Re: 2.15.9 regtests

[Phil Holmes]

----- Original Message ----- 
From: "Phil Holmes" <address@hidden>
To: <address@hidden>; "Keith OHara" <address@hidden>
Sent: Wednesday, August 31, 2011 3:06 PM
Subject: Re: 2.15.9 regtests


> ----- Original Message ----- 
> From: "Keith OHara" <address@hidden>
> To: <address@hidden>
> Sent: Wednesday, August 31, 2011 6:31 AM
> Subject: Re: 2.15.9 regtests
>
>
> > Keith OHara <k-ohara5a5a <at> oco.net> writes:
> >
> > > Two unexpected spacing changes that I will investigate:
> > > beam-extreme.ly  spacing-correction-accidentals.ly
> >
> > The /new/ behavior is correct in these two tests.
> > (That is, the output matches the intent of the code)
> >
> > I forgot to mention, tuplet brackets now span their spanned notes
> > from stem to stem, rather than being extended to include flags as well.
> > This is better, as it is consistent with the brackets over partially-
> > beamed tuplets.    Another un-advertised improvement.
> >
> >
> > > One concern:  Cell counts doubled in many .profile reports !
> >
> > Still a concern.
>
>
> On my pixel comparator I keep a list of how long it took to compile each 
> file.  The differences I see are minimal, so I don't believe there's a 
> general performance problem.
>
> --
> Phil Holmes


I've seen the discussion and issue report.  I'll do a detailed timing check 
and report soon.

--
Phil Holmes