Re: Staff Stretching problem in 2.13.50+

[Graham Percival]

On Wed, Mar 09, 2011 at 06:10:46PM +0000, James Lowe wrote:
> Graham,
> 
> From: Graham Percival <address@hidden>
> >> \relative c'' {
> >>   \time 3/8
> >>   \compressFullBarRests
> >>   R1*19*3/8 \break| cis4. | d8 r r | R1*2*3/8 | cis4. | d8 r r | \break
> >> R1*20*3/8
> >> }
> >
> >In which universe do you consider this to be a "tiny example" ??
> 
> Umm...6 bars of music of which 4 show the problem and the other 2 are
> necessary to show err the problem.

Honestly, all those * look complicated to me.  Granted, the only
time I did typesetting of weird time rhythms was in 2002... but it
still looked complicated enough that I ignored this discussion the
first two times it came up.

Be kind to people like me who don't know lilypond as well as you.
:)

> >Is
> >the \time necessary?  Are the *19*3/8 necessray?
> 
> As far as I was concerned, Possibly

If the answer is "possibly", then if I were in the bug squad, I'd
ask you to revise the example with no qualms whatsoever.

Look, it took me about 60 seconds to generate the Tiny example --
and example without weird rhythms.  Now, that example *also*
failed the "Tiny test" -- Keith created an example without any
\compressFullBarRests or multi-measure rests.  Whoops.  So the Bug
Squad should have rejected *my* version as well.  Or at least, I
should have gotten a snarky "please see the guidelines for Tiny
examples" email to teach me to spend 120 seconds next time!

> Well Keith did have a 'tiny example' but then started on about \markup
> which I didn't have in my example (there are no \markup) and yes I NOW
> know what he meant by \markup but that wasn't clear to me. So there were
> two examples in my email. Had I known what I know now then I wouldn't have
> bothered trying to contradict Keith and used his example, but how was I to
> know? For all I knew \header {} was something 'special' something
> 'different' and might have been key.

Sure -- so the thing to do is to test it.  If you commented out
the \header{} and put a \markup{} in there instead, then you would
either see the same behaviour (in which case it wouldn't matter),
or else you'd see different behaviour (which could be good info
for the report).


Nobody's perfect.  As discussed above, *I* screwed up when trying
to make a Tiny example.  I'm not (trying to) badger you
personally.  I'm just trying to get *everybody* into the right
mindset:
1. bug reporters: try to make it as small as possible.  Don't just
assume that feature XYZ is actually necessary to reproduce the
bug; test if it's necessary by commenting it out.
2. bug squad: if you don't think that a report is truly Tiny, then
officially ask the reporter to make it smaller.  Or, if you're
feeling merciful and/or have nothing else to do for your 15
minutes, go ahead and turn it into a Tiny example, then explain to
the reporter how he could have made the example smaller.

Cheers,
- Graham