RE: The glyph of the CAPITAL SHARP S

From: Philippe Verdy (verdy_p@wanadoo.fr)
Date: Thu May 10 2007 - 15:48:28 CDT

  • Next message: Philippe Verdy: "RE: glottal stop"

    Mark E. Shoulson wrote:
    > OTOH, I'm not a German-speaker or reader, and I confess that 90% of the
    > time when I see ß, it's the "B" circuits in my brain that fire first. So
    > my opinion probably shouldn't be relied upon (and indeed, very likely
    > should be ridiculed heartily).

    It's quite normal for a non-German speaker, to see a B or beta, but given
    the context (lowercase Latin text, both options are immediately cancelled).
    However, in a all-capitals context, we can't reliably keep the lowercase
    form which looks too much like a capital B; some better hint is needed and
    the best way is to not only avoid the vertical line on the left (something
    already done for the lowercase esszett), but also revert the yogh or
    esh-like assimilation that occurred when the modern esszett form was
    adopted.

    Using Z will not be helpful for modern usage (despite of the German name of
    the character which remains and contrasts with the modern interpretation as
    "doppelt Ess"), so it just remains S as the best visual hint (but we need to
    remove a part of both left and right components of the glyph to avoid
    confusions.)

    I won't speak about other traditional or decorative font styles that may
    adopt any other form, but with caveats when rendering something else than
    German or languages with basic Latin alphabets and very few extension base
    letters. If a language still has Yogh or Ezh, those alternatives will create
    confusions.

    As Unicode representative glyphs should be locale-neutral, as much as
    possible, something really distinct should be selected as the primary glyph;
    a German-language renderer could still use the Ezh-based variant, either by
    direct selection with a variant selector, or though language hints given to
    the renderer (but for multilanguage documents, this will still be a problem,
    notably if German proper names are inserted in non-German documents, this
    usage being the most convincing reason why one will insist on encoding the
    esszet explicitly, and making the distinction very clear, even within
    uppercase titles.)

    Philippe.



    This archive was generated by hypermail 2.1.5 : Thu May 10 2007 - 15:50:11 CDT