Lists Home |
Date Index |
> -----Original Message-----
> From: Simon St.Laurent [mailto:email@example.com]
> Sent: Sunday, October 19, 2003 10:53
> To: XML Dev
> Subject: [xml-dev] more fixed lists (was UTF-8+names)
> firstname.lastname@example.org (Tim Bray) writes:
> >Check out http://tbray.org/tag/utf-8+names.html
> I have (yet) another concern about this document.
> Sections 6 and 7 appear effectively to create
> yet-another-fixed-list for processors to incorporate into
> their code, with no provision for future change.
I don't see how the lists can be changed in the future, since we are talking
about an encoding. Changing the lists would mean that the encoding becomes
a different encoding, and must have a different name (UTF-8+names-1,
UTF-8+names-2, etc.). The last thing we want is interoperability problems
between different versions of the *same* encoding that decode the same byte
sequence in different ways. So any list that is initially specified must be
carved in stone.
Or do we want a parameterized UTF-8+names encoding, depending on a named set
of lists? Something like "UTF-8+names(myLists)"? This would probably solve
the original use case better than a fixed set of lists. But aren't we
getting too far along this path?? I think we are.
It seems to me that we are really addressing the problem at the wrong layer.
The more I think about it, the less I like the proposal. I can understand
the use case that inspired this proposal, but I increasingly believe that
the problem should be solved at the XML level, not at the encoding level.
> We've seen how well this worked for XML 1.0 - until the
> appearance of XML 1.1, which provoked all kinds of discussion
> about the impact of change.
> What provision for changing these lists can be made?
> More importantly, what precedent does this set? If I want to
> create my own "UTF-8+names-in-Russian" and
> "UTF-16+names-in-Chinese", is there anything stopping me?
> It's not like screams "non-breaking space" to people
> who don't speak English, and localization of such things
> seems like more than a convenience as XML spreads.
> The xml-dev list is sponsored by XML.org
> <http://www.xml.org>, an initiative of OASIS
The list archives are at http://lists.xml.org/archives/xml-dev/
To subscribe or unsubscribe from this list use the subscription