Simon Wascher <[EMAIL PROTECTED]> writes:

> I looked into my files, and found that the abc files I
> transcribed are a pile of about 3 MB (plain ASCII). Assuming that other
> peoples who are listed as large collections at the abc homepage also
> have big collections besides what is in the net right now we are talking
> of at least 60 MB of content (another approximation is to multiply the
> 14000 titles of the www abc index with an single tune size of about 20KB
> makes 280.000 KB ).

I would say that much of this material is such that it can use the
current ABC definition till kingdom comes -- `Celtic' folk music and
other one-melody-line-plus-chords stuff. At least this applies to the
several megabytes of ABC-notated music on my machine. It seems that
therefore a great portion of that corpus will never have to be
translated into another representation.

If people do come up with a new notation that is better for multivoice
music, complicated classical scores etc. then by all means use that for
those kinds of music. I for one am quite happy with ABC the way it is
because it fits my requirements pretty much perfectly (with some
tweaking which could be made unnecessary without throwing all of ABC out
the window). Any completely-new notation had better be as simple as ABC
for my uses before I personally am going to jump ship.

Mind you, I'm all in favour of updating the ABC standard but not if new
functionality is invented wholesale. Let's try and bring the various
implementations together and build from there.

> So if an entirely new syntax appears how will this syntax interprete
> this pile ? what are your solutions?
> will you write an all plattform automatic conversion tool and is it sure
> that no part of thecontent gets lost in this process? 

I'm sure something could be worked out. A conversion tool doesn't
necessarily have to work on all platforms -- there could be a Web-based
conversion service for those who cannot or will not run, e.g., Perl
locally to convert their files.

Anyway, if a backwards-incompatible version of the ABC language (rather
than something entirely new and separate) is agreed upon then there
should be a way of tagging new-style files so that ABC processing
software can still work with both flavours without having to guess. We
could do it XML-style so that the first line of a file (or tune?) reads

  %%ABC version="2.0"

(and this would also be the natural place to put `encoding="utf-8"' and
such-like if desired).

Anselm

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

Reply via email to