Future of the `Karl Berry Font Naming Scheme'...

["Melissa O'Neill" <oneill@cs.sfu.ca>]

Recently Berthold posted a reference to support files for the LY1
encoding, which included a collection of TFM files for LY1 encoded
postscript fonts.  This files did not use the `Karl Berry Font Naming
Scheme', and this gave me pause for thought...

Should we care so much about operating systems that are so amazingly
primative that they can't handle more than 8+3 characters? Even using
the Berry scheme, my own fonts for Kepler can have names like pkplci9dx11,
which is 11 characters (actually, my fonts get worse names like
pkplci9dx1095). Since we can have fontname mapping files, perhaps we
should just pick random garbage letters on ancient DOS machines for the
file names and then name the files properly in the map file.

It seems to me, that naming fonts as they're named in the .fd file would
make much more sense. Thus, the font would be called T1-pkp9-lx-itsc-1095.
This would make the \DeclareFontShape line in the FD file pretty obvious,
and would only be a small about of blow-up.

I guess at the very least, there is nothing to stop me using a naming
scheme like this for my own fonts.

What do people think?

    Melissa.