Lists Home |
Date Index |
-----BEGIN PGP SIGNED MESSAGE-----
/ Gregory Murphy <Gregory.Murphy@eng.sun.com> was heard to say:
| Unlike SGML, which specified a syntax for public identifiers,
SGML defined a syntax for Formal Public Identifiers, but it never
required anyone to use that syntax. The loosest interpretation of
public identifiers in SGML is
 public identifier = minimum literal
 minimum literal = "minimum data" | 'minimum data'
 minimum data = minimum data character*
 minimum data character = RS | RE | SPACE | LC Letter | UC Letter | Digit | Special
So with the exception of some limitations on the characters that may
be used, SGML allows any string as a public identifier.
| XML treats them as simple strings, and
| suggests only that they may be used "to generate an alternative" URI
But that's what SGML public identifiers were for too.
| where the default is of course contained in the system
The system identifier can be used "to generate an alternative" too. :-)
| I am curious about what sorts of things developers are doing with public
| identifiers, if anything at all. Are you still using SGML-style formal
| public identifiers, out of habit, or necessity?
I'm still using them, but it's no more or less out of habit or
necessity than it was in the SGML days.
| Are there any efforts to
| formalize other semantics for the field, for example as a way of specifying
| alternative resolution strategies for the system identifier?
Does it really make sense to put those resolution strategies in the
Be seeing you,
P.S. You can even use public identifiers where designers failed to provide
an explicit mechanism (more's the pity) for them: see RFC 3151.
Norman.Walsh@Sun.COM | The art of living is more like wrestling than
XML Standards Architect | dancing.--Marcus Aurelius
Web Tech. and Standards |
Sun Microsystems, Inc. |
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: Processed by Mailcrypt 3.5.7 <http://mailcrypt.sourceforge.net/>
-----END PGP SIGNATURE-----