Re: 2.15.16 regtests - pixel comparator

[Keith OHara]

Phil Holmes <email <at> philholmes.net> writes:

> 
> A few differences of note - in particular span-bar-spacing where we now get 
> a collision between the bar line and natural symbol which did not exist in 
> 2.15.13.  The other attached images are for comment - but I think there's a 
> critical regression shown with span-bar-spacing.
> 
note-head-aiken
 The .pdf output looks fine to me

mozart-hrn-3-page2
baerenreiter-sarabande
 I raised issue 2005 for these, and confirmed they come from the fix for issue 
1951.  I see no better workaround than a manual extra-offset on every affected 
staccato mark.  These side-effects seem unintentional and worse than the 
original problem in issue 1951.  If the bug squad agrees, please adjust the 
priority of issue 2005 appropriately.
 
beam-break
 Different, but still good

span-bar-spacing.png
 Already listed as issue 2000.  Confirming that it came from the fix to issue 
1846.

slur-script.png
  Different, but still good

fret-diagrams-fret-label
 David's commit 8a459d03 changed the regression test itself.
(Whatever \revert TextScript #'(fret-diagram-details label-dir)
is supposed to do, it does not revert the direction fret number.
I can simplify the \overrides and get the original output, so I'll push that
within a couple days unless somebody else has a better fix.)