John Harrison writes:
> It seems contrary to interoperability to allow a situation where a
> client requests one thing, but is given another.
You're right, I'd not spotted that: under Ray's proposal, my server
would not be at liberty to support only the one true record metadata
schema -- or, at least, I'd have to make it reject some record
metadata requests, since they might request one of the other syntaxes.
This of course makes a solid pragmatic (not just philosophical) reason
to continue to dislike the multiple-schemas approach.
Sometimes one is just the right number.
> There is just 1 schema for the explain response, would it be so bad
> to have just 1 schema for a metadata response?
_Thank_ you, excellent point.
Ray Denenberg, Library of Congress writes:
> > I still object to it philoshophically, because it's dumb.
>
> It's dumb?
See above :-)
> It does exactly what we want it to do at base level with no
> additional complexity for base-level implementation, while
> providing extensibility for additional schemas. So I would say
> "it's smart!".
"Rule of thumb: every time Microsoft use the word
``smart'', be on the lookout for something dumb."
-- John Walker.
I think this generalises beyond Redmond :-)
_/|_ ___________________________________________________________________
/o ) \/ Mike Taylor <[log in to unmask]> http://www.miketaylor.org.uk
)_v__/\ "If your religion does not change you, then you should change
your religion" -- Elbert Hubbard.
|