May 18, 2006

XTech - day 2 - part 2

Native to a Web of data: Designing a part of the Aggregate Web

"Internet will not listen to reason" - is the image we can see in the beggining of this talk. Tom is working at Yahoo! and before he worked at BBC, but he's talking about he's views and not BBC's or Yahoo!'s.

Design and Web 2.0 - Blogger started it

Web as an environement is changing...

What is the web changing into? Lot's of stuff is changing, but this presentation is going to concentrate about data and information. We're trying to walk into "an aggregate web of connected date sources and services"...

Web of data - A web of data sources, services for exploring and manipoulating data, and ways that users can get toguether.

Mashups - First we got a set of pages, then we got mash-up's, and now we're getting mash-up's of mash-up's. We'll get a network effect of services: Every new service can build on top of every other existing service - the web becomes a true platform

Consequences - massive creative possibilities; accelerating innovation; increasingly competitive services, componentised services and specialised services

Making money using the web of data:
* Use API's to drive people to your stuff
* Make your services more attractive and useful with less central development
* Use syndicated content as a platform
* At the end, you can turn your API as a pay-off service (or a better version)

Choosing what to build (in this environement)
* What can I build that will make web better?
* How can I add value to the Aggregate Web?

Open up your data, others will beneffict from it, and so you'll bennefict from their bennefict

Or you're the only player, or the winner will be the company that first reaches critical mass via user aggregation, and turns that aggregated data into a system service

Services doesn't need to create data, but can just exploring data or manipulating it

Architectural principles:
* Last year there was a presentation about "Designing Data for Reuse" by Matt Biddulph
* Data sources
* Standard ways of representing data
* Identifiers and URL's
* Mechanisms for distributing data
* Ways to interact with / enhance data
* rights frameworks and financial

Principles:
1 - add valuue to the aggregated web
2 - you have to build to normal users, developers and machines
3 - Start by designing explorable data, not pages
4 - identify your first order objects and make them addressable (point to it, link to it, readable url's, ...)
5 - correlate with external identifier schemes (or coin a new standard)
6 - use readable, reliable and hackable URLs
* permanent references to resources
* have a 1-to-1 correlation with concepts
* use directories to represent hierarchy
* not reflect the underlyings technology
* reflect the structure of the data
* be predictable / guessable / hackable
* be as human readable as possible
* be - or expose - identifiers
7 - Build list views
* three core types of page: derstination, list-view and manipulation pages (flickr is an example)
8 - Create parallel data services (API's, Microformats, Parallel XML, RSS...)
9 - Everything should have an appropriate license (Creative Commons, for instance) so people know what they can do with your data (this is one thing I need to do...)

If you're running a business, use APIs to drive people to your stuff, make your services more attractive and useful with less central development

(this guy runs plasticbag.com or something like that)

(not really about this talk, but I need to take a look to feedrinse.com )
(not really about this talk, but I need to take a look to microformats.com )
(not really about this talk, but I need to take a look to odeo.com )

No comments:

Post a Comment