At 21.33 +0100 2001-02-19, Frank Mittelbach wrote:
>I don't mind either way and happily leave the decision to David but one
> > Compare with the case for the NFSS fontdef tokens. I understood the meaning
> > of e.g. \T1/pad/m/n/12 (when TeX was showling an hlist) the first time I
> > looked at it, but it took several months before I realised how tricky it
> > really was.
>how tricky was what?
>are you saying that the decision within NFSS was good or bad?
I'm essentially saying that the decision was good, because I understood
what that thingy meant (here we switch to font ...) without any
understanding of the mechanisms behind it. (Here I mean in particular that
for each font, TeX keeps track of a fontdef token which selects that font
and uses it to represent a font switch when showing a horizontal list.) In
NFSS the slashes work fine because you made clear from the start (or so is
my impression) that NFSS attributes are alphanumeric (OK, maybe the sizes
can contain a period), but for keyvals you've already opened the door for
miscellaneous symbols by a frequent use of hyphens.
As for splitting the names: You need to do a catcode change when defining
the macro that is to retrieve the parts of the names, and it's not (in
complete generality, although probably yes in practice) possible to get the
parts right simply by expanding one macro, but with the current naming
scheme there might not even be a unique identification of the name parts.