Carl Sorensen <c_soren...@byu.edu> writes:
On 4/27/12 4:00 PM, "d...@gnu.org" <d...@gnu.org> wrote:
Instead, one can just add a single context mod adding the new grob definition to the Score context definition.
<Snip>
It would make more sense to revert the respective commit creating the snippet completely.
http://codereview.appspot.com/6128048/
What about fixing the snippet in scheme-text-spanner.ly so that it creates the grob the proper way?
Since scheme-text-spanner.ly patches around a _lot_ of global variables (grobs, event classes...), that's not sufficient for fixing the problems with it. I am currently working on repairing the concept of redefining event classes. Takes enough concentration as it is without defining a reasonable interface for extending grobs. One can make the futzing of grobs less broken than Mike did it even without creating a proper straightforward interface, and I outlined how to do it if someone really wants to go ahead with that and/or creating some general-purpose extension code outside of a snippet. But this particular snippet plasters over so many different things that I can just fix them one by one. The grob definitions _are_ already stored in a context mod inside of a context def, so it is totally unnecessary to mess with global variables that will affect future files in the same run. For the event classes, this is harder to unglobalize. But the user interfaces should likely end up somewhat similar, to make things not harder than necessary to learn. So if at all, I am going to fix it after the event classes. But in the mean time, I don't want more crud like define-event-class to migrate into the code base. -- David Kastrup http://codereview.appspot.com/6128048/ _______________________________________________ lilypond-devel mailing list lilypond-devel@gnu.org https://lists.gnu.org/mailman/listinfo/lilypond-devel
