Re: bug report: partcombine and multi-measure rests

[Gilberto Agostinho]

Hi Simon

> there’s no point in discussing off-list.

My bad, I hit the wrong button. Just for the sake of having the 
conversation registered in the list, I am copying my previous message to 
the bottom of this one.

>> The output is ambiguous
> It is inconsistent

Yes, inconsistent is a much better word.

> I forgot to mention: You can change this indication

Thanks! I also like that better.

> If no one else will give some more founded comment on this, I’ll 
raise an issue from this.

Great, thanks for all the help and patience.

Cheers,
Gilberto

* * *

Hi Simon,

Sorry for the misunderstanding, I installed 2.19.28 and indeed there is 
a Solo indication above that b2~, as well as its stem is in the correct 
place. The rest r2 just before it is still in a wrong position though, 
see: 
http://s14.postimg.org/a6dhivry9/Screenshot_from_2015_10_18_17_22_44.png

> So as I understand it the main point of your request would be the 
behavior of \partcombineApart, if one voice has rests.

Yes and no: yes, that has always been my point, but no, that is not a 
request, that is *already* how \partcombineApart works. See the 
following case in which the \partcombineApart command is at the same 
position as the R1*N:

\version "2.19.28"

{
  \partcombine
  {\partcombineApart r2 b'2~ | 1 | \partcombineAutomatic c''1 |}
  {R1*2 | f'1 |}
}

Which outputs this (no solos, no single voice, just two voices apart as 
expected, regardless of full measure rests or not): 
http://s14.postimg.org/sw5gzmkox/Screenshot_from_2015_10_18_17_22_51.png

So my point is that it's broken if and only if there are R1*N happening 
before the command \partcombineApart.

> Which it can, and though it may not be up to your preferences, the 
output is unambiguous, understandable and correct.

The output is ambiguous because the output of LilyPond will differ if 
the R1*N coincide (or happens later) than the \partcombineApart command 
or if it happens before. It's also not correct as shown by the rest 
circled in red in the first screenshot on this message. Below is an even 
more concise tiny example showing the problem, wouldn't you agree that 
the output should be at least equal in order to argue that it's 
unambiguous?:

\version "2.19.28"

{
  \partcombine
  {R1 | \partcombineApart r2 b' | \partcombineAutomatic c''1 |}
  {R1*2 | f'1 |}
}

{
  \partcombine
  {R1 | \partcombineApart r2 b' | \partcombineAutomatic c''1 |}
  {R1 | R1 | f'1 |}
}

Producing: 
http://s14.postimg.org/m6yxjlzcx/Screenshot_from_2015_10_18_17_22_54.png

> Sorry I can’t help more. I may raise an issue though, but I’d like 
for someone else’s opinion before.

Sure, thanks anyway.

Cheers,
Gilberto