OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.


Help: OASIS Mailing Lists Help | MarkMail Help

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: using namespaces to version


> Why can't an application "guess" that "http://foo.com/myspec/0/1" is
> related to "http://foo.com/myspec/0/2".  Of course I quote "guess" because
> since computers are imbeciles, there would actually have to be some
> method/convention in place for this to be useful.  I see Eric's idea as
> one such convention.

We could also use query parameters to differentiate versions. It's a lot
more obvious to standard URI processing tools that
http://foo.com/myspec?major=0&minor=1 and
http://foo.com/myspec?major=0&minor=2 because they have the same scheme,
authority, and path parts. They only differ in their queries.

We could use the URI without the query part (http://foo.com/myspec) to refer
to the version independant namespace or append some parameters to refer to
version dependant namespaces.

The advantages this has over Eric's proposal is that it's very easy to tell
where the version independant and version dependant parts start and
end--just look for the question mark. It also gives us a little more
flexibility--we don't need to limit ourselves to merely positional, version
indicators. How about a lang parameter? http://foo.com/myspec?lang=en versus
http://foo.com/myspec?lang=jp anyone? (Bonus points for anyone who can dream
up why you'd actually want to do this.)

How about a parameter that indicates the current W3C status: Would
http://www.w3.org/XFoo?status=WD versus http://www.w3.org/XFoo?status=REC be
useful? (Only if the syntax didn't change too drastically, I imagine.)

Is it possible to use URL rewriting to redirect requests made to the same
path but with different parameters? If not, that could be a disadvantage. It
would take a servlet of some sort to do the redirecting. But I bet it
wouldn't be too long before somebody implemented mod_namespace if we needed

The biggest disadvantage to both approaches, though, is that existing
namespace-aware processors will treat them as different namespaces
regardless of how similar they look to us silly humans. Given that, I can't
see how we could make such suggestions this late in the game.