RE: Codepoint Differentiation

From: Peter Constable (petercon@microsoft.com)
Date: Tue Feb 22 2005 - 08:51:36 CST

  • Next message: Patrick Andries: "Re: Codepoint Differentiation"

    > From: unicode-bounce@unicode.org [mailto:unicode-bounce@unicode.org]
    On
    > Behalf Of UList@dfa-mail.com

    > SERBIAN T
    >
    > > This should be handled by language dependent glyph selection.
    > > That's a standard feature in OpenType and there's no need to
    > > duplicate that facility in the encoding.
    >
    > As I've already mentioned, switching language identifiers back and
    forth
    > every
    > word in an HTML Russian-Serbian dictionary is hardly an efficient
    solution.

    If anyone is compiling a dictionary, they should be using some system
    that maintains the data in a structured format in which everything
    that's Serbian is kept separate from everything that's Russian. Then,
    it's no problem whatsoever to present the data with sufficient info,
    including language tags if needed, or perhaps a different typeface, to
    display the Serbian different from the Russian.

    Frankly, I think having relevant presentation info generated from the
    structured data would be far more useful than scattering variation
    selectors into the data -- something that would have to be done manually
    and therefore prone to error.

     
    > And this is not a job for language identifiers anyway -- at least
    > that'sUnicode's opinion (??).

    That's not at all Unicode's opinion. Unicode just wouldn't recommend the
    use of plane 13 tag characters.

    > That is to say, all the other language-
    > specific
    > Cyrillic letter variants have gotten codepoints. The problem with this
    one
    > isn't that it's Serbian, but that it's italic.

    Rather, it's Serbian italic. It *is* dependent on the language.

     
    > Not quite eligible for a codepoint, not worth switching entire
    language
    > identifiers over one letter. Perfect candidate to be defined as an
    > (optionally
    > ignorable) *sub-category* of the main Cyrillic "t".
    >
    > And *such* a ridiculously simple thing to do, to say, OK, you can add
    > Variation Selector 1 -- with no negative impact on anyone.

    No negative impact? You want every Serbian user to have to enter
    variation selectors when they type a Cyrillic "t" so that it will
    display correctly if at some point it is set in italic? Or to remember
    to enter the variation selector if they format it in italics?

    Peter Constable



    This archive was generated by hypermail 2.1.5 : Tue Feb 22 2005 - 08:52:38 CST