Re: BendSpanner-engraver

[Federico Bruni]

Il giorno dom 2 ott 2016 alle 19:30, Thomas Morley 
<address@hidden> ha scritto:
> Hi Marc et all,
>
> 2016-09-18 16:55 GMT+02:00 Marc Hohl <address@hidden>:
> > >  I'd prefer to solve that slow down first, otoh it's little effort 
> > > to
> > >  do a new tarball.
> >
> >
> >  If the estimated work on the slowdown is not that huge, I think it 
> > is ok
> >  to get a grip on that first and release a new version afterwards.
>
> it turned out it was _much_ harder then expected.
>
> I ended up writing a second engraver which sets the new defined
> grob-property "TabNoteHead.bend-me", which decides whether a string is
> bendable.
>
> To get there, I used the context-property `noteToFretFunction'. Using
> it in the engraver leads to the need to give it the appropiate
> arguments.
> Disadvantage: Actually `noteToFretFunction' is now used twice, in my
> BendMeEngraver _and_ in the default Tab_note_heads_engraver.
> If someone knows a better method please shout!
>
> Nevertheless, it's now 4 times faster than before.

Thanks Harm, indeed it's faster. These are the times I got in 5 (small) 
pieces:

1st version --> 2nd version
2,2" --> 1,8"
2,3" --> 2,1"
3,3" --> 2,3"
2,0" --> 1,7"
1,7" --> 1,5"

I noticed that I get this warning:

warning: no notes to start a bend from found. If you want to bend
an open string, consider override/tweak like: TabNoteHead.bend-me = ##t

if I don't specify the string of a default open string like b on second 
string:
ais\startBend b\stopBend

So I must use b\3\stopBend even if it's the note where the bending 
_ends_?
The bend can take place only between notes (explicitely) on the same 
string, right?