Re: MNX Proposal Overview - and reduction of "book-keeping and post-processing"

Hi Joe

Thanks for carefully answering all these questions.

About divisions: Having no divisions-value seems to remove a lot of 
problems. There is no problem with changes. Durations are rational 
numbers, they are exact by definition (1/3 is exact while 0.33333 is not).

Thanks also to James for his answers. It see that I have not described 
my use of timer-tick good enough. My MusicXML-timer-tick is related only 
to the notes. E.g. four quarter-notes will need 5 timer-ticks: In 
timer-tick 0 the first note starts, in timer-tick 4 the last note stops. 
And divisions-value is 1. (milliseconds are not relevant at this level)

I consider all timer-ticks because there may come <backup>- and 
<forward>-declarations, and I reserve a lot of memory because of this 
"GOTO any timer-tick". So MNX will save memory for me.

My pass 1 and pass 2 calculates new divisions. Pass 4 handles verses and 
lyrics in case the number of verses differ in the parts - The MNX-group 
will properly also find a solution to this. (Note the MusicXML proposal: 
"Lyrics - To which note does a syllable belong? #151".

My pass 2 also preprocesses repeats, segno, code, fini etc.. I look 
forward to see a solution to repeats etc. Hope that focus will be on the 
semantics, what does it mean? Not how it looks - e.g. a forward repeat 
on the first measure shall be defined. The note writer may not want this 
shown graphically. We have Coda and ToCoda, still some note writers want 
the Coda-symbol shown on ToCoda.

But there is a chance, that I only need one pass to process an MNX-file.

Kind regards
Mogens


PS: Can the declaration <measure> coexist with "480t"? Yes, but "117t"? 
Are we saying: We have no <divisions> and the value is 960?

On 2017-03-27 19:01, Joe Berkovitz wrote:
............

Received on Wednesday, 29 March 2017 22:55:45 UTC