Re: Human Friendly Trix

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Jeremy Carroll wrote:
| Yes the TR for xml-stylesheet has a rationale section which reads
| somewhat like an apology. I find it unconvincing. PIs might have
| weaknesses but the strength of not requiring any change to anything
| else. The problem with additional functionality as xml attributes is
| that it is opt in, the existing schema and DTD need to be modified to
| permit the attribute.
| I guess in the TriX case that is probably acceptable.

Yes, that's what I meant.

| It is hard to permit multiple transforms, which you can do with
| stylesheet PI.

Can you explain what you mean by 'multiple transforms' -- alternative
ones or sequential ones? What's the use case for them in TriX?

Of course if you have a TriX mechanism for specifying TriX transforms,
that doesn't prevent you in any way from also including an
<?xml-stylesheet?> for a transform to HTML or XSL-FO or whatever. Having
alternative transforms of an XML file into the TriX plain triples format
doesn't seem very useful, a single deterministic mapping seems the only
reasonable thing to do there :-)

|> OTOH, introducing your own "media descriptor" for TriX should also be
|> possible without causing havoc, even though the specs don't allow it.
|
| I read HTML 4.0 as permitting a profile to include new media
| descriptors, without providing any mechanism for so doing.

Ok, I didn't know about profiles, was just reading the section that
talks about media descriptors (it didn't link to profiles).

| Admittedly
| there is no mechanism for using a profile in xml-stylesheet

Gn, then that doesn't help :)

- - Benja
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org

iD8DBQFANhSVUvR5J6wSKPMRAnZfAJ9x6GI2idLr27Bp0BR3tCGgLZYgeQCePnUe
wpZvYUpPONnbsYWGYBfZPUE=
=N6Ow
-----END PGP SIGNATURE-----

Received on Friday, 20 February 2004 09:07:26 UTC