Slashdot Mirror

← Back to Stories (view on slashdot.org)

Series of Linux localizations coming out of Asia

Posted by timothy on from the hmmm-there-are-a-lot-of-people-there dept.

Imran William Smith writes "The Asian Open Source Centre is reporting a series of recent successes in translating Linux to Asian languages, where there are also many complex font issues to be solved. This site has a wealth of information about open source and free software activity in Asia."

11 comments

  1. Re:___ by Surreal_Streaker · · Score: 0, Offtopic
    Damn Slashcode filters. That was supposed to be:

    Yao!

    /Rimshot

  2. As long as it's not called... by Anonymous Coward · · Score: 0

    SARS GNU/Linux...

    It can make your CPU fan stop spinning in 5-10 days if not treated properly.

    1. Re:As long as it's not called... by Anonymous Coward · · Score: 0

      The new OS is called Rinux Opelating System

  3. not as promising as it appears by Anonymous Coward · · Score: 3, Interesting

    Yes it is good to see more languages supported, but in this world software = English. The mass of programmers largely either live in English speaking countries, or countries like Germany and Scandinavia which are switching over to English. The attitude towards languages in software development is "learn English, tough".
    Multilingualism is barely even an afterthought in software development.

    1. Re:not as promising as it appears by Anonymous Coward · · Score: 0

      Tell that dozens of Chinese and Indian folks (not programmers) who hardly will speak english in near future. They are wrong race and language and get what deserve, right?
      Think twice when posting, moron.

  4. This would be great by Krellan · · Score: 3, Insightful

    I think this would be great. I don't know how to read Asian languages but like to have support for them nonetheless, because I'm fascinated by various writing systems used all over the world.

    Problems I see:

    * What to do about the text console? Would the framebuffer console become mandatory, since VGA text can only support 512 unique characters (nowhere near enough for Asian languages)? One of the strengths of Linux is its extremely fast text console updating, and on many systems I've chosen to keep the standard VGA text instead of going to framebuffer mode for this very reason.

    * What to do about fonts? Making full Unicode fonts is very expensive and time-consuming. There are some free fonts already but nowhere near the level of fonts available for standard ASCII.

    * What about software that is hardcoded to use 8-bit characters? The tcsh shell comes to mind.

    * What about IME's (input method editors)? Are they smoothly integrated with all X applications, or are they hardcoded to only work inside of one specific application? A global IME will be needed, if it does not exist already. Also, an equivalent IME will be needed for console mode, if that is supported.

    I'm assuming UTF-8 will become the standard encoding format. That will be a good thing, as it provides a smooth upgrade path.

    Translating error messages and the like should be easy. Accepting input from the user, and handling it properly, will be harder.

    My guess is that full I18N will be possible only in graphical mode, and that text mode will still be assumed to be English. Many distributions are defaulting to graphical mode these days, leaving text mode solely for the advanced user who has manually customized their setup. I use text mode all the time, for speed and other issues, and would hate to see it deprecated due to the difficulty of getting it to work internationally.

    --

    Dr. Demento On The 'Net!

    1. Re:This would be great by unmadindu · · Score: 1

      I do not know about all the Asian languages - but I am actively involved in the Indian language scenario, and I would try to answer your questios from that perspective.
      1. We cannot do anything inthe text mode, as text mode only supports "fixed width" fonts. This won't do for Indian languages. However, some developers have tried to modify the kernel and other hacks, but the results are not very interesting.
      2. Fonts are a real bottleneck. We cannot work with plain old TTF fonts - we use a new format called OpenType (search microsoft.com). That adds a lot of complications - and usually each font works with one language. For example, you may check ot http://www.nongnu.org/freebangfont which works with Bangla (the 4th largest language in the world in terms of the number of people speaking it). I am the coordinator of this project - and working on this can be really troublesome.
      Recently, there has been proposals on a grand unified Indian font - let's see what happens.
      3. Hardcoded software has to be changed - nothing doing - no alternative way exists.
      4. Input method (IM) is yet another problem as a result of lack of proper well publicised standards. X has a fairly good Input mechanism - and GTK2 has one of its own too. You are free to choose whichever you want to use.

  5. Utter nonsense by GCP · · Score: 1

    Too bad I'm so late responding, but in case anybody reads this:

    All major platforms are expending serious resources to internationalize. Java and .Net are both based on Unicode. The only two text encodings that all conformant XML readers/writers are *required* to understand are both forms of Unicode. The document character set for all HTML documents changed to Unicode in HTML 4 (and thereafter). Perl, as of 5.6, changed over to being entirely Unicode internally. Oracle and Microsoft both recommend buiding all new databases as Unicode. The Linux Standards Base on which most major distributions are based is about to make Unicode (UTF-8) the default system encoding.

    And the list goes on and on. Why Unicode? Because it lets you handle not just English, and not just Western languages, but ALL of the world's languages with the same applications. Big players in the world of software are big players all over the world. Limiting themselves to English is not a possibility, and the smaller players are falling in line behind the big ones.

    "Learn English, tough" may have some validity when talking about *content* (an online discussion among an international group of specialists may take place in English only with no translations, for example), but it is nonsense when talking about software architecture. In software architecture, "English only" approaches are increasingly obsolete.

    --
    "Those who have never entered upon scientific pursuits know not a tithe of the poetry by which they are surrounded."
    1. Re:Utter nonsense by Moridineas · · Score: 1

      The poster couldn't be more right!

      And once again, OpenBSD is taking the innovative lead--pf now has better bilingual support than any other packet filtering software out there!