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

distribution of symbols on MSP/MS1


> No, as far as I can see, it is not inherited. The recipe for the \rmoustache
> in MX1 is

> \setslot{bracerightmid}[This slot growing the delimiter `\(\big\rmoustache\)'.]
> % There must be at least an extensible version.
> % This encoding has just that.
>    \varchar
>       \vartop{bracerighttp}
>       \varbot{braceleftbt}
>       \varrep{braceleftex}
>    \endvarchar
> \endsetslot

OK, fine.  Perhaps I was mistaken and I saw it somewhere else.

> PS: How do you like the new font tables ? Over the weekend, it occured
> to me that we could improve the tables further by moving the \curlywedge
> and \curlyvee from MS1 to their \precc and  \succ friends in MSP.  

Please don't.  It will cause problems regarding their availability.

I wanted to come back to this issue and prepare and alternative
proposal for the distribution of symbols in MSP/MS1 anyhow.  

The motivation is this: while design similarity is not a bad idea by
itself, it is easily taken a little too far leading, to the inclusion
of symbols that are difficult to find in non-CM/AMS fonts.  If you put
more emphasis on the availability of symbols (as in the case of the
Mathematica symbol fonts), it seems preferable to break some of the
design similarity groups and reduce MSP to those glyphs required for
compatiblility and those that are easily available, while MS1 would 
be the place for all the more exotic ones.

For example:

* glyphs that are available:

  <, >, \leq, \geq, \lesssim, \gtrsim, \leqslant, \geqslant

  \nless, \ngtr, \nleq, \ngeq, \not\lesssim, \not\gtrsim,
  \nleqslant, \ngeqslant

* glyphs that are NOT available:

  \lessapprox, \gtrapprox, \eqslantless, \gtrslantless, etc.

This would break the \leqslant, \eqslantless design similarity group,
but it would avoid filling MSP with exotic or rare symbols that could
not be implemented without resorting to AMS symbols that don't match
the style of the remaining symbols taken from Math1/Math3.

Hope you'll get the idea.  I'll try to work out the details during
the next few days.

Cheers, Ulrik.


P.S.  Would there be any significant advantage in trying to place all
the LASY glyphs into the MSP, so that not only LaTeX base, but also
LaTeX base + lasy could be implemented in 4 families, thereby making
it possible to provide the full LaTeX 2.09 symbol complement with one
family less than previously required.  It appears that all it would
take is relocating the triangles from MS1 to MSP, while some of the
AMS glyphs, e.g. the sub/superset-or-not-equals group could easily be
moved to MS1 to make room.