Re: Representing Unix filenames in Unicode

From: Hans Aberg (haberg@math.su.se)
Date: Tue Nov 29 2005 - 00:27:12 CST

  • Next message: Chris Jacobs: "Re: Representing Unix filenames in Unicode"

    On 29 Nov 2005, at 04:35, Doug Ewell wrote:

    >>>>>> Most facts points to that the Unicode/10646 is a human
    >>>>>> interface,
    >>>>>> not a computer to computer [] interface.
    >>
    >>> ...I mean saying that "Unicode/10646 is a human interface." It's
    >>> a coded character set.
    >>
    >> I abbreviated "human to human interface and human to computer
    >> interface" to "human interface", as I felt that the former was too
    >> long. Sorry if some clarity was missed out by that.
    >
    > 1. I don't see how Unicode, or any other coded character set, can
    > be considered a human-to-human interface. Very few humans
    > communicate directly with each other by means of character codes.

    What about the email you just wrote?

    > 2. Coded character sets exist so that computers can deal with
    > text. This has to do with input and output, which is human-to-
    > computer, but it also has to do with processing and interchange,
    > which is computer-to-computer.

    I am speaking about the interface required for the processing. The
    processing may forward data which is not strictly speaking for the
    processing itself. In the given context, low level filesystems need
    not know anything about character encodings, as it suffices to use
    byte string. One can use character encodings on this low level, and
    ot was reoprted one, but not two such systems. There appears to be no
    benefit in doing so. So this is a strictly computer-to-computer
    interface, even though for debugging pruposes human interfacing
    should probably be at hand.

    > 3. The original statement sounded as though "Unicode/10646" were
    > somehow different from other coded character sets, in terms of
    > being human-to-human or computer-to-computer. If that was
    > intended, I don't agree.

    No. It is just the general principles of computer language design.
    The human to human interface means that the humans among themaselves
    must be able to interpret the computer language. The human to
    computer interface means that computer must be able to translate it.
    The computer to computer to computer interface means that computer
    programs among themselves must be able to translate it. Different
    needs in these areas leads to different computer languages.

       Hans Aberg



    This archive was generated by hypermail 2.1.5 : Tue Nov 29 2005 - 00:29:03 CST