Re: Fonts

From: Hans Aberg (
Date: Mon Nov 13 2006 - 09:08:32 CST

  • Next message: Adam Twardoch: "Re: Fonts"

    On 13 Nov 2006, at 14:47, Adam Twardoch wrote:
    > > There are the Computer Modern and AMSFonts in PostScript Outline
    > Form available at <>, developed by Blue Sky
    > Research and Y&Y. I wonder what it might take to make Unicode fonts
    > out of these - there is probably some xopyright issue involved. But
    > these fonts produce very good math results, and it seems
    > unnecessary work to develop a wholly new fonts.
    > First of all, I don’t agree with the thesis that since existing
    > fonts provide "very good" results, there is no need to develop new
    > fonts. Both Computer Modern and the AMS fonts are heavily flawed
    > and have some serious typographic problems. In general, CM is not
    > really very suitable for offset printing (since it is too thin),
    > the quality of diacritic characters is very questionable etc.

    I am only addressing it from the mathematicians point of view, were
    one wants something looking good enough for expressing the correct math.

    > Also, these fonts, just like the AMS fonts, lack many characters,
    > for example properly drawn Cyrillic. Typesetting math mixed with
    > English may be fine with them, but if you’re trying to typeset even
    > just Czech, not mentioning Russian or Greek, you run into problems
    > (mostly, the Greek characters in those fonts are optimized for
    > math, not for Greek text).

    In particular, I only write math in English, though I imagine, there
    must be someone somewhere out there using a different language. :-)

    > As many of you know, typesetting mathematics is more complicated
    > than setting regular text. Just providing some glyphs in the SMP is
    > not enough.
    > Cambria Math is a font that includes a special OpenType table
    > ("MATH") that defines the relationships between different glyphs
    > used in typesetting mathematics. Microsoft is planning to release
    > the specification for the mathematical OpenType extensions along
    > with some tools in near future. It is also likely that XeTeX will
    > be the first TeX system that will be able to use the Unicode/
    > OpenType mathematical fonts.
    > I personally think that investing the effort of developing a font
    > with just the mathematical glyphs defined in the Unicode standard
    > included, without the OpenType math-specific extensions that will
    > allow the font to work in Office 2007 and XeTeX (for example), kind
    > of misses the point.
    > Microsoft’s Murray Sargent has been writing about this recently:
    > You can contact him for more information on the subject.

    Actually, I came across the link. It seems to contain a good idea,
    namely, an language using ASCII for inputting Unicode.

    For more complex math, one needs something corresponding to a macro
    system; perhaps some lambda calculus may be used here, as a macro
    system quickly becomes rather crippling. In addition, I think an
    analysis of the math (human, natural) language is needed, to one can
    have develop a semantically correct syntax. I do not pretend this
    will come easy. :-)

       Hans Aberg

    This archive was generated by hypermail 2.1.5 : Mon Nov 13 2006 - 09:09:41 CST