[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [RSS-DEV] Re: [syndication] Time for XHTML-RSS?
Tristan Louis wrote:
WARNING: Long rant on XHTML and RSS and how the two SHOULD be separate....
requirement is dropped from the XHTML spec (it could be made optional); and if
the root element of the document no longer must be <html> (that too should be
made optional) then XHTML becomes modular enough to be embedded in anything,
including an RSS feed. That solves the first problem of embedding XHTML in RSS.
The next question is to figure out how to embed other elements in XHTML. Right
now, the W3C does NOT offer a way to properly embed other things in XHTML. As a
result, there is no way to add "smarts" to an XHTML document. This is a big flaw
as it essentially says that, if we are to build a smarter web, we will have to
do it in parallel to the existing one. If you look at this long term, it seems
like something that will be very unwieldy and difficult to maintain.
All you really have to do is stick the RSS elements within the xhtml
document, and leave out the doctype. Even for RDDL, the doctype has
been a problem with browsers, so everying leaves it out (including the
rddl page itself http://rddl.org/ -- they comment it out).
To be useful, a HTS (HyperText Syndication) document need be:
- well formed
- contain necessary tags to indicate items
- of reasonable size
An aggregator should process by:
- ignoring all content in the file until it finds the first syndicated
item (probably a channel element, although I have received suggestions
that only items are required), Just toss it away as it streams through
the document.
- refuse to process if too much client bandwidth is being consumed.
- syndicate only the channel portion
- stop reading and processing content once the channel element is closed.
A publisher should use this format to:
- easily publish documents for syndication that are not too big
- learn about RSS and syndication and how to publish and syndicate content
A publisher should switch to RSS if:
- they can manage running a tool to extract (not scrape) RSS from html and
- the number of user-agent requests from entities other that browsers is
signficant and
- bandwidth is a concern to them or their clients.
TNL
long winded, maybe, and trying to make sense with a single cup of coffee :)
--
Doug Ransom
Hate spam & pop ups? Try Mozilla for web/ëmail.