At 09.50 +0200 2005-04-27, Achim Blumensath wrote:
>the MinionPro project was asked about our wishes for an improved font
Say again? Who did the asking and who did the answering?
>Currently there are three things which we miss in
>o Separate values of \bfdefault,... depending on the font family.
>o Support for many-dimensional font shapes. We need:
> - one axis for roman/italic/swash,
> - one axis for u&lc/small caps/spaced small caps/versals,
Are you sure "versal" is a proper English term here? I see what you mean,
but I fear "all-caps" is what the English are stuck with for this concept.
As for "spaced small caps": what is that supposed to be for? Some variation
on the ugly spacing for e m p h a s i s tradition, or what? If on the
contrary it is semantically equivalent to smallcaps, then it doesn't belong
as a separate position on this axis.
> - one axis for text figures/lining figure, and
> - one axis for proportional figures/tabular figures.
That last one is rather a special case of width.
A point to note here is that these first two "o" level items could have a
common solution. \bfdefault is an example of a higher level (than
\fontseries, \fontshape, and friends) font selection, but in LaTeX2e this
is very rudimentary; the problem is IMO that there is no interface for a
font declaration file to supply the necessary information about what can be
chosen. Multiple shape axes could similarly be implemented on top of the
NFSS2 shape axis, if the higher level font selection was more intelligent.
>o The ability to change math fonts within the document.
Doesn't \mathversion provide precisely that?
> \SetSymbolFont and friends are declared \@onlypreamble. This is
> needed, for example, to switch to tabular figures when setting tables.
It could be argued, on the grounds that math mode material is full of glue
which would anyway spoil equal-char-width alignment, that in tables where
equal digit widths are important one should not use math mode, but instead