Re: a present for Graham

[Graham Percival]

On 2/19/11, Han-Wen Nienhuys <address@hidden> wrote:
> Also, could we dispense on esthetic changes on developer-only
> parts (like the regtest) ?

This was not an esthetic proposal.  I honestly (but mistakenly!)
thought that the lack of complete bars made the spacing more fragile,
which made it harder to spot actual regressions.  The amount of
regressions between 2.12 and 2.13 is the only reason that 2.14 wasn't
out in 2010 (and probably won't be out until this summer, if then!),
and I am always looking for ways to make our development process
smoother and more reliable.

Unfortunately my initial belief (bars->fragile spacing) was false, so
the conclusion was false.  But the intent was not esthetics.

> I'm not sure of the exact reasoning behind 'frogs' issues, but before
> they should be packaged up to be small, so nobody spends many hours on
> them.

The intent behind the "frogs" (as a group of people) is to give newbie
contributors a place to get mentored, teach each other "simple" things
like how to use git which they might feel shy about asking on the big
scary lilypond-devel, etc.  There's a separate mailing list, which
isn't used as often as I would like:
http://frogs.lilynet.net/

The intent behind the "frog" issues is that they *are* (relatively)
small, and require (relatively) little knowledge of lilypond's
internals.  In short, they are intended to be ideal tasks for new
contributors.

If the end-of-barline issue had been a valid one -- which it wasn't --
then I think this would have been an ideal Frog task.  Every potential
lilypond developer already knows how to add a few rests to a .ly file;
the regtests can be compiled individually, so there's no build issue
problems.  It would have been a nice introduction to git and creating
patches.

I will send a warning not to trust our issue tracker.
- Graham