Wednesday, January 23, 2008

Hebrew 2.0

Many programmers, administrators, and even users, are wasting a considerable amount of time dealing with issues regarding localization of software. When it comes to Hebrew, this amount of time, doubles itself. For start, Hebrew’s direction (Right-To-Left) is something many software vendors forget in their first versions of a product (take blogger for example), so Hebrew is aligned in the wrong direction. And even when this issue is solved, still there are many quirks. Such quirks can be when mixing Hebrew with English (or other Left-To-Right languages) or mixing with numbers and parenthesis. Other issues involve saving files in the correct code-page, or displaying data in the correct code-page (how many times have you changed the browser’s encoding on misbehaving sites). Sometimes, file formats (such as text) might not include magic headers about the encoding, thus making the client software guess how data should be displayed. Moreover, since there are multiple ways to represent Hebrew (8859-8, 8859-8i, 1255, using UTF-8 etc.), it leads to some conflicts and bad implementations.
My suggestion is a new way to represent Hebrew, at least in the computer software world. It involves a new language, written Left-To-Right, using western characters, or some other characters that are already included in UTF-8 (could possibly be Hebrew characters, though that would be very confusing). I also have a name for this language: Simplified Hebrew. This is the place to mention: I’m not looking to replace the holy-language, I like Hebrew. The words would be the same words, with the same meaning, and the same sound. Only the written language would look differently. Think about it, how many issues could be solved. Adapting software for Hebrew speakers would be like adapting American (US) software, to British (UK). As simple as it gets. Plus, Hebrew speakers won’t have to compromise on their language when they are using computers, since software would be in Hebrew – Simplified Hebrew!


  1. !שממל ןמזומ התא - ןיוצמ ןויער תויהל לוכי הז

  2. You're totally right!
    It's amazing how much of my time as an Oracle Applications sysadmin is wasted on Hebrew specific issues.
    It's even more amazing that no one have designed a sound solution for this issue until now.

  3. For a minute there I thought you were talking about reinventing the visual-Hebrew encoding but then I realized you are talking about something even more bizarre - changing the way we write.

    Hebrew subscript has a lot of f@#k-ups: LTR, sometimes hard to read with no nikud (not phonetic), But the fact is that Hebrew is a revived language after 2000 years it was dead, is and amazing fact - I can read old subscripts without much trouble and I think it's cool.
    This subscript is unique for us and something that defines us as a nation and therefore irreplaceable.

    I think that as long as everybody will stick with Unicode, there shouldn't be any trouble in the world of computers and technology.


  4. @shlomil:

    1) You are talking about the Hebrew script. A subscript is reduced characters printed at the foot of the line (LaTeX's _).

    2) The idea itself is not completely unthinkable; as you should know, the current script is not the first script that was used for Hebrew. Turkey provides a more modern example: they changed their script, less than 100 years ago, from an arabic-like system to a latin-like one.

    3) However, as your reply indicates, and as Turkey exemplifies, such a change is political and not just technical, and can probably only be done while revolutionizing a whole lot of other things.