Re: using Feta font in a research project/HTML5 canvas?

[Graham Percival]

On Mon, Apr 05, 2010 at 10:50:43PM +0000, Geoff Chirgwin wrote:
> Since our first implementation is in HTML5, we needed a
> lightweight way to render music notation to the HTML5 canvas
> entirely client-side in the browser.  Since we know of no
> existing software that does this, we've started writing our own.
> We're calling it Rastral.  

This is probably going to be an order of magnitude more work than
you think it will, unless you have a very strict feature set in
mind.  For example, if you only want to deal with pitches,
durations without tuplets, articulations, and slurs without caring
about collisions, then I think it would work.  If you want to do
much more than this, I wouldn't be surprised if Rastral ended up
being more work than the handwriting recognition part.


> There are two immediate questions about licensing (which may spark more):
> 1) Is it permissible to embed the Feta font in a web page and/or redistribute 
> it?
> 2) Is it permissible to embed fetaList.cxx in a web page and/or redistribute  
> it? (we're using this for the font name to code point mappings, and, believe 
> it
> or not, it interprets verbatim as Javascript)
> 
> We're assuming we'll GPL Rastral.

Don't just assume this -- decide on it right now.

- if you put Rastral under GPLv3, then there are no problems using
  portions of LilyPond.  However, if you (or your university)
  wants to commercialize your work, there might be problems with
  GPLv3 software.

- if you put Rastral under GPLv2, you can only use feta stuff from
  a few months ago; the later material is under GPLv3.  This
  could give you more leeway with "web apps" if the rendering
  is done server-side.  Since you've explicitly mentioned
  client-side rendering, though, this point is moot.
  The same problems with potential commercializations apply.

- if Rastral is not going to be GPL'd, then you certainly cannot
  use fetaList.cxx, and off the top of my head I doubt that
  you could use Feta either.  (I don't claim to be certain
  about the latter point)


> We're not expert in software licensing, but  we're hoping to
> release this software in the spirit of open source, in such a
> way that the Lilypond developers approve and that the community
> as a whole may benefit.  Any input on this is appreciated.

I would recommend GPLv3, then -- but make sure that your
collaborators agree to this as well, before progressing too much
further in this project.


On a personal note, I'm a PhD student in the Center for Music
Technology at the University of Glasgow, and I'm quite interested
in this work -- both as a user, but also as a potential developer.
I'm currently generating notation for musician listening
experiments using flash, but this is a quite unsatisfactory state
of affairs, both from a licensing/openness standpoint, but also
from an ease-of-development standpoint.

Cheers,
- Graham