Re: [musicxml] Suppress the 'non-controlling' attribute in measures

> Still I wonder: Is the MusicXML-file OK? Finale shows it wrong and 
> Musescore will not load it at all. I think that
> I have seen some example with mixed 2/4 and 3/4 that worked with Finale 
> and Musescore.

This is a great illustration of the consequences of not having a real
specification for MusicXML.  Over and over, we see questions "Is this valid?
Finale does X, <other application> does Y, which is right (if either)?"  A
real specification would tell us whether or not the input is valid, and
whether Finale's or <other application>'s behavior was conformant or a bug.

I know that we are not going to produce a real specification as part of the
current MusicXML cleanup activity, but given the history of MusicXML
evolution to date, I feel it is important to be reminded that the desire for
/ lack of a real specification has real consequences and is not just
pedantic.

> Here Finale adjusts the measures to be same length and drops the last two
> measures of the 2/4-part.

I have seen it argued that MusicXML importers should be free to disregard
any and all elements of the input in favor of their own ideas of layout and
notation, with no approval by (or even indication to) the user.  I have
argued, and will continue to argue, vociferously to the contrary, since this
stance is contrary to all industry practice in my experience.

						L Peter Deutsch

Received on Friday, 15 January 2016 19:39:32 UTC