Re: self-alignment-X has negligible effects on dynamics

[David Kastrup]

David Kastrup <address@hidden> writes:

> David Nalesnik <address@hidden> writes:
>
>> On Mon, Jul 28, 2014 at 9:25 AM, Noeck <address@hidden> wrote:
>>
>>> Your snippet produces a different output here with 2.19.8
>>> (i.e. more like what I would expect)
>>>
>>>
>> Yup--and you get the same with 2.19.9.
>
> Easy to suspect
> <https://code.google.com/p/lilypond/issues/detail?id=3978> here.
>
> I'll see what bisection does.

09412c25cce3fdffce7cefe491cb3849ece0ae3e is the first bad commit
commit 09412c25cce3fdffce7cefe491cb3849ece0ae3e
Author: Janek Warchoł <address@hidden>
Date:   Sat Mar 30 00:11:28 2013 +0100

    Clean up DynamicText horizontal alignment.
    
    Until now, DynamicText alignment was messy, as there were 3 different
    callbacks involved - some of them interacting in a confusing way:
     - in define-grobs.scm, X-offset property was initialized to
       ly:self-alignment-interface::x-aligned-on-self,
     - Dynamic_engraver called set_center_parent on every DynamicText
       attached to a note, so that half a NoteHead width was always added
       to its X-offset (producing confusing results for example when
       user requested dynamics to be left-aligned),
     - DynamicTexts living in a Dynamics context used a completely
       different offset callback, which aligned them on NoteColumns.
    
    Since aligned_on_parent is now able to correctly align grobs
    with PaperColumn parents (issue 3254), we can use it for all
    DynamicTexts and have a single interface for the job.
    
    Expected changes in output: DynamicTexts in Dynamics context
    aligned to suspended noteheads may be placed up to 1/4 NoteHead
    width further to the right.  This shouldn't be a problem.

:040000 040000 38a27adc238e97dc202923e87d75a7074c19688c 
174d4ae439a343221e4c17ce01facee7e28bfc00 M      lily
:040000 040000 6f9024af65d9bd279d1bf3423de22dc262aee908 
80b24c0099842d727d6a76ed28d9f5fc75fd3d57 M      ly
:040000 040000 8ef68b110e3449080678f7d13c97e7074c5f0f20 
f1f2f44ff454c3568a4286e82a7c611ef97c66a0 M      scm

Indeed, the first commit of
<URL:https://code.google.com/p/lilypond/issues/detail?id=3978>

-- 
David Kastrup