accent and marcato shouldn't be quantized

[Mark Polesky]

Wow.  It's been just over 94 (and a half) *days* since my last transmission 
here.  If anyone is curious, I am still alive, but have been away from LP for a 
long time, and I'm finding myself extremely rusty!  If you don't use it, you 
lose it, I guess...  I'm also fully liberated from Microsoft et al, and my 
shiny new Linux desktop just turned 3 (months).  I was hoping to have become a 
Linux guru in that time, but that was unrealistic.  I am, however, using two 
spaces between sentences now, so that must count for something.

Anyway...  I said I'd be back, and here I am.  I had hoped that I would hit the 
ground running, but in reality I have a lot of catching up (and re-learning) to 
do, and perhaps not as much time as I used to have, but that's okay.

On the plus side, I may end up typesetting a new orchestral reduction of a 
relatively obscure woodwind concerto for a New York publisher (details being 
withheld for now).  The existing transcription in print is so riddled with 
errors (almost 2 per measure), that I'm considering just re-doing the whole 
thing myself.


Which brings me to my question:


Currently there are 11 scripts for which 'quantize-position is set to #t 
(accent accentus circulus comma espressivo ictus marcato semicirculus staccato 
tenuto varcomma).  I used my handy ly:filter-alist procedure for this (see 
http://lists.gnu.org/archive/html/lilypond-devel/2009-05/msg00438.html):

#(display (ly:filter-alist default-script-alist 'quantize-position))

In case you didn't already know, quantize-position basically means "center this 
script vertically between two staff lines so it's close to the note head but 
not touching it".  However, in common practice the accent and marcato marks are 
(more often than not) kept out of the staff entirely (see Kurt Stone, pp.5-6), 
ie. quantize-position should not be set to #t for those two scripts.

Is this an oversight, or is there some reason for this?  Would there be any 
opposition to removing the relevant lines in script.scm (currently lines 24 and 
149)?

**********

On a related note, the excellent solution* (by Mats) for controlling this 
individually for each script (using scriptDefinitions) should go into the docs, 
I think.  Are any of these places suitable?

1.3.1 Attached to notes
4.4 Placement of objects
5.4.2 Direction and placement

*http://lists.gnu.org/archive/html/lilypond-user/2009-03/msg00210.html

Thanks
- Mark

ps. The LM reads a lot better than I remember it!