Re: partcombine/divisi framework - let the fun begin!

[Simon Albrecht]

Am 29.08.2014 um 08:53 schrieb Keith OHara:
> Janek Warchoł <janek.lilypond <at> gmail.com> writes:
>
> > I'm working on putting together a set of functions for handling parts
> > that split into multiple voices and/or staves
>
> > Attachment (definitions.ily): text/x-lilypond, 2359 bytes
> > Attachment (example.ly): text/x-lilypond, 825 bytes
> > Attachment (example.pdf): application/pdf, 33 KiB
> I like the way 'definitions.ily' uses <<>> parallel music constructions
> to ensure that the parts re-synchronize even if there is a note missing
> in the input.
>
> You should do something similar for the \together segments.
> As you have it now, the together segments are un-tagged, so are included
> in all three staves (two of which are not printed).
> But if someone uses these definitions with music that uses tags for
> some other purpose, the tagged music will be left out.
A problem which always comes up when trying to use tags in more than one 
"layer". A workaround is the use of \removeWithTag in place of 
\keepWithTag, e.g. replace
\keepWithTag together.part
by
\removeWithTag divI.divII.score
, if you understand what I mean. (in complex projects, finding the right 
solution so all the included files and tags and layout definitions etc. 
work together well is likely a headache; but when the solution finally 
is there and it compiles as it should, relief is great...)

Or is it worth a thought to add support for different layers of tagging? 
For example
- \tag, \tagA and \tagB  or
- \tag 1 foo and \tag 2 bar (optional third argument).

Best, Simon