Re: 2.15.41 Regtests

[David Kastrup]

Keith OHara <address@hidden> writes:

> Phil Holmes <email <at> philholmes.net> writes:
>
>> Please look at it and see if these are intended changes.
>> 
> The changes to slurs across line-breaks are intended.
>
> I don't know what caused the changes to slurs, probably the Stem/Flag
> work, but
> I've updated issue 1327.
>
> The stem failing to meet the triangle note-head is issue 2441, so I'll
> update that issue.

I don't think it is issue 2441.  Issue 2441 claims to occur only with
flagged notes, the problem occurs with quarters.  2441 was treated with

commit 3c4b0bec5953774ee9e6e128bdf1edc44bbeae07
Author:     Mike Solomon <address@hidden>
AuthorDate: Wed Apr 4 09:36:28 2012 +0200
Commit:     Mike Solomon <address@hidden>
CommitDate: Wed Apr 4 09:36:28 2012 +0200

    Correctly calculates stem-begin-position for merged stems.
    
    Previously, the lookup for stem-begin-position was potentailly
    triggered before note collision resolution occurred, which meant
    that it ignored Y_AXIS shifts from collision resolution.  Now, through
    using pure property lookups, stem-begin-position is not set until after
    the collision resolution takes place.

whereas the problem occurs since

commit 30f2bb655e6f240a795af9f52c1476849b378ead
Author:     Mike Solomon <address@hidden>
AuthorDate: Fri Jun 15 08:27:17 2012 +0200
Commit:     Mike Solomon <address@hidden>
CommitDate: Fri Jun 15 08:27:17 2012 +0200

    Gives stencilless unbeamed stems an empty Y-extent.

which has been introduced much later.  This is a new regression created
last month.

-- 
David Kastrup