[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Replies to H Sami Sozuer



Hi everyone, good to see the list coming to life!

H Sami Sozuer writes:

>Concerning the new decisions on the encoding scheme to be used in
>LaTeX, I will  make the following points:

A small technicality, but it's not just for LaTeX, it's for any flavour
of TeX!

>1 - It is my understanding that new encoding schemes are being
>planned to accomodate the usage of calligraphic, fraktur and
>blackboard letters in math formulas, and that these will
>have to be assigned to distinct families (6 in all).

The calligraphic, blackletter and open lettershapes are being assigned to
the MSi encodings.  This means you can use \cal, \frak and \Bbb in any
TeX document without having to load any extra fonts.  The alternative
would be to provide \cal, \frak, \Bbb etc. as Cork-encoded fonts, and try
to reduce the number of MS fonts to 2.  However, \cal would always have
to be loaded for upward compatibility, so we wouldn't have saved any
famililes, and for applications like the AMS, where they need all three,
we'd have used two more families.

If we weren't faced with the 16 family limit, providing calligraphic,
open and blackletter as separate fonts would be very nice, but 16
families is all we've got!

>I believe
>this is a serious mistake. Many PostScript fonts are available
>for calligraphic, script and fraktur typefaces and these can
>be used in math instead.

This is the basis of the `orthogonality' argument, which the rest of the
math font group is very bored with me going on about :-)

>The advantage of this is that
>one does not have to waste many fonts even if they may never
>used in a particular document.

Yes, but the disadvantage is that when they are used, they end up using
another family (OK, OK, or you do clever macro coding about \mathchoice
and jumping into horizontal mode, at the cost of a lot of time!).  There
is a playoff between using fonts and using families.  And the family
limit is a lot more pressing!

>2 - I believe that NFSS is a major improvement over the
>default scheme used in LaTeX 2.09. However, it leaves a lot
>to be desired for. Namely:

Oh dear, this isn't really the place for a discussion about NFSS, but...

>      NFSS assigns *all* the math fonts at each *size* command.

NFSS2 does slightly better than this, in that it can demand-load any math
fonts that are only accessed by control sequences such as \frak, \cal,
etc.  But it's impossible to demand-load the fonts that are accessed by
\mathchardef.

>      This requires that \mathchardef command (and a few similar commands)
>      be replaced by new commands such that, for example, \Gamma
>      could be defined as \MathCharDef\Gamma7\rm00 .

Exercises for the interested reader:

1. Estimate the increase in running time for mathematical setting if all
   \mathchardef's were replaced by \MathCharDef.

2. Explain how to write such a macro in such a way that it will survive:
   a) being placed in an \edef
   b) being placed in a \write
   c) being used with plain TeX and LaTeX

3. Explain why such a demand-loaded \MathCharDef will still load in three
   fonts even if only the textstyle variant is used, unless _ and ^ are
   redefined.

(Note: :-) :-) )

It's worth noting that any proposed encoding has to work with plain TeX
with no fancy demand-loading of fonts, so we can't propose an encoding
that will only work under any particular font selection scheme.

>I believe that all greek characters
>should be available in all text fonts.

The issue of text Greek is beyond the scope of this list.

As far as math Greek is concerned, a math version (to borrow the NFSS
terminology) can use as many MC-encoded fonts as it likes.  So if you
want serifed, sans serifed, bold serifed and bold sans Greek, just load
the four MC-encoded fonts and you're away.

>>A font made for maths does not have the same adjustments as a font made for
>>text.

>I always thought it was OK to use \rm \it \sf etc in math. Are you
>telling me I shouldn't have?

The problem is that the subscript and superscript positioning isn't great
when the nucleus is taken from a text font, because of the bizarro way
that Appendix G uses horizontal escapement and italic correction to
indicate script positioning.  Anyone can use any font they like in math
mode, but you're not guaranteed wonderful results.  (Take a look at
$\Gamma_0$ for example.)

Alan.