What happened to the interaction, reading text, reading comments.
And how many feeds do you really need? And how hard is it to just click onto their page and see whatever there is to see via html?

Feeds work for video only page, like rocketboom.

AND.. now people age watching vlogs on those little videopod things.. All shrunk down to 3″x3″.
Seeing video on the web isnt a new thing. I dont understand the hype.

Dave may be writing tools, but he’s writing tools that will work with Dave Winer systems (RSS 2.0, OPML, XML-RPC, Frontier/Manilla) rather than anyone else’s. That’s another benefit of specs – you’re not tied to proprietary systems. I’m not actually concerned one way or the other whether Dave develops cool outlining/collaboration tools. What I am bothered about is having to support badly designed formats in tools I write for the Web. Even Dave himself had to write converters for his outliner to handle lists found in the wild, even though they were spec-valid. Work that could have been avoided had there been decent specs. But it’s to be expected – bending a format designed to support simple hierarchical text editor (a 1980’s tool design, no less) out into every possible data application is just plain silly.

HTTP and HTML are largely defined by their reference implementations, which at the time of conception were developed by NCSA and later transitioned to Apache and Netscape. However, now if we look at standards for web development, it would be nice if everyone complied with the specifications, but the reality is not even Safari and Firefox, which are standards compliant browsers, render everything identically, even though the specifications are quite explicit. Internet Explorer is another matter, and in the end development requires workaround for different clients.

RSS and OPML have been able to avoid a lot of these problems due to their simplicity. Also, when there’s a doubt in the specification, which for both is far from complete, there is a reference implementation to turn to to see how you should implement something.

Often times people get together to define a standard that will solve a problem, take for example the Yahoo Media RSS group. There is a problem with enclosures. The specification is unclear as to multiple enclosures, and having to have a seperate for every enclosure isn’t ideal. However, after a lot of work a specificaction is out there, except there’s no reference implementation and no major adopters, so it isn’t gaining traction. There has to be an application to go along with the specification or it’s largely meaningless.

Nobody’s writing software doing the same thing with XHTML formatted lists. The reason OPML is succeeding and will continue to succeed is because Dave Winer is developing software people are using to write outlines and collaboratively share information. In the end, that’s what matters.

Atom format was 2 years in the making, RSS 2.0 took about 4, OPML has taked 5 so far. There are (informative) reference implementations for Atom – most notably the validator. But if you follow a good spec you don’t need a validator, it should just work. That’s the whole point. You agree on conventions, and your applications can work together.

“Why is Microsoft Word the default format for exchanging documents and not OASIS?” – because one is a proprietary word processor application and the other a standards organisation?

What about HTML, HTTP and XML? Proper specs or doesn’t it matter?