Re: [relax-ng] annotations

[John Cowan <jcowan@reutershealth.com>]

Michael Fitzgerald scripsit:
> 
> I think annotations are a necessity and that they should be distinctive. I
> have tried to come up with alternatives to our current compact syntax, but I
> just succeeded in convincing myself that what we have is fine.

I agree that the current syntax is as good as we're going to get.  I
still think that annotations corrupt the compact syntax, because they
have to do with how the syntax is translated, not what it means.
Without annotations, a translator is free to do whatever it wants in
translating the syntax (e.g. it can rearrange definitions, or preprocess
include and external, or whatever), _salva veritate_ (that is, as long as
the same class of documents is recognized).  With annotations, translation
is far more restrictive.

-- 
Not to perambulate              || John Cowan <jcowan@reutershealth.com>
    the corridors               || http://www.reutershealth.com
during the hours of repose      || http://www.ccil.org/~cowan
    in the boots of ascension.  \\ Sign in Austrian ski-resort hotel