[Groff] Re: Composed glyphs - quick and dirty solution

[Werner LEMBERG]

> TeX has huge amount of time invested in 8bit fonts, 8bit encodings,
> virtual fonts - and commands to create accents by hand look natural for
> mathematicians who use them in formulae.
> TeX is not likely to go Unicode any time soon -

Ha!  Are you aware that I've written the CJK package for TeX to
process Asian scripts, with all the nifty details like suppression of
linebreaks before CJK punctiation characters, and that I've recently
added full support for Unicode up to U+10FFFF, together with automatic
Unicode bookmark creation?  Everything is on the macro side, of
course, but it *works*.

> > .de ps-accent
> > [...]
>
> Will this code work for marks below base? ;)

I have no idea :-) It shouldn't be too difficult to handle this case
too, or to provide a different macro for it.

> Why the skew is computed for 'x' and not for accent glyph ('\\$2')?

Again, I've no idea :-) The code snippet was taken from ps.tmac,
without analyzing it in detail.  I've simply forgotten how it works.

> > No.  For example, my AvantGarde-Book.afm file version 001.006
> > (directly from Adobe, I think) has this:
> >
> >  StartComposites 56
> >  CC Aacute 2 ; PCC A 0 0 ; PCC acute 183 163 ;
>
> Are there Courier/Helvetica/Times versions with CC lines?

Yes.  All files have `CC' lines.

> OK, the question of composition tables for base fonts seems to be
> better shelved until Bruno and you shout military style Go-go-go to
> testers and send them into mine fields of new features!

Hehe.

> PS. I did not quite follow the developments in hyphenation support.
>     What has to be done to, say, russian hyphentaion file to adapt
>     it to the next groff version?

Currently, there are no changes w.r.t. hyphenation.  Just take, say,
ruhyphal.tex and use a bunch of .hpfcode lines to map its koi-8
encoding to iso-8859-5.  Then set up iso-8859-5 as the input encoding
as usual.


    Werner