Winston took a slightly different tack - let users choose for
themselves and avoid taking a religious position. I'm not sure I buy
this. I mean, I get the sentiment - giving users control over how
things get done is never a bad thing, but I'm left wondering if, in
this case, that would be akin to letting the user pick which language
we write the application in. Isn't it enough to allow them to specify
whether or not they would like their particular blog to offer a
syndication function to their readers? RSS might be a brand, but I'm
not sure that publisher-users have the same loyalty to RSS 2.0 as they
would to Tide?Dave Winer, RSS 2.0 chieftain, chimed in with a feature request indicating that he'd like to see Blogware support the "...guid, comments, pubDate, and if possible, category" in our RSS 2.0 feeds. Done, done, done and done. We've been that way since the beginning, hopefully it does make things easier for various aggregators in the way that Dave suggests.
Through the discussion it became clear that there really is no "right" answer and also that no one really seems to give a rats-ass if we continue supporting 0.91. I suppose the bigger question is whether or not we build in support for ATOM. The problem with ATOM is that "its not just about feeds anymore". To do it right, I have to rip out a whole whack of code that deals with trackbacks, remote posting and feed generation and rewrite it to suport ATOM. We could also choose to only support a subset of the spec, but I'm more inclined to wait for an answer to the question of "why?" Why should we care about ATOM?
At this point, I honestly don't think that anything is going to change. Keeping 0.91 feeds in the UI doesn't cost us anything and users aren't going apeshit with support calls because we offer them three syndication options. And as far as Atom goes, well...we'll see I suppose.
1As an aside, I'm not sure what TheRoss means when he says "if you value RDF"...I don't tend to value technology in the way that I think he's implying I might. If they are useful to me, then I value them, but I don't think that they are sacred cows in the sense of "standards must be preserved at all costs"...this aside is turning into a post of its own...to me, RDF (RSS 1.0) is simply an open agreed upon standard very similar to the standards that have gone through the IETF process whereas RSS 2.0 is a widely used pseudo-defacto "standard" that sits on a publicly documented specification - much more similar to open source software. In both cases, as an implementor, I have the option not to comply with the details that I think I know better about. It results in a non-standard implementation, or perhaps a fork, but the dynamics and risks associated with this type of behavior are well understood by the open (code/standard) communities. In any event, my point is that open standards/code only get my respect insofar as they deserve it. The fact that we didn't create our own syndication or publication specification should serve to speak to the degree that we support the current "standards".)
