commits for translation updates: 401
  • Most patched packages: anaconda (19) and evolution (19), followed by openoffice.org (8) and gtk2 (7).
  • Number of bugzilla bugs fixed over upstream: 349

Random notes:

  • Upstream for us was actually an updated Fedora Core 4+Extras+Livna, instead of original upstream packages. So 19 patches for evolution really means 29, since the latest update to FC4’s evolution already had 10 patches.
  • By patches or patched I mean files ending in “.patch”. The changes to the “.spec” files and new “Source”s are only counted in the CVS commits. Translations fixes were not done as patches, but SRPM “Source”s.
  • Bugs means things we had bugzilla’d in our internal bugzilla. Some fixed bugs were only reported upstream (to gnome, fedora, OO.o, ...) or were fixed by a developer or translator without filing. The real number should be about 500.
27 May 2006 (updated 27 May 2006 at 15:00 UTC) »
giulia: Microsoft’s Michael Kaplan writes about changing the user interface language at runtime and how hard it would be, even if you had the tools for it:
The distance between the possible and the practical here is substantial; the distance between the possible and the maintainable is orders of magnitude greater still.

How often would it really and truly be worth it?

Danilo thought we needed a UI-switch support system for giulia when we were discussing giulia. But from the Sharif Linux 2 experience, I found that it is already hard for application maintainers to keep the code properly internationalized. I really think making the applications UI-switchable would not be worth the time really.

23 May 2006 (updated 23 May 2006 at 20:23 UTC) »
Our Tower of Ivory: We lived in a tower of Ivory, and we had no clue about it. We have been experiencing technical problems with our primary (academic) internet service provider, so we were using the fallback internet service for the last few days. This has the level of website blocking a man in the street gets from his internet service.

And we found interesting things. It’s not only marginallly controversial web sites like Orkut, BBC Persian, and the Kurdish Wikipedia. Many other things are blocked: the whole livejournal.com is blocked. Flickr is blocked. Not only searches for pornographic words are blocked, but also any search that contains “woman” is blocked. LaTeX is blocked because latex has sexual uses.

And the blocking software is so simple-minded that it even stops you from doing ‘positive’ things. Today, I was searching the Encyclopaedia Iranica for calendars used historically in Persian/Iran, and the most important article for me was blocked because it had the term “Islamic period” in the URL. That is blocked because it has the word “period”, a synonym for menstruation which may supposedly be used to find pornographic pictures on the internet...

The thing amazes me the most, is that my own name is blocked! My family name, “پورنادر” (Pournader), starts with the transliteration of the word “porn” into Persian, “پورن”...

23 May 2006 (updated 11 Jun 2006 at 16:44 UTC) »
Duplicate
Linux and hardware: I’m feeling sick after a few hours of trying to get 1280x1024 output on an external LCD connected to my Vaio TR2...

Whatever I do, I can not get a one X pixel for one external LCD pixel with this combination of laptop and LCD. Even if I don’t want to use the whole display, the grapbical card thinks the external 1.33:1 display is a 1.66:1 display (like the laptop’s own LCD).

20 May 2006 (updated 20 May 2006 at 19:57 UTC) »
Sharif Linux: Sharif Linux 2, our Persian-enabled enterprise GNU/Linux distribution, will be publicly released on May 28, 2006.

For those of you inside Iran, there will be a ceremony and an introduction to the operating system on the same day (7 Khordad), 16:00-18:00, in Sharif University of Technology, Jaber ibn Hayyan hall (map point 1 on this map). The ceremony is open to the public.

Ah, forgot the screenshot. This is rhythmbox, completely localized to Persian (there are two minor localization issues, see if you can spot them). [more screenshots here]

I will post a hall of fames chart soon (try to guess which package got the most number of patches).

Christian, of course Google has every right to do whatever he wants with his money. But let me quote Rocky Bernstein:
“This is a non-sequitor. I don't think the issue is just this year or within the confines of the timeframe of the summer of code. For example, should some other organization want to follow google's very generous example, they might be tempted to make the same (perhaps arbitrary, or worse, wrong) restriction.

Governments especially bad ones, often work by intimidation and innuendo. If it is the case that there is no applicable law but you/google just don't want the possibility of a hassle, in a sense the repressive forces have succeeded in repressing as they have too often in the past.“

I don’t think it’s about frustration. For me, it’s about oppression. US and Iran not on good terms? Who knows that more than the likes of me who are oppressed by both of the governments? I have to deal with it? I am, but I’m also trying to change some things, at least by trying to raise some minimal awareness.

Ah, and the sanctions are quite effective. I would appreciate it if you take a look at this older post of mine to get some ideas. If the goal of the US government is increasing the power of the Iranian government over the Iranian people and making sure Iranians can’t thrive even outside Iran, they are effective. Iranian government working on atomic energy and at the same time saying things about Israel? Iranian people lose. US government worried that Iran may develop atomic bombs, so increase the sanctions? Iranian people lose. This is a lose-lose game for the Iranian people (who are not even players in the game, some of which don’t even care about Iran or US or atomic energy), living both inside and outside Iran. And Google is helping the governments to make Iranian people lose more.

BTW, I really don’t know how happy or unhappy Iranian students living abroad are about this. It’s been a long time since I’ve been a student, and I’ve never lived abroad. I know some Iranian students are participating in Google’s SoC under their foreign friends’ names and some have even applied to GNOME under these fake identities. How do you like that? Somebody doing open source work and not being able to get the credit?

Behdad may know, and he has promised to write about this with a lot of background information on recent sad things happening to Iranian students in the US (some of which are related to Google).

PS: Chris DiBona is probably the sanest guy involved in this matter at the Google’s side. He has helped more than I had expected. Google’s lawyers are the guys who started this situation, and they are simply sticking to it for not even wanting to think.

Google and FSF: Wow! RMS could be soooo great sometimes. I really like the way he can punch:
There is a total disconnect between your reply and my message that you replied to. Your reply has no relationship to what I said.
So I can't actually change this ruling from our legal team.
I didn't ask you to change it. I asked for an explanation of why it doesn't correspond to what the law says.
I sympathize ... But my personal feelings matter not one bit in this case.
Exactly, and we're not talking about your personal feelings.

Are we supposed to interpret this as meaning you are a prisoner and someone has a gun to your head?

(boldface is from RMS, while italic is from Chris DiBona.)

To which Chris DiBona replies:

I'm sorry Richard but I've given you all the reasoning I can provide. If you cannot take part in the Summer of Code because of this, I understand and ask that you let me know by the 17th so that I can plan appropriately.

I believe that translates to “FSF, if you really want our money, you should cope with whatever arbitrary reading we have of the US law. No Iranians, OK?”

I really can’t guess what’s happening inside Google, but the original story is getting confirmed again and again. To explain its actions, Google is referring to a law that says something totally different. Perhaps the proper term is “discrimination through self-imposed ignorance”? If it’s not that and it’s not Google doing it intentionally, the only remaining chance is the US administration having restricted Google in a secret way that Google may not talk about. RMS is practically saying the same thing is his last line, but he is very probably joking.

15 May 2006 (updated 15 May 2006 at 18:24 UTC) »
Sharif Linux: We just fixed that last blocker bug in Sharif Linux 2.0 which had took us two whole days. This means I can take some rest, and read the emails and blogs again after a few months of just looking at them.

Hamed and Farzaneh helped a lot in fixing this last bug (the fix was a one-liner patch, as you may guess), which was an OpenOffice.org crasher. Caolan also helped me on IRC to debug the monster.

I really hope that tonight’s build will be the one final Release Candidate.

113 older entries...

New Advogato Features

New HTML Parser: The long-awaited libxml2 based HTML parser code is live. It needs further work but already handles most markup better than the original parser.

Keep up with the latest Advogato features by reading the Advogato status blog.

If you're a C programmer with some spare time, take a look at the mod_virgule project page and help us with one of the tasks on the ToDo list!