Danny Ayers wrote:
> we don't have a link from the feed that can be guarenteed to point
> to the feed URI itself.
If these links were in the feed's <head/> element it would make life
easier for us at PubSub.com. We currently insert a link pointing to the
source feed into every entry that we insert into one of our synthetic feeds.
We do this for both RSS and Atom.
It should be noted, however, that there are potentially two "source"
feed links that are interesting. When an aggregator is in use, you could
have the "source" link that is reported by the feed itself and you could
have a different source URI which would be the URI at which the feed was
actually discovered. At PubSub.com, we'll probably want to report both of
these -- particularly if they differ.
The difference between these two URI's can, of course, have a number
of different causes. These include: 1) Typographical error by the feed owner
when inserting the source URI. 2) The feed might have been copied to a
location other then the location it was intended to have when first created
and when the source URI was written. 3) Feeds might be stored in locations
that have multiple equivalent names. For instance, the feed might have a URI
whose "host" field is a domain name, yet, the feed was retrieved using an IP
address as the "host" (i.e. pubsub.com vs "209.11.36.154" -- both
equivelant.) The same sort of multiple naming can result from symlinks,
multi-hosting servers, etc.
bob wyman
