Danny Ayers wrote:
I nearly missed this thread, but coincidentally I'd expressed myopinion onOPML this very morning in another mail, below. But first - Re. "Since OPML is already supported by every single aggregator on this particular planet..." - not for the purpose being specified.The amount ofwork needed to tool up for the new job would be the same if youuse OPML ora completely new XML syntax.Huh? OPML is used for *exactly* the purpose specified. Passing lists of feed urls around.Not in the form described, otherwise we wouldn't be having this discussion.
Two ways to proceed:Design-to-fit: pick an XML format, and then reject anything which does not match this criteria. This is not as crazy an approach as it sounds. But in any case, if you go this route, then certainly OPML is a candidate.
Actually figure out what we want, before we decide how to implement it. This sounds better in theory, but often leads to paralysis by analysis. If we go this way, the right next step is exactly what Dave Winer previously suggested: more use cases. The end result may end up being OPML. It may not. If we go this route, such a decision is premature.
- Sam Ruby