Re: Script L

From: Asmus Freytag (
Date: Fri Mar 17 2006 - 04:58:03 CST

  • Next message: Antoine Leca: "Re: Representative glyphs for combining kannada signs"

    On 3/17/2006 1:41 AM, Raymond Mercier wrote:
    > "Anto'nio Martins-Tuva'lkin" <
    > <>> wrote
    > >>Considering that U+1D4A7 is one of the many "wholes" in the Mathematical
    > Alphanumeric Symbols block, which correspond (considering their place in
    > the alphabetic sort order) to those in the Letterlike and Mathematical
    > Symbols blocks, I assume that U+2112 is a regular script L, not "locked"
    > to the glyph(s) used in Maths (Laplace transform).
    > <<
    It is indeed correct that if you were to use a script font that uses a
    different type of script, that you can use 2112 for the L in that font.
    the intent is that this would still be for Math.

    Knuth has experimented with his Euler typefaces - they are used in place
    of the more traditional renditions, but they designate semantically a
    variant of mathematical variables, even though the corresponding
    shape is not a textbook 'script' shape.

    Varying individual characters independently of a consistent type
    design was not intended or considered when encoding these,
    so 2112 is intended to be constrained by the font used for
    the entire 'script set'.

    This is one of the reasons why we disunified the small l, the shape at
    2113 was effectively frozen.

    If anyone were to adduce evidence that the same is true for the cap L
    as used for the Laplace, we might have to consider that one as well
    - however, by now, the third party mappings to Unicode are so
    firmly established as to make such a move nearly impossible to consider.

    It's also not as widely used as the lower case l at 2113 which is used
    for liter in a very conventionalized, frozen form, which made it imperative
    to separate from the generic mathematical script l.
    > As far as I can see (Tables in Unicode 4) U+1D4A7 is 'reserved', with a
    > reference to U+2112, which is not what I was after.
    > I have just noticed however that civitype.ttf (which I acquired from
    > somewhere almost 10 years ago) has the right sort of thing, but still not
    > close enough.
    > Well, it is not of huge importance - I was just trying to keep to the
    > notation of a (French) book that I am translating.
    What sort of book, and is the shape there anything other than
    a font variation? In other words, is it notationally significant?
    For example, the precise shape of 2118 is notationally significant
    and unrelated to any of the math alphabets (despite its name).

    If such a thing happens in modern notation (or significant historic
    notation) we might consider adding such a character outside
    any given math alphabet. A lot would depend on the circumstance

    Care to share some details.
    > Raymond

    This archive was generated by hypermail 2.1.5 : Fri Mar 17 2006 - 05:01:04 CST