e martë prill 26, 2005

Bosworth's Web of Data: "Google's Adam Bosworth's keynote at the 2005 MySQL Users Conference was a call to audience members to "do for information what HTTP did for user interface." The web was successful because it offered a simple, sloppy, standards-based, scalable platform, and the challenge is to take a database and do the same. Daniel Steinberg covered Bosworth's talk, and provides this report."

...Bosworth predicts that RSS 2.0 and Atom will be the lingua franca that will be used to consume all data from everywhere. These are simple formats that are sloppily extensible. Anyone who wants to can use these formats to consume content or to author content...

(Via O'Reilly Network Articles.)

Looks like the ideas I express in our common JavaOne talk are not completely off topic: I'm in good company:-)

Very graphic, I like the dynamics of the arrows, which illustrate very well my belief that filtering syndication applications will be very important. I prefer it to his second entry.

Kenny Redman #1: "

Rome Logo Contest posted a photo:

This logo design basically shows how versatile ROME can be. I tried to use the colors of your web page.

wiki.java.net/bin/view/Javawsxml/RomeLogoContest

"

(Via Rome Logo Contest's Photos.)

Both hilarious and elegant.

Michael Lewcio: "

Rome Logo Contest posted a photo:

Rome burning - that's news.

wiki.java.net/bin/view/Javawsxml/RomeLogoContest

"

(Via Rome Logo Contest's Photos.)

I just sent Dave Johnson my slides for our common session on syndication at JavaOne. The first part is an introduction about syndication history, formats and standards. I've given similar talks many times for the past 3 years and always had too many slides. This time I concentrated on a typology of syndication applications, and reduced my 10 slides on the topic in a single graphic. This typology is based on technical characteristics, as opposed to the typology used by business people or VCs to partition these applications. If you want to read a more business oriented analysis I recommend Richard McManus' comments about the JupiterResearch report on RSS readers, and Brad Feld's thoughts about why he's hot on Feedburner, where he identifies 8 segments for RSS and blogging applications (Brad is a VC who invested in FeedBurner, Technorati and NewsGator).

Partitioning syndication applications based on their technical characteristics, client vs server, producer vs consumer, helps to highlight where the opportunities in this space are, from a technical perspective (after that you need to find the business model that goes with it:-).

My own take on it, and what I've been advocating at Sun for the past year, is that the biggest opportunities are on the horizontal axis, ie applications that both produce and consume syndication formats.

On the client side I call these Desktop Syndication Suites, examples being the venerable pioneer Radio UserLand, and the newcomer dynamic duo NetNewsWire + Marsedit on the Mac. I expect Office Suites (Open Office, MS Office) to follow this trend and integrate an aggregator and a feed producer.

On the server side I call these server side filters Server Syndication Engines. Examples are the grandfather O'Reilly Meerkat, that I use since a long time to filter out O'Reilly feeds, Feedster, oriented on search, Feedburner, oriented towards publishers, and AllConsuming which provides topical aggregation based on books. I expect this last category to grow a lot in the next few years. One of the category of software that will merge with this one is server side aggregators: today they are sinks for feeds, they read all, and spit it out in HTML for browser clients. I think they will evolve into server side filters, letting you consume feeds in feed readers (it could be client side, but also other server side readers or filters). Bloglines/NetNewsWire feed synchronization is a first step in this direction.

The other trend I talk about in my JavaOne presentation is syndication formats as a universal envelope format for time based queries. What this means is that for many web services application, when the operation is a query type, the envelope format won't be SOAP but Atom or RSS 2.0. The payloads will be embedded as namespaced extensions under the entries. The next evolution in syndication applications is UBL embedded in Atom! This is for this type of application that we built the modules support in our ROME library. Amazon's recent OpenSearch standard, where they added 3 elements to RSS 2.0, mainly for managing result pagination, in order to aggregate search results from various engines on a9, is a first step in that direction. I will also expand on this theme at our Xtech 2005 presentation in May (this time with Alejandro only).

Crazy ideas? Comments are welcome.

calendar

search

navigation

links

rss feeds

referers

Today's Page Hits: 314