Re: Regression test analysis - 2.13.35

[Neil Puttock]

On 30 September 2010 09:57, Phil Holmes <address@hidden> wrote:

> 1) In harp-pedals-sanity-checks.ly we get:
>
> +warning: Harp pedal diagram contains dividers at positions (4). Normally,
> there is only one divider after the third pedal.
> +warning: Harp pedal diagram contains dividers at positions (1 3 5).
> Normally, there is only one divider after the third pedal.
> +warning: Harp pedal diagram does not contain a divider (usually after third
> pedal).
> warning: Harp pedal diagram contains 5 pedals rather than the usual 7.
> -warning: Harp pedal diagram does not contain a divider (usually after third
> pedal).
> -warning: Harp pedal diagram contains dividers at positions (1 3 5).
> Normally, there is only one divider after the third pedal.
> -warning: Harp pedal diagram contains dividers at positions (4). Normally,
> there is only one divider after the third pedal.
>
> There's similar in harp-pedals.ly.  Is this a bug or an update to fix a bug
> that's caused different error messages.

It's a bug.

Joe's recent fixes for outside-staff estimation sort markups (and
other outside-staff grobs) based on their 'outside-staff-priority.
This causes the stencils to be evaluated in a different order, which
then causes the warning messages from the harp pedal markup command to
reverse their ordering in the regtest.

Here are two minimal examples showing what's happening:

\version "2.13.35"

\relative c' {
  c1^\markup \musicglyph #"1st"
  c1^\markup \musicglyph #"2nd"
  c1^\markup \musicglyph #"3rd"
}
%{
  warning: Cannot find glyph 2nd
  warning: Cannot find glyph 3rd
  warning: Cannot find glyph 1st
%}

\relative c' {
  c1-\markup \left-brace #0
   \override TextScript #'outside-staff-priority = #0
  c1-\markup \left-brace #-1
}

%{
  warning: no brace found for point size -1
  warning: defaulting to 10.5 pt
  warning: no brace found for point size 0
  warning: defaulting to 10.5 pt
%}

Cheers,
Neil