Re: Uppercase is coming? (U+1E9E)

From: Marnen Laibow-Koser (
Date: Tue May 08 2007 - 01:58:38 CDT

  • Next message: Marnen Laibow-Koser: "Re: Uppercase is coming? (U+1E9E)"

    On May 8, 2007, at 2:04 AM, Werner LEMBERG wrote:

    >>> See
    >>> where I have given an F, J, U, ß, long-s, and two capital sharp
    >>> esses.
    > Sorry to say but I dislike both of them.

    As do I. The second doesn't look uppercase (perhaps for the reasons
    that John Hudson mentioned) and the first is just plain ugly (but I'm
    sure you knew that already, Michael -- it doesn't look like a
    finished character). I recognize that these are just first steps in
    that direction, though, because I've seen too much of Michael's work
    to believe that he'd be satisfied with these.
    >> I spent some time today drawing uppercase eszetts, and found it a
    >> pretty gruesome business.
    > I believe you. My thinking is that an uppercase form of ß should
    > always look similar to `SS'

    Why? The lowercase form certainly doesn't look like ss, at least in
    Roman type where the long s is not otherwise used.

    > -- as TeX shows, which has introduced an
    > `SS' glyph about 20 years ago to make uppercasing give correct visual
    > results.

    And some all-caps Windows fonts have this as well in the ß slot
    (haven't seen it in Mac fonts for some reason, although I wouldn't be
    surprised if an example turned up). I think it's little more than a
    hack, and I would *not* rely on it as a model for how [ß] should look.

    > To make an `SS' glyph distinguishable from `S' `S', as
    > needed in official documents like passports, why not adding a small
    > `ligature connection' similar to the bow in the `ct' ligature glyphs
    > formerly used in French typography (see attached image)?

    To me, that looks too much like two glyphs, where the lowercase ß is
    clearly only one, at least in Roman type. about an S with an extra bowl, sort of like


    (If I have time, I might try to draw one and see how it looks, but my
    skills as a type designer are limited at best.) This might pass the
    Trajan test...

    On the other hand, does the lowercase ß pass the Trajan test (or
    lowercase equivalent)? In many fonts, I think not. It's kind of a
    strange letter that looks out of place in Roman type. Perhaps the
    Swiss were wiser than I realized.

    Historical question: does anyone know how recent the use of ß in
    Roman type is? I have a German grammar (in English) from sometime in
    the 20th century (I'm guessing 1930s, since it's early enough that
    all the German is in Fraktur, but I can't lay my hands on the book
    right now to cite). There's a prefatory section that introduces the
    Fraktur alphabet, along with Sütterlin and German Antiqua
    handwriting. It also mentions the various Fraktur s forms, including
    the difference between Fraktur ſſ and ß, and follows that with the
    comment: "In Roman type, 'ss' is used for both these forms". So it
    appears that at least at the time this book was published, Roman ß
    simply didn't exist.

    (One would assume that ß made its way into Roman lowercase for the
    same reasons that it may now be making its way into Roman uppercase.)


    Marnen Laibow-Koser

    This archive was generated by hypermail 2.1.5 : Tue May 08 2007 - 01:59:55 CDT