[
Lists Home |
Date Index |
Thread Index
]
On Thu, Mar 07, 2002 at 04:15:26PM -0800, Tim Bray wrote:
> At 04:11 PM 05/03/02 -0500, Daniel Veillard wrote:
> > For me there is no doubt that the document is the set. Well formedness
> >is defined for the set, and a well formedness error detected when parsing
> >an external entity affects the whole document.
> >
> > Anyway, even if the REC may be ambiguous, from a programmer viewpoint the
> >document instance will likely to be based on those extra sets, XPath
> >for example requiring them.
>
> More good arguments for getting rid of entities. -Tim
If 1.0 had come without entities, I for sure know that libxml code
would be at least twice smaller, easier to read and would have probably
taken 4 time less time before reaching the current stability state.
Changing it now is a two edged sword:
- new incompatible XML version could be made simpler, easier to
understand, etc ...
- a bunch of application will never be able to upgrade and people
shipping XML parser will anyway have to handle the complexity
of 1.0
It's like HTML, now that the market has expect an HTML parser to digest
and render anything which could contain a <p> or <html> tag, it's very hard
to pull the plug and say "sorry you have to change", so far the cleanup
versions attempted like XHTML are not a success, I don't see why/how this
could be made to work with XML (I would love this but I don't see why this
would magically work this time, especially since contrary to HTML violations
this is correct behaviour for the previous version).
So while I understand the theorical practical use of such a major revision
I still don't see in practice how it can be useful, seems very hard to release
a parser that would understand only that "stripped down" version. I didn't
see any SML specific application flagged as such in the last year.
So would such a large refactoring of any practical use? (independantly
of the possibility of cleaning up some of the grey area where some of the
core specification interract in bizarre ways.)
Daniel
--
Daniel Veillard | Red Hat Network https://rhn.redhat.com/
veillard@redhat.com | libxml Gnome XML XSLT toolkit http://xmlsoft.org/
http://veillard.com/ | Rpmfind RPM search engine http://rpmfind.net/
|
