http://codereview.appspot.com/4536068/

Passes make and there is a difference in the reg test but I am not sure if this is incorrect. There have already been a lot of reviewers, but I added this because it evidently isn't a trivial change. So at least we have a track for the reg test difference.

James

:o\ ... Sorry to bother you again with regtests, James, but you should get what is attached. You probably forgot to compile lilypond between the baseline and the check. As you can see, there should be a breve in every 3/2 measure.

I can attach PNGs of regtests in the future, but it's much better if someone else than the patcher does it.

Thank you, Bertrand

Bertrand,

Not a problem for me (takes me less time to build from scratch and reg check than make a decent pot of coffee).

You were indeed correct - I even went back to deliberately forget the second make after the patch is applied - and my attached reg test is here, now. Looking the same as yours.

I think I know how I made this mistake; multiple 'terminals' open with different histories stored in the shell.

Sorry for this unnecessary work for you to tell me - but I guess this is why it is good to post any reg differences from someone who isn't the author of the patch :) if if the checker needs checking themselves sometimes.

So this looks ok.

Pushed as 74ef44d10b7bb20e68117c787e78371d0d19997b