[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [syndication] Re: SuperOpenRatings
At 08:00 PM 2/5/01, you wrote:
The idea is to deal with people who have CMSes that change the link as the
document is moved.
(Okay, this might be slightly off topic...)
Please forgive my naivete, but if I publish a syndication file pointing to
new articles on my web site, there are CMSes out there which then change
the URL I provided that points to those items? Sounds like they are broken
to me.
Combine this with misspelled / reused / missing titles
Then use the unique link.
Aren't we talking about third parties (customers) rating content generated
by other people?
and the fact that the link is option in RSS 0.92 and you've got a problem.
I guess this is what I get for ignoring 0.92 until it is finalized...
taking a look.... ugh... I have to say I think this is overloading RSS
beyond what is appropriate.
I thought RSS was for *original* content authors to be able to provide
notifications about what they are publishing?
If we want a description language for general web pages (collections of
information and links pointing to/from many sources) then I suggest we need
another schema.
To me RSS 0.91 already was a bit too verbose, and these large-scale changes
in 0.92 just really complicate matters (as well as being semantically
muddled (are we calling them "link"'s or "url"'s?)).
(I'm not sure we need a new element for "enclosures", why can't they just
be another item? (the description would indicate their association...) And
I think the source really should point to the web site/page where the item
is offered by the original author, the idea of recognition being to drive
traffic to the site, most people will be taken for quite a loop by pulling
up XML source...)
Well it couldn't be centralized, of course. I was thinking more along the
lines of a permanent, generated URI for the site, or a unique ID on a
per-feed basis.
Then where does the unique ID for the feed come from if it is not centralized?
-Robert