Re: Font naming rears its ugly head again

[karl@owl.HQ.Ileaf.COM (Karl Berry)]

    if you receive a dvi file containing the font ptmrq,
    whose ptmrq should you use?

ptmrq defines a unique font -- Adobe Times Roman in the Cork encoding.
It shouldn't matter what generated it.  If it does matter, then there
must be some other difference in the font.

In your case, I guess it is letterspacing.  So the way to fit this into
the Karl scheme (which I wish we could call something else :-) is to
make a new variant letter for whether the thing is letterspaced or not.

I suppose I could give up one of my precious two remaining variant
characters (7 and 8) for that, if you are really married to
letterspacing your font, which I bet you are ...

    use a `unique' prefix for the fonts I generate (say `f1' for `fontinst') 

Can't one generate a font that would be identical with the current
ptmr.tfm (say) using fontinst?  Conversely, I bet one can get a TFM/VF
identical with yours using afm2tfm.

I think the method of generation is not important to encode -- it's the
final resulting font that counts.

     use the remaining three characters (after the `f1'
    prefix and the 3-letter family name)  to encode a 15-bit number

As long as you realize this doesn't cover the possibilities, either ...
The only way to do that is to do what Don (and you) suggest elsewhere,
just go to variable length names.

If you adopt this, I advise a single character for the source; I'm not
even close to using up the 36 possibilities for the source yet.

      Don> Before I send you a DVI file, I process it with a program which
      replaces instances of rptmr with Times-Roman-etc. You then run
      on your system a program which reverses the process and creates
      a DVI file where the names Times-Roman-etc. is replaced with
      Times-Roman. 

I think this is a good idea.  web2c and my other programs have had a
mapping file (texfonts.map) to allow precisely this for a couple of
releases now, but as far as I know no one is using it.  (I'm not.)

The only problem is that the long Times-Roman-etc filenames have to be
replaced with *something* for the actual filename -- and so why not make
the actual filenames the same on different systems?  At least, that's
the reasoning I make to myself.

    Alan> simple as an agreement between TeX implementors as to a standard
    character to use as a directory separator

That doesn't sound simple to me, because why do TeX fontnames
necessarily have to map onto particular directory structures?  Unless
you want to limit names to 8 characters, they won't ...