Anybody looking at XML syndication, RSS or "weblogs"?

[Prentiss Riddle]

This isn't strictly a library topic, but as usual when I run across an
interesting puzzle, I immediately wonder what the digital librarians
make of it.

By popular demand I'm trying to think about the architecture of a
"portal" site for my university (by which terribly overloaded term I
believe people now mostly mean a site featuring extensive
personalization and/or customization[1]).  I'm convinced that the basic
navigation backbone of our site wouldn't benefit much from
customization, but there is a great deal of potential in customizing
"news" -- the highlights and hot links that have become an increasingly
important feature as various university communities have waked up to
the web's ability to grab eyeballs.

The question of collecting and distributing news items from many
sources in a highly customizable fashion has led me to looking at XML
syndication and Rich Site Summary (RSS).  RSS is a standard for
encoding "channels" of news items and announcements so they can be
shared among web sites.  The February, 2000 Web Techniques has an
excellent summary of RSS.[2]

A glance at some sites like My Netscape[3] and SlashDot[4] shows what
can be done with RSS.  Both let users create their own customized
"newspapers" by selecting among channels of interest.  Netscape is
starting a Yahoo-like hierarchy of channels available to its users[5].
A more comprehensive site like Userland[6], though, shows that RSS
syndication may have serious problems of scale: with hundreds
(thousands?) of channels publishing news in RSS format on a weekly,
daily or even hourly basis, a straightforward merge of everything in
more than a handful of channels is quickly overwhelming.

Complicating things further is the rise of "weblogs", the RSS analog of
personal web pages[7].  Filtering and adding point of view is perhaps
the main benefit of RSS, but when individuals with eclectic interests
create channels, the result can be chaos.  Perhaps there's a
granularity issue here:  maybe subscribing by channels is only the
first step, and services should allow further filtering by keyword or
by subject classification.  Alas, if RSS provides a place to plug in
classification information or controlled vocabulary keywords I don't
see it, particularly not at the item level.  Perhaps the answers lie
not in RSS but in XML-News[8], ICE[9], or some other syndication
standard.

So, to bring this back to Web4Lib: anybody out there looking at XML
syndication?  Using it?  Any insights on where this is going in
general, or how it might best be applied to the problem of collecting
and selecting news for a customized university portal?

Thanks!

-- Prentiss Riddle ("aprendiz de todo, maestro de nada") riddle at rice.edu
-- Webmaster, Rice University / http://is.rice.edu/~riddle 

[1] For definitions of personalization and customization see:
    http://www.useit.com/alertbox/981004.html

[2] http://www.webtechniques.com/archives/2000/02/eisenzopf/

[3] http://my.netscape.com/

[4] http://www.slashdot.org/

[5] http://dmoz.org/Netscape/My_Netscape_Network/

[6] http://my.userland.com/

[7] http://www.weblogs.com/about

[8] http://www.xmlnews.org/

[9] http://www.webreference.com/xml/column5/5.html