intertwingly

It’s just data

Mr Safe


Q: It has been 31 months since I last talked to Tim Bray, Jon Udell, Joshua Allen, Chad Dickerson, Dave Winer and John Gotze.  Can I have an update?  Are Funky extensions still an issue?

A: Only if you want to name an author without revealing their email address to spammers, or to include the full content of your data in addition to a summary, or if you want to include the date the item was published on platforms like JSP which automatically try to adjust for localization based on HTTP headers — but don’t worry, the latter can be solved with enough work, and every widely used consumer of RSS ignores the “funkiness” issue.

Q: Hmmm.  And what’s this I hear about silent data loss?

A: Well, the effective content model for RSS has been described as “Here’s something that might be HTML. Or maybe not. I can’t tell you, and you can’t guess.” so things that look like HTML are routinely misinterpreted.  The author of FeedDemon indicates that he gets multiple bug reports per week on this issue alone.

Q: Couldn’t that simply be solved by identifying which elements are HTML and which elements are plain text?

A: Yes

Q: How long could that take?

A: Well, lets put it this way.  Remember that Echo effort Jon told you about?  We started based on experiences with RSS, but otherwise with a clean slate.  We had interminable discussions over the name of the effort.  Then we had interminable discussions over which standards body to pick.  Then we had interminable discussions over dates.  Eventually we solved all the issues above as well as other issues like relative URIs.  And we had to wait while the slow standards process ground away.  And finally, we got an official RFC issued by an internationally recognized standards body: the IETF. The same organization that issued the HTTP and SMTP standards — these are the protocols that power the web and email respectively.

Q: I don’t want to hear about Atom.  Everybody is talking about RSS. Surely in that time there was an update to RSS?

A: There were several.  There was an RSS 1.1.  And an individual I never heard of who is working on an RSS 3.0.  In RSS 2.0, an example was added to indicate how HTML could be used in one of the elements, but plain text remained valid.  Oh, and ownership of the 2.0 spec transferred to Harvard.

Q: Harvard?  They are a respectable institution.  That should be pretty safe, right?

A: Well an RSS Advisory Board was set up, and they did all the work.  Harvard reportedly is “delighted to know that many members of the RSS community continue to work on relevant issues to move the industry along in various ways, including related to the spec itself, Harvard has no involvement with any of these efforts.”

Q: No involvement?  What a pity.  OK, tell me about this RSS Advisory Board then.  How can I join?

A: Beats me.  Several are recruited never to be heard from again, and are quietly purged from the list.  Others replace them without any documented nomination or vote.  And reportedly the original founder of this group claims that the RSS Advisory Board no longer exists. Something that came to the evident surprise to the current chair.

Q: Ouch.  Does this claim have any weight?

A: Perhaps.  The role of the Advisory Board has always been, well, advisory.  Both the owner of the spec, and the sole author of the spec, are clearly stated — in the spec itself.  Once the author of the spec left the board, the board’s role in recommending changes was rendered somewhat irrelevant.  When he was asked about reviving the board, his response — according to the very words that the chair posted on his website — was that you could “carry on the business in any way you want” with conditions that included that “the spec, to be left as-is”.

Q: OK, so they have shut down then?

A: No.  They are still working towards producing a draft-1.

Q: Can they do that?

A: I’m not a lawyer, but the spec is under a Creative Commons by-attribution share-alike license, so apparently literally anybody can create a derivative work as long as they attribute Harvard and license their work under an identical license.

Q: Anybody?  So, we have the possibility of even more forks?

A: I’m afraid so.

Q: But Microsoft is planning to include RSS support in LongHorn, I mean Vista, in 2006 or 2007 or 2008 or something.  Microsoft is no dummy. Why did they pick RSS over Atom?

A: You have to realize that Microsoft uses the term RSS like many use the term Kleenex.  I’m not sure why they insist on confusing the market this way.

Q: So you are saying that Microsoft supports Atom too?

A: Yes, their Feed API works just fine with Atom feeds.  They were also amongst the very first to produce feeds that conform to the Atom 1.0 specification.

Q: But surely when Joshua Allen told me that it was safe to go with RSS, he was referring to RSS 2.0, not Atom which didn’t exist yet?

A: Actually, I think Joshua secretly prefers RSS 1.0.

Q: Oh, this is so confusing.  But being so new, Atom can’t be very widely supported yet, can it?

A: At this point, pretty much everybody supports Atom.  But I have to be honest with you, we are still working with the vendors to address conformance issues based on results of running their tools against the set of Atom compliance tests that have been defined to date.

Q: Ah ha!  Since RSS 2.0 has been around longer, tools undoubtedly do better job on the RSS 2.0 compliance tests, Right?

A: There are no RSS 2.0 compliance tests.

Q: Where do you think we will be two to three years from now?  Will there be new versions of Atom?  Of RSS?

A: As we discussed, work is actively continuing on a new draft of RSS 2.0.  Ultimately, we’ll either get a real spec for RSS 2.0 or see more people fleeing to Atom and RSS 1.0.  Unlike RSS 2.0, the core of Atom is already clearly specified.  Instead, efforts are focused on extensions.

Q: Meanwhile, it finally is safe to put out feeds... based on Atom 1.0?

A: Yes.