From: Philippe Verdy (firstname.lastname@example.org)
Date: Fri May 23 2003 - 09:41:26 EDT
From: "Asmus Freytag" <email@example.com>
> > > I hope Unicode will not need to redefine styled variants for
> > > ALL existing letters in defined alphabets or abjads of the BMP...
> I hope so to. And not only hope, but given the understanding of the
> way the development of that 'script' is operating, I am fairly
Do you include educational/vulgarization publications in your hope? They have a very distinct need to introduce concepts with a notation that can easily be recognized and learned by students using a non-Latin non-Greek script for their native language... Mathematical symbolic notations in this case are adapted to local scripts and languages.
I am quite confident that scientific publications will adhere to a standard, but as stated in the TR document for Maths and Unicode, mathematicians are very creative, and will sometime refuse to use a past standardized notation that may conflict with the new concepts they need to represent, notably if both semantics are needed simultaneously and do require separate notation/symbolic systems.
Just consider the difficulties introduced with quaternions and how this conflicted with the symbolic representation of complex numbers, and other difficulties encountered in quantic physics for the notation of quantas or probabilistic wave functions as mathematical concepts/objects. Due to possible ambiguities in formulas that combine multiple "numeric" systems, a new notation is often introduced with its own set of symbols, most often derived from existing scripts (so why mathematicians would restrict themselves to use only Latin and Greek letters/digits or to a restricted set of font styles ?).
That is still true that Unicode offers new alternatives to mathematicians, notably by using other standardized and widely implemented scripts. Will it be enough? The future will answer to this question...
Another possible source of conflicts: if mathematicians use some RTL Semitic script as symbolic notations within standard LTR mathematical formulas, how will Unicode will adapt to this? Does it mean that bidi-override format control characters will need to be used for these symbols ? Is there some existing local usage of mathematical notations with RTL direction ?
How Unicode manages the case of matrix notations ? It's not clear because Unicode only suggests using parentheses for grouping cells in rows (or columns?) and does not encode properly the dimension by which such linear notation must be interpreted... Here also, the interchange needs may require defining some other type of parenthese-like format characters, whose 1D and 2D layout expansion would need to be specified coherently somewhere... The effective glyphs used there may be considered representative but not mandatory (there may be several rendering choices, using parentheses, square brackets, or curly braces, or even angle brackets): for maths, most graphic symbols are replacable, provided this is done coherently throughout the text with the same layout rendering rules...
This archive was generated by hypermail 2.1.5 : Fri May 23 2003 - 10:58:34 EDT