Re: Aw: Re: Re: Pseudo-handwritten font

[Werner LEMBERG]

[please use plain text for your e-mails]

> > What do you mean with `rigid'?
> 
> Some things related to Feta/Emmentaler access are still being
> handled deep inside non-public scheme coding or even C source code
> and consequently very hard to tweak.

Please open bug reports about those issues.  The more detailed, the
better.

> Amongst the most "stubborn" stencils are multi measure rests where
> the glyph names are assembled in the C code by concatenating
> strings.  These things are very hard to manipulate from "outside".

Hmm.  For me, it's OK that glyph names are constructed in the C code.
What's probably missing is a possibility to map those glyph names to
other glyph names.  This belongs to the abovementioned bug report :-)

> Another example: It's easy to change the font used for ChordNames,
> but the accidentals will be Feta accidentals and there is no
> straight-forward way to replace them by anything else.

Ditto.

> > To be really independent you have to add `LILC',
> > `LILF', and `LILY' tables to the font.
> > I can imagine that it is not too complicated to
> > write corresponding scripts for FontForge.
> 
> I don't even know the format/contents of these lookup tables, let
> alone how to coerce FontForge to have them embedded into the font...

It's Scheme code.  If you build lilypond from the source, you get
those files, for example mf/out/feta11.{otf-gtable,otf-table} and
mf/out/emmentaler-11.subfonts.  The FontForge scripts are built by
scripts/build/gen-emmentaler-scripts.py.


    Werner