nextPage Route planning algorithm now needs X requests. The fragments are cacheable (serverload/extra user ↓) Clients perform the algorithm on their own @PieterColpaert
services? → service/api is parameter-driven while websites are self-descriptive ◦ Actually, there is no difference, websites are apis too ◦ And, through hypermedia we can inform clients to more actions ⇒ There’s a thin line between data publishing and data services → Open Data is great, but some companies want to keep underlying things hidden, but do want to answer questions on the web. How? ◦ They don’t need to publish everything? Documents for public and API for others ◦ This is an access control problem: there is a spec for Access Control and the problem can be tackled in very flexible ways: class level, This is about data that is already public through e.g., a HTML or PDF: make it more useful by adding an open license, and by making it Web-queryable through wisely chosen fragments