(6) Location-Based Services

(7) Location and the Web

• The current Web is unaware of location as a concept
  • it supports other concepts such as media types and languages
  • many applications use location and expose location in the UI
  • mobile access [The Mobile Web (1)] and location-oriented services become more widespread
• How to express location as a Web-level concept?
  • should there be a URI scheme for locations?
  • should there be location support in HTTP?
  • should there be location metadata in HTML?
  • should there be location access in Web Apps? (there is [http://www.w3.org/TR/geolocation-API/])
  • should there be location concepts in RDF?
• Thinking beyond location-oriented applications
  • applications typical are closed verticals
  • mashability and repurposing requires exposed application models

(9) Mobile Phones

(10) Native vs. Web

• iPhone switched from Web-based to native
  • iPhone OS 2.0 introduced native apps and the App Store [http://www.apple.com/iphone/appstore/]
  • the native app model have proven to be the killer feature of the iPhone
  • other mobile platforms try to replicate the model
• In-App Purchases allow apps to use micropayments
  • it could be The Rise of The Micropayments [http://dret.typepad.com/dretblog/2009/03/the-rise-of-the-micropayments.html]
  • so far it has not seen a lot of widespread use
• Phones will have increasingly powerful browsers
  • phones might also have only browsers (ChromeOS)
  • Web Apps [Web Apps (1)] might be a better long-term strategy
• Currently too many players on the mobile OS market
  • consolidation is inevitable
  • the Web is moving to Web Apps [Web Apps (1)], why not the mobile Web as well?

(12) HTML5 for Advanced Web Apps

(13) The CSS3 Effect?

• HTML5 itself defines the language and the DOM
• Web Apps need additional functionality in terms of APIs
  • push services: Server-Sent Events [http://www.w3.org/TR/2009/WD-eventsource-20091029]
  • TCP over HTTP over TCP: Web Sockets [http://www.w3.org/TR/2009/WD-websockets-20091029]
  • key/value database: Web Storage [http://www.w3.org/TR/2009/WD-webstorage-20091029] or WebSimpleDB [http://www.w3.org/TR/2009/WD-WebSimpleDB-20090929]
  • SQL database: Web Database [http://www.w3.org/TR/2009/WD-webdatabase-20091029]
  • threads: Web Workers [http://www.w3.org/TR/2009/WD-workers-20091029]
  • file system access: File API [http://www.w3.org/TR/2009/WD-FileAPI-20091117]
• Turning the browser into an OS
  • the Java vision without the Java language
  • where do you stop? (3D rendering, device access, introspection)

(15) Understanding the Web

(16) Understanding Resources

• Finding needles in a haystack
  • search engines are amazing but limited in their understanding
  • more understanding of resources allows better resource access
• Semantics are always embedded into perspectives/models
  • who has the right to understand/describe resources?
  • how do I know how the resource was interpreted and described?
  • how can I find alternative descriptions?
  • where is the line between data and information/knowledge?
• The tough challenges are not on the technical level
  • kid-friendly search engines
  • browser-level or OS-level content filtering
  • Internet-level or Web-level filtering

(18) Connecting Real-World Objects

(19) Internet vs. Web

• Internet of Things is the current buzzphrase
  • a useful and necessary first step (establish connectivity)
  • not enough to enable ubiquitous accessibility and usability
• Apply Web architecture to the Internet of Things
  • expose Things via REST [Representational State Transfer (REST)]ful Web services
  • from the Web's perspective, things are just resources
• Some new twists and challenges have to be solved
  • how to find things (you cannot search for them with Google)
  • how to find the interesting things (there are too many things)
  • how to interact with them (is HTTP [Web Foundations (URI & HTTP); Hypertext Transfer Protocol (HTTP) (1)] sufficient as the uniform interface?)
  • how to support real-time [The Real-Time Web (1)] interactions (push instead of pull)

(21) Tracking in Real-Time

(22) Pull vs. Push

• Pulling is the underlying principle of the current Web
  • interactions are initiated by the client
  • servers don't have to keep track of their clients
• Translating between the Pull and Push patterns
  • translating push into pull is trivial: store the push, wait for the pull
  • translating pull into push requires some optimization
• Pushing is faster but more expensive
  • scalable pushing infrastructures can be built (iPhone push notifications)
  • building it in a distributed way makes it more flexible and scalable
• PubSubHubbub [http://code.google.com/p/pubsubhubbub/] builds a push model around feeds
  • feeds declare that the content is also available through a hub
  • hubs can be configured into a multi-level hierarchy
  • clients connect to hubs and are notified by their hub