Re: Beam subdivision bug in 2.15.22

[Carl Sorensen]

On 12/16/11 12:12 PM, "Xavier Scheuer" <address@hidden> wrote:

>On 16 December 2011 19:48, Thomas Scharkowski <address@hidden>
>wrote:
>> \times 2/3 { b16 b b } b8 b4 b b
>>
>> 2.15.22 produces a (wrong) beam subdivison.
>> 2.14.2 output is correct.
>> See attachments.
>
>Result is fine with 2.15.20.
>I did not make a git-bisect but I suspect Carl's fix for issue #11
>to be the cause of this.
>http://code.google.com/p/lilypond/issues/detail?id=11

I am quite sure that this is the case.

>
>I'm not a git expert but it seems Carl's commit was pushed a few hours
>after 2.15.20 release, so issue #11 should have been tagged as
>fixed_2_15_21, isn't it?

That is probably correct.

>
>A better wording for this issue report would be something like
>"unexpected beamlet in tuplet sixteenth".  But maybe it is better to
>reopen issue #11.

No, do not reopen issue #11.  This is *not* issue 11.  It is a separate
issue caused by the fix to issue #11.

We should *never* reopen fixed issues unless we see that the issue was in
fact, not fixed.

The regtest for issue 11 shows that issue 11 has been solved.  That fix
introduced a regression that was not tested by the existing regression
tests.  We now have a new problem, and a new potential regression test.
When I've fixed the new problem, the regtests will show that 11 is *still*
fixed.

So a new issue should be added for this bug.

Thanks,

Carl