Re: smart rest positioning - i'm giving money!

[David Kastrup]

Janek Warchoł <address@hidden> writes:

> I will pay 80 Euro to a fearless hero ;-) who will implement smart
> rest positioning - that is, make LilyPond:
> a) render the example used by Daniel just like Gould says it should look,
> b) handle the attached example as well.
>
> The rules are:
> - it must be a built-in, enabled-by-default functionality (not a snippet),
> - the code must pass LilyPond code review and be merged into master branch,
> - it has to be done quite fast, let's say until March 10 (if you think
> you could do it if you just had more time, we can negotiate).
>
> I could prepare more examples to narrow down the desired behaviour.

It would make more sense to quote "just what Gould says it should look
like".  You are apparently assuming that every programmer has a copy of
Gould available to him.

Given that this would appear to take something like a melody-following
approach, it would be a very large change to existing scores.  It would
be overoptimistic to assume that it would make it into master in three
weeks without causing significant regressions requiring a number of
cleanups afterwards.

Since not even the comparatively "harmless" Melody_engraver (which
actually seems like a reasonable candidate for doing the data collecting
legwork for this job as well) is enabled by default, pressing a change
through design, implementation, review, commit, documentation,
corroboration with the whole test suite, enabling by default will allow
about three days for each of those steps.

That seems somewhat ambitious.  In particular since "enabling by
default" alone will require another review, commit, documentation pass.
Add some more melody-line based positioning into the fracas, and we are
at a somewhat realistic GSoC project already, namely a three month
challenge for a student.

-- 
David Kastrup