Manual custos and automatic custos (on clef change) have inconsistent spacing

[henryso]

The manual custos of a figure like (g+::c3) has additional space behind it compared to (z0::c3). It is actually coded this way: \GreManualCustos adds the space before calling \GreCustos. As such, it works as coded.

Neither I nor @rpspringuel know whether this is the desired behavior or is just some inconsistency that worked its way into the code (and thus would be considered a bug).

Does anyone know the history here? Is this a bug?

[henryso]

If no one knows the history behind this (probably @eroux would be the best to speak to that), then does anyone have an opinion on the correct behavior?

My not-very-strong opinion is that we should set that space to 0 by default and apply it before both the g+ and z0 custos, but not to the end-of-line custos.

[henryso]

Sr. Maria Ruth OSB, who brought up this issue on the mailing list, says this:

My simple thoughts are that all custos should be equally treated, obviously when they are at the end of a line shouldn't be followed by a space.

[eroux]

Neither do I have a strong opinion, I guess treating all custos inside a line equally is consistent, and maybe set the space to 0 is good too...

[henryso]

I'll work on this and submit a pull request. Should this be in 4.0.0?

[henryso]

It turns out that spacebeforecustos is actually applied to the end-of-line custos, so zeroing it out is a bad idea. However, if I apply it to both the g+ and z0 figures, the space is very wide because (I think) the end-of-element space (whichever case it happens to be), which would have been swallowed at the end of the line, is placed before the custos (and spacebeforecustos) in the middle of a line. Can anyone confirm or disprove this finding and, if true, give me a hint on how to combine the end-of-element space with the spacebeforecustos so that the space is the maximum of both spaces and not the sum of both spaces?

[rpspringuel]

I haven't done any checking to see if what you're getting is true (though it does sound right), but perhaps the easiest thing to do would be to split spacebeforecustos into spacebeforeinlinecustos and spacebeforeeolcustos. In this scenario one could then zero out the inline value without affecting the end of line value.

[henryso]

I tried that, and that does indeed help if there was a note before the custos. To reduce the "too wide" space in this case, I set spacebeforeinlinecustos to be 0. However, using those settings, if the custos immediately follows a bar (like at the end of gabc-output/bugs/fix-569.gabc), the custos touches the bar. This might be the reason (in the annals of history) that the space was there in the first place.

I tried a few things, but I can't figure out a good way to deal with this in TeX without having the gregorio program detect it and use a different macro at that point. Can you think of a way to fix this problem in TeX?

[rpspringuel]

So you're saying that a bar doesn't surround itself with the appropriate spacearound..., at least under some circumstances? That itself sounds to me like another bug. Is there some set of circumstances under which spacearound... should intentionally not be placed both before and after the corresponding bar? Maybe @eroux can remember programming that behavior?

[eroux]

Well, sorry, I cannot remember anything about this...

[henryso]

That's a good clue. The problem with touching is only on the final divisio finalis. The final divisio finalis uses the end-of-line finalis macro as opposed to the in-line finalis macro, and changing this might break some edge cases where a final divisio finalis without any custos after it appears at the end of a line. I don't have any more time to look into it this morning, but I'll most likely use a special macro to handle this case unless someone has a better suggestion.

[rpspringuel]

Perhaps the final divisio finalis macro should only be used if the absolutely last element of the score.

[henryso]

I tried that, but found that the rubber space after the normal in-line divisio finalis can make things look a bit funny at the end of a line (this is from tex-output/bugs/fix-43/fix-43.tex):

I personally am not happy with how that looks, but I am willing to accept it if others think it is OK. Otherwise, I will try to come up with some way to tighten that space with a special \GreFinalCustos macro.

[henryso]

Any comment on this? Should I try to tighten that space or is this satisfactory?

[eroux]

I think it should be tightened

[henryso]

With a final custos macro, I can get it to look like this:

... with the space tuned to be like the one after the dagger in:

Comments?

[eroux]

looks good!

[rpspringuel]

I agree.

[eschwab]

I also agree. It looks much better tightened.

[henryso]

The big question: release-4.0 or develop?

[eroux]

I don't have a very strong opinion, but I would say 4.0...