Yet Another RSS History: "
[You don’t expect me to work out the CSS right after making it semantic, do you?]
Shift to another universe. It’s sometime in the late 1990’s. Ramanathan Guha, Tim Bray, Dave Winer, Tantek Çelik, Dan Libby and Dan Connolly are sharing a jacuzzi*. As they sip Marghueritas, their conversation goes like this:
- DanL
So, we’ve got this idea for publishing content that’s a bit like CDF, but we’ve made the system more of a service than just a desktop thing.
- Guha
Sounds cool. Might be a good fit with this RDF thing I’ve been working on.
- Dave
Hmm, Dan’s stuff does sound cool, but with all due respect dude, RDF does seem a bit complicated. I really don’t think the folks out in userland would get it. And they majored in graphs.
- Tim
Maybe we could make it a bit more straightforward, you know, like put pointy brackets around it?
- Dave
Straightforward’s good. Better still, simple. They like simple.
- Tantek
But what about the rest of the Web, you know, like HTML?
- DanL
Hmm, but how do we do the timestamping kind of thing, and wrap it up in a ‘microposty’ way, the things that makes this distribution mode work?
- Guha
Yeah, metadata is cool. Keep the metadata.
- Tim
Not cheap though. The Web must be cheap. Did Andreesen show you his pictures..?
- Dave
…’Microposty’? you mean like my newsletter thing, but on the Web?
- DanL
Yep, like Cool Diary Entry of the Day
- Tim
But do we really need 1000 pages of spec for that?
- Tantek
…Incidentally, did you see my Box Model Hack?
- Guha
Yup.
- DanL
Yup.
- Tim
Yup.
- Dave
Yup. I explained that on DaveNet last year.
- MarcC
Hey! I’ve got it: ‘MyDigitalCocktail’..?
- DanC
Hang on, that gives me an idea…
There was a tangible outcome to this conversation: a document format which supports content and unambiguous, explicit, data and metadata, timestamping and much, much more. It’s viewable in a regular browser. Can be syndicated; can be aggregated. Unlike forgetful RSS, archives are almost always retrievable using regular HTTP methods. In this universe there was no RSS. No syndication wars. No talking-at-cross-purposes conflict between docheads and dataheads, syntax fans and model fans. No-one had to publish simple data in Byzantine RDF/XML. No-one had to deal with doubly-escaped content and silent data loss. There was no need for any new format for business cards, calendars, blogs, link lists, reviews, pet profiles. XHTML with CSS was more than enough. DanL got the MyNetscape he wanted. Tim got the simple, tight format he wanted. Guha got the AI. Tantek got to do presentations in a cool black raincoat. DanC finally got his schedule on his Palm Pilot. Dave got the credit. MarcC got the parasols and a grass skirt none of the others would admit to having brought.
Shift back to this universe. Check out hAtom. It’s not finished yet, but David’s been methodically working through the (utterly sound) microformats process. Looks good to me.
* apologies for the imagery, but how else do think Silicon Valley might seem to someone raised in the cowpat-coated hills of Derbyshire?
PS. Apologies to everyone mentioned. And before you suggest it, blogging *is* therapy.
"
(Via Raw.)