Re: Font: Including the jazz font for chords (issue3972048)

[Simon Tatham]

address@hidden removed from cc-list on the
grounds that I doubt this is any longer on-topic for any actual code
review]

James Lowe <address@hidden> wrote:
[Gonville]
> I'd contact the author directly. Last year when I helped write what
> documentation we have on it in our NR, and he was very helpful, and patient
> with me. I am sure he would be flattered that his font could be in the main
> source.

I'm actually still reading here, even if only intermittently. Hello :-)

Indeed, I would certainly be pleased to see Gonville shipped as
standard with Lilypond; at least one person I know has mentioned to
me that they like the font but are currently put off using it by the
horrible installation procedure.

Gonville's MIT licence is compatible with the GPL, so I don't see
any legal reason they _couldn't_ be shipped together. The bit about
having the 'compiled' OTF/TTF/PFB/etc form of the font be completely
unrestricted is primarily there for the convenience of people
incorporating it into PDFs who don't want to have to worry about the
font copyright as well as the music copyright. (If I remember
rightly, isn't there an exception in the licensing for Feta for the
same use case, to avoid GPL-contaminating users' output PDFs?)

Failing that, I'd still like to see it be possible to make the
Gonville installation procedure easier than the one I currently
document on the web page (of constructing a symlink mirror of the
Lilypond data directory with all the font files replaced, and
pointing LILYPOND_DATADIR at that directory).

I confess I haven't kept up with all this recently, but as I recall,
the situation last time I looked was that it was now possible to
drop one Gonville font file into the existing Lilypond fonts
directory and issue a directive in Lilypond source which switches
over _almost_ all the glyphs to that - but the parts of the font
that fit into normal ASCII (time signatures and dynamics) still
didn't switch, and couldn't be persuaded to switch by any means
short of the LILYPOND_DATADIR hack.

Is that still the situation? If the time signatures and dynamics are
now fixed, I'd be very happy to relegate the hacky approach to the
status of a historical horror and document a far simpler method that
works with modern Lilypond.

Cheers,
Simon
-- 
Simon Tatham         "infinite loop _see_ loop, infinite"
<address@hidden>     - Index, Borland Pascal Language Guide