Re: Google Summer of Code 2015

[Janek Warchoł]

Hello David, Georgy, all,

I got a message from Paul that you're interested in working on
LilyPond during GSoC - that's great!

I think that both musicxml support and tie formatting can be shaped
into GSoC projects, although I expect that these projects would be
quite open-ended.  I also expect that it may be impossible to
_completely_ fix tie formatting during 3 months of GSoC.  Still, any
progress achieved would be very welcome.
As I see it, the formatting engine probably needs to be rewritten -
right now it uses some heuristics to determine tie formatting, and in
my opinion these heuristics are fundamentally broken (but I'm not 100%
sure about that).

As for slurs, I think that they are well outside of the scope of a
GSoC project.  There are a few small things that could surely be
fixed, but writing a generic tool for formatting slurs would take much
more time than 3 man-months (for someone not already familiar with
LilyPond internals).

Long time ago I have collected all of my research concerning ties in a
repository on github:
https://github.com/janek-warchol/tie-crusade
I suggest that you take a quick look at it (follow the first paragraph
of the README) to get an overview of the task at hand.  It probably
needs some cleanup - I'll review and update it over the next two days.

Let's discuss this in more details in two days, shall we?

best wishes,
Janek


2015-03-07 9:42 GMT+01:00 David Kastrup <address@hidden>:
> Georgy Frolov <address@hidden> writes:
>
>> Hello,
>>
>> I too would like to do a lilypond-related GSoC project. I'm doing the
>> final year of my MSc degree in mathematical physics, sing in a choir
>> and occasionally use lilypond to typeset some choral scores; not yet
>> familiar with its internals though.
>>
>> I've been working much with numerical codes in C++ and python,
>> including much interpolation, optimization and tweaking heuristic
>> models, so the tie/slur appearance project seems a great fit for me.
>> It's also something where I  want improvement as a user of lilypond.
>>
>> Could you comment why it may not be suited for GSoC?
>
> Personally, I think it is a reasonably well-contained piece of LilyPond
> to work on.  I think it would be good if you had a copy of "Behind Bars"
> at your hand and hopefully a reasonable bout of piano music from
> renowned publishers in some hand-engraved editions: piano extracts for
> choir tend to be newer, not with large challenges to slur appearance,
> and not with a particular focus on piano note quality.
>
> Of course, LilyPond manages to have ties/slurs look ugly even in
> comparatively simple cases.
>
> Starting point would be ties, I think: limited task that _really_ should
> be computer-soluble.  And the code already gets all the data it needs
> for making its decisions, so you can focus on doing the work rather than
> fighting the data flow (the data flow in LilyPond is _way_ complex).
>
> The disadvantage is that it's a bit of a hit&miss project where I don't
> see a lot of leeway for "gradual" improvement.  So at the end of the
> project, you'll either have something to show.  Or not.
>
> If you want to have a good impression of what kind of code you would be
> working with, the main work is likely currently done in
> lily/tie-formatting-problem.cc.  Unifying the tie formatting and slur
> formatting would be nice in theory but far beyond a GSoC project.
>
> --
> David Kastrup
>
> _______________________________________________
> lilypond-devel mailing list
> address@hidden
> https://lists.gnu.org/mailman/listinfo/lilypond-devel