Re: User comments on R shorthand

[David Kastrup]

Kieren MacMillan <address@hidden> writes:

> Hi David,
>
>> See page 3, at the bottom, for two, three, and 
>> four measure rests. (I'm obliged to cite and give
>> credit for the scans.} How can these be implemented
>> without R2, R3, and R4?
>
> Assuming that staff is in 4/4 time (there is no time signature
> currently), then R1*2, R1*3, and R1*4 should work perfectly — do they
> not work for you?
>
>> There should be no problem detecting the measure lengths.
>
> That does seem intuitively true…
> However, David K. (who knows way more about this than any of us) says
> it's false — I trust his judgement/opinion, until seeing concrete
> evidence to the contrary.

Not reliably, and I don't see a reliable way to complain about the
unreliable cases.  LilyPond needs to be able to calculate the length of
music in terms of whole notes in a number of "dry run" situations.

If we want to stipulate that bars and whole notes are interchangeable,
how many bars are there in

music = { \time 4/4 a1 a1 a1 a1 } ?

The answer appears intuitively simple.  Now look at the output of

music = { \time 4/4 a1 a1 a1 a1 }

\new Staff << \music \\ { f1 \time 12/4 f1 f1 f1 } >>

which is a perfectly valid though slightly capricious LilyPond file.
\music takes two bars.

-- 
David Kastrup