From: Kenneth Whistler (kenw@sybase.com)
Date: Thu May 22 2003 - 21:37:19 EDT
> From: "Kenneth Whistler" <kenw@sybase.com>
> > Yes. Recent additions of large numbers of mathematical
> > symbols were done at the request of and with the expert
> > participation of the STIX Consortium of mathematical and
> > technical publishers, nomenclatural and font experts of
> > the American Mathematical Society, participants in MathML
> > development, and developers of commercial mathematical
> > formulae handling software.
And Philippe (Johnny Come Lately) Verdy opined:
>
> In my opinion, Unicode made an error by accepting such encoding.
Well, you are entitled to your opinion about this, but I was
responding to your prior claim that the addition of math symbols
to Unicode would not be of use because math cannot be represented
in plain text anyway. To which I responded that the additions
were at the request of various mathematical and technical publishing
experts, because they asserted that they *wanted* to use them.
> If style is significant in the interpretation of text, then it
> should have been much more coherent to define supplementary
> diacritics that modify the semantic of the base character,
> by assigning them a well-defined style or form variant, in a
> way similar to an invisible nukta in Brahmic scripts...
Such an option was fully debated years ago before the
mathematical alphanumeric characters were added to Unicode 3.1,
and was rejected because it would play havoc with the
representation of nonmathematical text in Unicode.
> This would have also allowed:
> 1) to use any existing letter or digit in any script (for
> example the AE letter or some Runic, Hebrew or Arabic letter,
> or some Brahmic digit) as a mathematical symbol that needs
> various styles for meaning different mathematical semantics.
Yes, anyone *could* do such things, but in fact mathematical
practice does not do so. Mathematicians avoid even
using accented Latin or Greek characters, because the accents
would interfere with the use of other math diacritics and would
confuse the point of the formalism.
The UTC had advice from some of the world's leading experts on
mathematical publishing, and the repertoire that was chosen
reflects their best advice. (And I can assure you, they have
reviewed many more mathematical publications than you are likely
to have done.)
> For now the set of allowed letters is restricted to a few
> Latin or Greek letters
As intended.
> (plus standard diacritics)... So in the future, Unicode will
> probably receive requests for many other mathemetical letters
> or digits...
From whom? If the professional mathematicians say the current
set is sufficient.
> I hope Unicode will not need to redefine styled variants for
> ALL existing letters in defined alphabets or abjads of the BMP...
Why stop there? I am *sure* that someday the mathematicians
will need Bold Fraktur Linear B Ideograms for their formulae.
Drat, it seems we made a big mistake in not encoding a
BOLD FRAKTUR STYLE COMBINING MARK..... NOT!
--Ken
This archive was generated by hypermail 2.1.5 : Thu May 22 2003 - 22:30:21 EDT