Re: interleaved ordering (was RE: Phoenician)

From: Kenneth Whistler (kenw@sybase.com)
Date: Fri May 14 2004 - 18:30:57 CDT

  • Next message: Philippe Verdy: "Re: Multiple Directions (was: Re: Coptic/Greek (Re: Phoenician))"

    > jameskass@att.net a écrit :
    >
    > >Dean A. Snyder wrote,

    > >
    > >>The issue is not what we CAN do; the issue is what will we be FORCED to
    > >>do that already happens right now by default in operating systems,
    > >>Google, databases, etc. without any end user fiddling?

    The trend for such systems is to build in generic, default
    behavior (for collation or for other aspects of localizable
    information), to support a number of high visibility and
    high demand particular behaviors "out of the box" and then
    to open the systems to end-user customization of particular
    combinations of behavior.

    The IT industry is, of course, a long way away from perfection
    here, in part because the entire field of internationalization of
    software is considered bizarre geekiness even among your
    run of the mill programming geeks. But the globalization of
    information technology is inevitable, in my opinion, and as
    that globalization proceeds, the inevitable tension between
    central control and end user demand will play itself out in
    ways that make the technology eventually more flexible and adaptive.

    People complaining about software now need to take occasional
    reality breaks and remind themselves of what things were like
    in the 1970's, when you still typed in your Fortran programs on
    punchcards, fed in your decks, and then waited 2 hours for
    the computer operator to deliver your line printer output
    to get a list of the syntax errors in your program.

    > >
    > >That's the question.
    > >
    > >Since search engines like Google survive based on their ability to serve
    > >users' wants and find what users seek, why wouldn't Google make such
    > >a tailoring?

    Google is on the frontline for this stuff:

    https://services.google.com/tcbin/tc.py?cmd=status

    The applicable model is to *distribute* the work of adaptation to
    the people who understand it best -- the people who have the
    requirements and the knowledge of what they want and the means
    to specify the local behavior.

    In this particular case, we are talking about translation of
    UI into dozens of languages, rather than end-user specification
    of search criteria. But in principle the concept of distributed
    definition of string filtering or string matching criteria for
    a search could also be implemented. If not Google, then somebody
    else, will come up with a way to do this and let the users devise
    their own criteria for how they search into the massive internet
    (and other) databases involved.

    Patrick responded:

    > Because the Phoenician user community is very very small ? Same goes for
    > Microsoft on some collations already mentioned (French Canadian sorting,
    > Khmer) and those are much larger communities.

    And the answer is to democratize the approach. One shouldn't be
    demanding that The Borg centrally define and implement all uses
    for all users, so that users simply dial Channel 621 and then
    sit there passively assimilating and get dished up their content.
    Instead, the users should demand of The Borg that user-definable
    requirements be supported actively, so that the *people* get
    to define what they do and how it is done at the point they
    interact with the software.

    --Ken



    This archive was generated by hypermail 2.1.5 : Fri May 14 2004 - 18:32:13 CDT