RE: [RSS2-Support] Re: Summary of issue with xmlns attribute
>--- In RSS2-Support@y..., "Dave Winer" <dave@u...> wrote:It sounds like the best bet, and I believe it's correct by the W3 specs, but
>> Interesting analysis Marcus -- but I wonder if the breakage is
>> really worth it.
>...but the spec could remain correct and most of the breakages could
>be avoided - I believe - if the spec explicity made the requirements I
>went into in the second half of that post. I don't think there _has_
>to be a complete tradeoff between maintaining correctness and avoiding
>breakages while allowing for namespaces.
>Anyone else able to confirm if this would be true?
the only way of finding out would be to test it.
Expecting parsers to ignore prefixed elements/attributes would have struck
me as a total long shot, but that was before I learnt that there isn't a DTD
The problems with 'xmlns' have demonstrated that it isn't possible to have
full backwards compatibility and add namespaces, so the current status of
the RSS 2.0 spec is broken. Unfortunately it's also frozen. RSS 4.0 anyone?
- I don't think Sam is suggesting backing off current version numbers.
I think he is suggesting two different version numbers for the
newest Userland format. Use RSS 0.94 for the non-namespace format
and RSS 2.0 for the version with the namespace option.
--- In RSS2-Support@y..., "Dave Winer" <dave@u...> wrote:
> I thought of that of course, but it doesn't work -- because of the
existence of 1.0.
> One possible back-off is for the RDF folk to change the name of
their spec to something other than RSS 1.0. I don't see that
happening anytime soon, it's been debated ad nauseum, I can't devote
any more cycles to that option.
> ----- Original Message -----
> From: Sam Ruby
> That sounds like a very easy problem to solve. Resurrect the
> All RSS 0.91 feeds continue to work.
> All RSS 0.92 feeds continue to work.
> All RSS 0.93 feeds continue to work.
> All RSS 0.94 feeds will work.
> This also gives RSS content producers an unambiguous way to
> namespaces contained herein" via the use of an 0.9x version
> for RSS content consumers to get fair warning that the
> troublesome to some consumers) namespaces are present when they
> encounter a 2.0 version number.
> I love it when an apparently intractable problem has a simple
> - Sam Ruby