RE: Letter from Planet Web on identifiers

["Roger B." <roger@xxxxxxxxxxxxxxxxxx>]

> cookies do muddy the waters but have the virtue that
> middleware can safely ignore them...

Tim, 

Not if the middleware is trying to accurately identify something. To the
extent that a "good enough" approximation does the job, this is fine... but
the bottom line is that resources change --sometimes in significant ways--
depending on the value of a cookie.

> and the Web doesn't do sessions, it
> uses stateless protocols].

The Web does sessions every day. It does them with UUIDs in query strings,
cookies, embedded form fields, and so on. You don't have to like it to
recognize it as a fact. :)

> And, by the way, ever since about fifteen minutes after the dawn of the
> Web, voices have been raised in horror...

I'd like to clarify that mine is not a voice raised in horror. I'm simply
saying that, while it would be nice if one URI == one resource, that's not
the way real people use the Web. And it will never be. So as long as we've
got a use-case for definitive identification of a piece of microcontent, we
need an atom:id.

> So Norm, you're changing the styles (which BTW can suppress whole
> <div>s) and the links (a defining part of web content) and the context
> and you're really sure I should still think these are the same thing?
> Think you could leave that choice up to me?

Whether we like it or not --and I don't in many cases-- syndication is
becoming a secondary layer atop the Web. We're already seeing many RSS
converts who resent having to leave their aggregators to visit an actual
page. They don't care about the styles, links, etc.... if the feed items
match, then they are effectively identical from the user's perspective, and
the user doesn't want dupes.

--
Roger Benningfield