RE: Ancient encodings (was: Where is the First> Last> convention documented?)

From: Jonathan Rosenne (jr@qsm.co.il)
Date: Fri Sep 14 2007 - 04:15:01 CDT

  • Next message: Doug Ewell: "Re: [icu-support] complete binary/utf mapping"

    EBCDIC is alive and well (sort of). A fair proportion of your financial
    transaction use it, and much more. The main problem with EBCDIC is that most
    PC and Unix specialists don't understand it, especially the fact that the
    common conversions to and from EBCDIC to common 8 bit codes cannot be used
    safely for binary data.

     

    Moreover, I still have production code in operation that was last compiled
    in 1994. It was written earlier, in 1994 it was just recompiled with no
    change. The copyright notice says 1990.

     

    ISO 8859 is in no way out - there is no reason to do it. There are huge
    databases in these codes, and all modern systems just require is that you
    identify correctly the encoding used.

     

    Jony

     

    From: unicode-bounce@unicode.org [mailto:unicode-bounce@unicode.org] On
    Behalf Of Michael Maxwell
    Sent: Friday, September 14, 2007 12:38 AM
    To: unicode@unicode.org
    Subject: Ancient encodings (was: Where is the First> Last> convention
    documented?)

     

    > The IBM EBCDIC control functions

    Relax, I'm not going to talk about control functions. Rather, I have a side
    interest in what I call the "half life of software", meaning how long one
    can expect a particular piece of software to be usable on anything but
    museum-piece computers (my guess is 5-10 years). I have found next to
    nothing on the topic.

     

    It occurs to me that character encodings are one part of this issue. I
    assumed that there was little or no data still in existence (aside from
    NASA's archives, maybe) that used EBCDIC; but maybe I'm wrong. Another
    encoding that has more or less disappeared is ISCII, as are (I hope) some of
    the proprietary encodings for Devanagari (although I wouldn't count on it
    :-(). Perhaps the ISO 88NN encodings are on their way out, although I'm
    sure there's still lots of data in that format today.

     

    Has anyone looked into this? I'll suggest responding to me personally, as I
    doubt that this is of general interest to this list. (Feel free to tell me
    otherwise.)

     

       Mike Maxwell

       CASL/ U MD



    This archive was generated by hypermail 2.1.5 : Fri Sep 14 2007 - 04:18:00 CDT