Re: Why XML is a bad idea[longish] (was Re: [abcusers] draft for V:)

[John Chambers]

Laura writes:
| I think the problem is that parsing lilypond files is more complicated
| because they potentially have more structure than ABC files, and can
| have includes, etc.  I don't know how difficult what you do with the
| ABC would be in practice.

I think you're right. Includes are especially troublesome. The reason
is  that  the directory structure that a program sees from the web is
not necessarily the same as what a local program  sees  in  the  file
system.   If  the  include  is  of a file in another directory, a web
program can't always correctly calculate  what  the  URL  should  be.
Probably  the only includes you could implement would be for files in
the same directory.

Of course, if lilypond had an include  that  used  URLs  rather  than
local file names, this wouldn't be a problem. Such includes are quite
easy to implement.

Another big problem I can see:  If my tune index includes, .lp files,
people  will be expecting the tune finder to extract single tunes and
return them in PS or GIF or MIDI or whatever formats. The obvious way
to do this is an lp2abc script ...


To subscribe/unsubscribe, point your browser to: http://www.tullochgorm.com/lists.html