# Re: on skewchar = space

• To: Ulrik Vieth <vieth@thphy.uni-duesseldorf.de>
• Subject: Re: on skewchar = space
• From: Matthias Clasen <mclasen@sun2.mathematik.uni-freiburg.de>
• Date: Wed, 1 Oct 1997 17:44:32 +0200 (MET DST)
• Cc: math-font-discuss@cogs.susx.ac.uk


> Have a look at
>
>  http://www.unicode.org/Unicode.charts/normal/U+0370.html
>
>for a Greek font table.  The \varbeta is in slot U+03D0 and seems to
>look like a normal \beta without a stem descending below the baseline
>and the upper bowl being perhaps a little more curly than usual.

Many thanks for that pointer! Lots of glyphs of which I only knew the
names suddenly come to life.

>Now, the question remains whether there are any technical reasons
>against using character 0 for visible symbols.  I seem to recall that
>the 8r-encoding of PostScript fonts avoided the use of characters
>0, 10, 13 (i.e. NUL, LFD, RET) in case we meet dump software''.
>Since we did not worry about slots 10 and 13 before, I really don't
>seem why we should treat slot 0 any different.

If the dumb software' that cannot grok NUL, LFD or RET is the same
that demands a space in 32, then this argument could be read as saying
that we should either reserve all four slots (0, 10, 13 and 32) or dump
the space in  slot 32 as well. Perhaps we need to be more specific about
the dumb software' which we might meet. Is there any driver which has
problems with NUL, LFD, RET ? Does it also demand space in slot 32 ?

Anyway, since we have decided to reserve one slot for the skewchar, it
could as well be slot 32. The only technical reason for not putting an
ordinary glyph in slot 0 which comes to my mind would be the special
treatment of \uccode/\lccode 0 in TeX, but this wouldn't be a problem for
math fonts (seemingly it is not even a problem in text, if you put a
non-letter in position 0, as the ec fonts do).

As a side-effect, this change will eliminate the clash between the
reservedness of the skewchar slot and the desire to keep the accents in
their T1 positions (in the MSP encoding). So, if nobody finds technical
obstacles, I will move the skewchar to slot 32 in all fonts. The next
question is what to put in slot 0 instead; I think that the font tables
of the extensible fonts can be improved by just shifting the bigops
to close the gap at 0.

>I've just grep'ed through the mail archives concerning the origin of
>the convention to use of slot 0 as the skewchar.  It appears to me now
>that there was no technical reason for choosing this particular slot,
>just the intention to use a character not intended for typesetting
>and simplicity:

Thanks for the history work.

Regards, Matthias

--
Matthias Clasen
Albert-Ludwigs-Universitaet Freiburg
Institut für Mathematische Logik
79104 Freiburg, Germany.   Tel.: +49 (0) 761-203-5606