Re: [syndication] Re: generic minimal set desired in a data syndication message?

[Benet Devereux <benet@cs.toronto.edu>]

> Since Ian isn't here, I'll try and respond (of course Ian is free to correct
> me when he gets back).
> 
> Ian is tying to start from a clean slate, in regards to syndication. For
> him, RSS has too much legacy stuff as well as trying to do more than just
> syndication. Ian is looking for...well, as the subject line says: the
> generic minimal set desired in a data syndication message. Perhaps the
> result of this work will be an extension to RSS (as well as perhaps another
> spec) but in the mean time, it's important to begin work on finding exactly
> what people mean by syndication, and what they need to do it.
> 
> That's what Ian's trying to do right now, he's not trying to create a spec
> or a module or anything. I hope other people can help out.

Yes! That is, more or less, what Ian (and I) are trying to do. In software
engineering speak, for which I apologize in advance, the goal is to
express a set of requirements for a syndication system, which will drive
out either:

      a) the use of pre-existing modules and protocols,
      b) the creation of new ditto and ditto, or
      c) some symbiosis of both.

And at this stage, we want to avoid implementation bias. While tossing
around some requirement, it is easy to look at it and say "Oh, you can do
that with RSS1!".. and that may indeed turn out to be the most efficient
way to handle the requirement, but we want to avoid committing to such a
thing until the requirements and the trade-offs between them have been
somewhat more fully elucidated.

-
Benet Devereux
Dept. of Computer Science
University of Toronto
benet@cs.toronto.edu
-