Re: [GLISS] differentiating pre/post/neutral commands

[David Kastrup]

Xavier Scheuer <address@hidden> writes:

> On 11 September 2012 14:04, David Kastrup <address@hidden> wrote:
>>
>> Ok, and now for something completely different.  I think there has been
>> one proposal to bring \[ \] in line with the post-event nature of [ ]
>> and ( ), but the one thing I have been thinking about recently is
>> whether we should not actually be going the other way round.
>>
>> (snip)
>>
>> One argument might be that
>> c( c)
>> might look ugly, but less ugly than
>> (c )c
>> looks.  Of course, neither is symmetric.  But if our logic is strained
>> enough that people want to invent new constructs for preserving the
>> current _order_ of writing while differentiating the logic, maybe it
>> would make more sense to rethink the order.
>
> Do I understand correctly that you suggest to change completely the
> LilyPond language so that every command that is currently post-note
> should be before the note?

No.  Just those commands that are not intrinsically attached to a note
within a voice, like dynamics and phrasings.  Basically those things
that you'd occasionally attach to <> or s1*0 for lack of something more
suitable.

> Ugh, maybe that is more logical (for you) but the postfix syntax has
> been used for many many years.

No question about that.  I assume you have seen the [GLISS] tag in the
subject line.  The question is whether there are some constructs for
which it has been abused for many many years.  At the stream event
level, those events are _never_ part of a note event (as opposed to
articulations).  That makes the mandatory attachment at the input syntax
level somewhat suspicuous.

> That sounds to me like people wanting
> to change the electric current convention so that it is the same
> direction as the electron flow.
> http://xkcd.com/567/

Except that \p\< c \! and c\p\< <>\! are not dual in nature.  The second
requires an additional artificial element.

And we don't write c\time 4/8 either.

-- 
David Kastrup