2016-01-20 13:11 GMT+02:00 Dirk Laurie <[email protected]>:
> the sequence
>
> (+4 f2d { e4 | e } )+3
>
> which is supposed to be equivalent to
>
> f2d s+4 e4 t | e t s+3
>
> (if I code that explicitly, it works) behaves as if it is equivalent to
>
> f2d s+4 e4 s | e s s+3
>
> I haven't coded straight PMX without the aid of M-Tx for almost
> 20 years, so I don't know whether this has always been like that
> or whether it crept in with a recent release.
The answer is: it has always been like that. As Don rubbed in
in a reply to this post, the perpetrator of the () and {} syntax in
PMX was a certain Dirk Laurie, venturing rashly to do some
modifications of someone else's Fortran code [1] and of course
not understanding all the subtleties.
One may allege that Dirk Laurie knew about it years ago. The
accusation is supported by what M-Tx does to that sequence
(which is perfectly legal M-Tx code too). Lo and behold, it is
translated to
(S+4 f2d { e4 | e4 } )S+3
I.e. M-Tx always puts an explicit label on an unlabelled pair
of parenthesis slurs. This fact is even documented in the
Corrections file that comes with the M-Tx source code.
But the truth is that M-Tx does so because
(+4 f2d ( e4 | e4 ) )+3
would also be quite legal. So I don't think Dirk Laurie,
who claims not to have coded straight PMX without the aid
of M-Tx for a very long time, knew in 2004 that PMX allowed
nested unlabelled slurs when one is type s and the other
type t. His PMX modifications were just enough to support
the use that M-Tx would make of them. Selfish if not
solipsistic, I grant, but that's Dirk for you.
[1] He was lured into this indiscretion because precisely
that happened to be the main daily grind of his first job.
-------------------------------
[email protected] mailing list
If you want to unsubscribe or look at the archives, go to
http://tug.org/mailman/listinfo/tex-music
