bounties

[Graham Percival]

On Tue, Jun 15, 2010 at 08:25:27PM +0200, Jan Nieuwenhuizen wrote:
> Op dinsdag 15-06-2010 om 16:50 uur [tijdzone +0100], schreef Graham
> Percival:
> 
> > That has been attempted before... hmm, 2007?  Han-Wen tried to work on
> > lilypond full-time, but there just wasn't enough people offering
> > bounties
> 
> I am considering to offer commercial support and may be able to do
> that on a part-time basis.  However, working on two bounties has
> illustrated that bounty work can be quite tricky.

Indeed; there's almost no relationship between the amount of work
required and the amount of money being offered.

> It would be very
> nice for someone doing this for a hobby and getting to know LilyPond,
> but commercial support requires some level of predictability.

Actually, somebody pointed out (privately) that chasing bounties
is less appealing for inexperienced developers: a $100 bounty
could very well take you 50 hours to complete (i.e. if it's your
first time working on spacing code), making the job $2 / hr.

> Also, if the amount of work is not consistent but takes the form
> of a few thousand euros once a year, you would be very lucky if I
> (or whoever else would take this on) would happen to be available
> within a reasonable time frame to work on those.

Yes.


I'm not trying to discourage people from offering bounties -- it's
certainly better than nothing!  However, there's very good reasons
why programmers don't immediately start working on any issue that
has a bounty being offered.

One idea I've toyed with is seeking a grant to work on lilypond.
Various governments and agencies give research grants; I'm pretty
certain that we could get a grant to improve medieval chant
notation or contemporary non-Western scales or whatnot.  However,
this would probably require
- a bunch of grant applications
- collaborating with some musicologists (i.e. a medieval chant
  expert, or John Cage scholar, or whatever)
- overhead of writing reports about deliverables, giving
  presentations to people, etc.
- etc.
In the process of doing the specialized notation, the developer
might fix a few "normal" bugs as well.


If there was a concerted effort, particularly between the European
academics involved with LilyPond, it could be done, and we might
even be able to fund a full-time developer for 6 months or even a
year.  However, I'm not certain the effort would be worth it --
writing grants is a lot of work; we'd probably have to make
multiple attempts; dealing with the administration of the grant
would be a lot of work; etc etc.

Cheers,
- Graham