Oh My! Netscape
by Rael Dornfest
sporting hundreds of channels carrying content from individual
providers, has shed its free content and become YAM*, Yet Another My.*. In the process, they also broke RSS 0.91.
A little background
When introduced in 1999, the "My" concept itself wasn't anything
earth-shattering; after all, My* (semi-) customizable, personal
portals were all the rage. What was unique about My Nescape was its
content-gathering mechanism. Rather than simply carry the standard
for-fee fare, Netscape created a channel description framework called
RSS ("RDF Site Summary") which allowed individual content providers to
syndicate snapshots of their latest offerings to the portal.
By providing a simple snapshot-in-a-document, web site producers
acquired audience through the presence of their content on My
Netscape. End-users got one-stop-reading, a centralized location into
which content from their favorite web sites flowed, rather than just
the sanitized streams of content syndicated into most portals. And My
Netscape, of course, acquired content for free. [ href="http://www.xml.com/pub/a/2000/07/17/syndication/rss.html">RSS:
Lightweight Web Syndication]
The original version of RSS, 0.9, was based on href="http://www.w3.org/RDF/">RDF ("Resource Description
Framework"), a framework for describing and relating resources on the
Web. This was followed shortly by RSS 0.91, re-dubbed "Rich Site
Summary," which had dropped its RDF roots and was instead described by
a static DTD at href="http://my.netscape.com/publish/formats/rss-0.91.dtd">http://my.netscape.com/publish/formats/rss-0.91.dtd.
Don't bother wandering over there though -- I'll explain why in a second.
Over time, folks started using RSS as a method for general metadata
and content syndication. As its use increased, so did its application
broaden. But the format itself remained unchanged, locked into its
static DTD and apparent disinterest by Netscape. Its semantics began
being overloaded, title and description elements being stuffed chock
full of dates, author information, and ad-hoc extensions began
cropping up. Extending RSS became a subject of much debate, resulting
in an eventual fork. The
technical point of contention was over a larger flat-file core with
centralized iterative extension versus a decentralized relational /
modular framework. But that's a story of another time.
With the fizzling and eventual acquisition of Netscape, the
My.Netscape portal atrophied; down-time, stale content, and broken
links foretold of its imminent demise. Recently revamped and turned
into a My.Yahoo!-alike, its distinctive framework and catalog of associated
channels have been taken offline.
In the decommissioning process they broke RSS 0.91.
RSS 0.91's reliance upon a DTD living on my.netscape.com has long been
an irritant. The site's staccato reachability caused some consumers
of 0.91 to turn to non-validating XML parsers so as to still have
access to RSS feeds even when the DTD was unavailable. Many did not.
Matt Sergeant and href="http://www.xmlhack.com">Edd Dumbill today noted the complete
removal of any trace of RSS, including the 0.91 DTD itself
from the Web. Wrote Dumbill in a posting to the href="http://www.egroups.com/group/rss-dev">RSS-DEV mailing list:
This breaks all RSS 0.91 feeds that are processed by a validating XML
parser and that reference the DTD, ie. huge numbers.
It presents us with several problems:
* has anyone got a cached copy
-- and --
* the DTD is (C) Netscape if I remember so hosting a cached copy is likely to be a violation of that copyright.
Failing any resolution the best things seem to be:
* short term, remove the DTD decl
* quick-as-you-can, move to a standard that doesn't rely on DTDs, ie.
RSS partisanship aside, this episode strikes yet another blow against
the use of centralized (specifically copyright) DTDs in an
increasingly distributed computing environment.