Of RDF, RSS, and clumsy backpedalling

by Uche Ogbuji

Here is a hilarious exchange between Dave Winer, John Cowan and Danny Ayers. I shall minimize my editorial comment (except in my choice of excerpts from messages). Draw your own conclusions.

In an XML-DEV message, Dave Winer says:

Reading your last paragraph, it would have been good if the RDF advocates
had recognized the work that had gone into RSS before they tried to hijack
it. To this day they don't recognize it. Look at the design of RSS 1.0 and
how disrespectfully it treats 0.91, which to this day dwarves its installed

If RDF wants to be considered, it should make a thoughtful proposal -- not
be the bull in a china shop that it has been.

John Cowan responds:

One could equally well say that RSS 0.91 hijacks the RDF-compliant RSS 0.9.
A plague o' both your houses. My company supports both.

To which Dave Winer rejoins:

Yes, I've heard a lot of people say that, but it's not true.

0.90 was not in any way RDF-compliant.

At this point Danny Ayers jumps in:


see the cached version of Netscape's RSS) 0.9 spec:

If you still have any doubts, try copying the sample into
the W3C RDF validator:


Lovely graph!
Lovely triples!
Valid RDF!

A certain amount of FUD is to be expected from vendors that have staked
their reputation (and probably their dollars) on a different horse, but this
is remarkable. On xml-dev. I'm speechless.

To which Dave Winer makes response as follows:

What I meant of course is that RSS 0.90 was in no way a foundation for all
the dreams people have for RDF. It's basically an XML format, and not a very
widely supported one. Don't be confused. Dave

Am I the only one who did a double-take while ROTFL?


2002-11-17 08:00:51
Joining forces
Just imagine what RSS could achieve if they quit these "discussions" that are happening all over the place (weblogs, mailing-lists etc.) and moved on hand-in-hand.


2002-11-17 11:18:09
Joining forces
That *would* make sense and has been suggested numerous time, however neither side are willing to compromise.

RSS 1.0 steadfastly insists on RDF's full presence in the format and are unwilling to modify to minimize the "RDF tax" (RDF syntax and structures) in addition to breaking backward compatibility with a format (0.90/1.0) that is in the vast minority. Userland developed RSS 2.0, which descends from the RSS 0.91 format, like an angry bull in a china shop. Its focus is on “simplicity” and backward compatibility with the larger install base of 0.91 feeds and will have no part of RDF.

Both sides are neither right nor wrong really. However I am more prone to say both sides are generally wrong and because this nonsense continues at the detriment of us all.