Re: Wrong glissando for notes with duration longa

[Thomas Morley]

2017-10-04 22:39 GMT+02:00 David Kastrup <address@hidden>:
> Thomas Morley <address@hidden> writes:
>
>> 2017-10-03 18:18 GMT+02:00 David Kastrup <address@hidden>:
>>> Thomas Morley <address@hidden> writes:
>>>
>>>> Version: 2.16.2: longa-note-head-stencil Y-extent: (-0.550006 . 0.550006)
>>>> Version: 2.19.65: longa-note-head-stencil Y-extent: (-0.72 . 3.519982)
>>>>
>>>> It's caused by:
>>>>
>>>> commit f494ef0afc8aed378bd778cd51a294804a8ee5d4
>>>> Author: Phil Holmes <address@hidden>
>>>> Date:   Mon Jul 29 16:09:29 2013 +0100
>>>>
>>>>     Corrects bounding box for longa (Issue 826)
>>>>
>>>> related is:
>>>>
>>>> commit b27354f5fd0df898310a354c4031c2c701ecebaa
>>>> Author: Phil Holmes <address@hidden>
>>>> Date:   Mon Jul 29 15:21:46 2013 +0100
>>>>
>>>>     Fixes bounding boxes for breves, improves longas (Issue 3480)
>>>>
>>>> Those patches tackle the skyline-problem, which causes the cut off 
>>>> previews.
>>>> So it's no good option to revert them.
>>>>
>>>> Maybe glissando should look at staff-position?
>>>
>>> That sounds sensible.
>>
>> Thinking more about it I'm not longer convinced it would be the best
>> fix.
>
> Interesting.

Why?

>> We have other glyphs exceeding their bounding-box, like the forte-sign, see:
>> \markup \box \musicglyph #"f"
>
> They don't feature in glissando end point positioning, do they?

Ofcourse not, take it as an example where the skylines catche all
relevant stuff even if exceeding the bounding-box.

>> But the skylines are done properly:
>> #(ly:set-option 'debug-skylines #t)
>> { c'1_\f }
>>
>> Though, not for mensural-glyphs:
>> #(ly:set-option 'debug-skylines #t)
>> \new MensuralStaff { \stemDown c'\longa_\f c' }
>>
>> So I tend to get the skyline-code improved to deal properly with
>> default-longas and similar mensural-glyphs as is done for the
>> forte-sign.
>>
>> Not that I'd have an idea how to do...
>
> I don't really see the connection to glissandi here.  Isn't that more of
> an independent problem?


Well, if we would revert the above mentioned patches and improve the
skyline-code the glissando and the cut-off-problem would be solved,
imho.
Or am I overlooking something?

Cheers,
  Harm