The scenario

["Dave Winer" <dave@userland.com>]

Some of the comments here are off because the commenter isn't considering
the scenario.

First, this is not happening in a Web browser, it's happening in an
aggregator.

We don't have the source of any HTML text around, ready to be examined for
links.

The server is Yahoo. They use lots of static pages, and a highly optimized
server. 404's are cheap.

The goal is to quickly get a list of feeds that are available from Yahoo.

Now I may not understand all I need to understand about Yahoo, and there may
be problems with putting a file at a specific location for other servers,
but as usual, the religious arguments are all we're hearing. Talk about
clogging up things and wasting time.

I plan to implement this in a few hours on an experimental basis on the
Berkman server, just to see what issues come up. Perhaps someone else would
like to try it out on their server, and report back practical, not religious
issues. I don't want a format that has deal-stoppers, but I'm also not going
to let religious arguments stop development of yet another useful idea. Had
enough of that.

Dave