public virtual MemoryStream

I'm curious. I keep hearing people talk about what is wrong with RSS. What exactly is so wrong with RSS that is going to be fixed with Atom that can't be fixed by clarifying stuff in the RSS spec?

I've heard most of the rants and they sound flat to my ears. However you're an aggregator writer so you may have a different perspective from all the other "RSS sucks" ranters I've read.

Seeking common ground

Dare: one significant difference that is developing with the Atom specs is that Atom has several different document types: feeds, entries, archives, introspection (API), maybe a real "site summary" type, and possibly some others. In other words, a *set* of document types for users and tools to use -- some via an "API", others with plain old HTTP GET.

Really Simple Syndication only addresses one of those document types directly (channel) and one indirectly (an item within a channel). RDF Site Summary can address all of those as RDF, but only specifies vocabularies for channels and items, and a syntax profile only for feeds. Basing off of RDF Site Summary was suggested, but I think most believe that's a bad political move regardless and many won't go near RDF anyway.

*Could* Really Simple Syndication be modularized and extended with new document types for this purpose? Of course it could. My take on the process for doing that *today* is "write your own copy of the spec, publish it in some random place on the web, beg people to adopt it, in a couple of years maybe we'll take a look at making it the *real* RSS spec, but until then it's confusion and competition and we'll fight it tooth and nail, and for goodness sake please don't call it RSS during the uptake process."

The transfer of RSS to Berkman is an indication that "competition and confusion" might be less an issue today than it was, say, a few weeks ago, but there is still no indication of a willingness to adapt "RSS" in these ways and provide equal billing for those adaptations.

Ken,
So basically there's little wrong with RSS besides the fact that it doesn't have a strong family of specs around it.

That's what I thought.

You're welcome.

Now, where's the home page for putting this strong family of specs, and what are the rules for putting them there?

Ken, we have a directory on the RSS site at Berkman with lots of room for more specs. Yesterday we added a pointer to a spec written by Rael Dornfest and Chris Nandor, for example.

If you have some specs you think should be in there, let's get them in there. That's what the directory is for.

Done. First two links are for the ssf-dev mailing list and wiki. I think I submitted them to the wrong sub-directory though, due to the link from the home page and not seeing a link to the submit feature from other pages.

> The transfer of RSS to Berkman is an indication
> that "competition and confusion" might be less an
> issue today than it was, say, a few weeks ago,
> but there is still no indication of a willingness
> to adapt "RSS" in these ways and provide equal
> billing for those adaptations.

Here's one indication: I'd be delighted to see the extension mechanism of RSS 2.0 used to solve problems not solvable within the core.

FWIW, I believe that "funky" was meant not to deprecate any use of namespaced extensions in RSS 2.0, but rather to deprecate uses intended to supersede synonymous core functionality.

I wrote a column the other week about my experience as an app developer working in a Lisp-like language with a team who built the Lisp-like engine supporting my apps. It was incredibly productive because the boundary between the "core" and the "periphery" was flexible; things could and did move in both directions. When we got an extension mechanism into RSS 2.0 I really hoped we'd see that kind of dynamic, and I continue to hope that we will.

- Jon

There seems to be some confusion of what was actually transfered to Harvard from UserLand. Some people have seemed to reach the silly conclusion that the RSS format itself and control of it has been transfered. That, of course, cannot be the case since UserLand has disclaimed ownership of the format.

What was transfered was the text of the specification that resided on UserLand's server, and the copyright thereof, not the format itself, not the control of it, nothing more. How, exactly, is it a Big Deal (TM) that a document that describes RSS was moved and not the format itself and not the control of it? Or are UserLand or Dave Winer re-claiming control and ownership of RSS?

From the Advisory Board FAQ:

"Is RSS controlled by one person?

No. Each of the advisors must make up his or her own mind on the issues that come before the board. Each member has one vote. In order for any member's point of view to prevail he or she must convince at least one other member to vote the same way."

This insinuates that not one person -- but all three people of the board -- *control* RSS. What happened to that "nobody owns RSS" thing? Was that just a lie, a joke, or what?

Consider the possibility Dave Whiner has very personal reasons for keeping RSS as 'his baby'. It should be OS'd - and that's that.