> On Tue, Apr 24, 2012 at 12:41 AM, David Kastrup <dak@...> wrote:
> Using define-event-class makes the file unfit for inclusion in
> multi-file runs of LilyPond since define-event-class permanently
> changes LilyPond.
> I'm a little confused by the word "permanently". Here I understand
> that the new event-class, measure-counter-event, will persist between
> the files if I run a group of files containing my file, but will be
> gone in a new session not involving that file. Any problems which
> might result will only affect that multi-file run?
Basically every use of lilypond-book is a multi-file run. So are our
> I am working on a replacement
> <URL:http://code.google.com/p/lilypond/issues/detail?id=2449>, but
> the pending patch is just one of several changes needed for
> changing the event class hierarchy into a per-parser item instead
> of a global entity.
> Thank you, David--I appreciate this very much.
I doubt that Erik Sandberg would appreciate it. His idea of an
executable-independent event stream is moot if the stream event
hierarchy is session dependent.