Re: U+00D0, U+01b7 -- variants or distinct chars?

From: Peter_Constable@sil.org
Date: Mon Nov 05 2001 - 20:18:00 EST

  • Next message: Doug Ewell: "Re: Gothic"

    Kenneth Whistler wrote on 03/17/2003 06:20:09 PM:

    > > Should this be
    > > considered a glyph variant of U+00D0,
    >
    > In my opinion, yes. By the way, I also consider it an *ugly*
    > glyph variant, unlikely to gain much traction, typographically.

    Yes, I agree it is ugly. I haven't actually seen any indication of it's
    use, though there is some potential motivation: the need to provide a
    contrast between capitals for 00f0 and 0256 in a single language -- I
    don't specifically know of a language in which that is necessary, but I
    think it was that potential that must have lay behind the Niamey
    conference proposing this.

    > > or should it be considered a
    > > distinct character?
    >
    > To do otherwise would be to introduce casing problems for eth.

    Yes, I realised that a recommendation to encode a distinct character would
    probably entail recommending to encode a *pair* of distinct characters,
    creating a clone of 00f0 that's part of a distinct case pair.

    > Again, a glyph variant, for the same reason. If you try to encode
    > a different *character* each time someone experiments with creating
    > capital forms for a lowercase letter that originally had no capital
    > form, you just create casing problems for processing the data.

    Right. Some months ago, you told me you felt it would be best, if a
    distinct character was needed in a situation like this, to create a
    distinct case pair to avoid the casing problems with Turkish and
    Azerbaijani. Of course, calling things glyph variants is a way to avoid
    the need for a new character. The other side to this issue, though, is
    that, if we call it them glyph variants, then the likelihood of these
    typeforms being supported in commercial fonts is significantly reduced, so
    the user communities that prefer these variants are at a disadvantage.

     
    > This is no more extreme than the two common glyph variants of the
    > capital eng character

    I don't disagree, but I will observe that your comment begs the question
    of what metric you might be applying to determine "extremity of glyph
    variation" -- we don't have any well-defined criteria that we consistently
    apply for deciding when two typeforms are considered similar enough to be
    "glyph variants" of the same "character", or are different enough to be
    counted as distinct characters.

    - Peter

    ---------------------------------------------------------------------------
    Peter Constable

    Non-Roman Script Initiative, SIL International
    7500 W. Camp Wisdom Rd., Dallas, TX 75236, USA
    Tel: +1 972 708 7485



    This archive was generated by hypermail 2.1.5 : Mon Mar 17 2003 - 23:33:09 EST