This is great stuff!
What would be amazing, is to use this to make a web-application that lets webmasters run a Twister-'endpoint', i.e. that applications not directly connected to the Twister network can call to receive messages and user information about a certain search query. Just like e.g.
twisterio.com, but returning simple JSON (or XML) data. The best thing would be to decentralize this a little, by having these systems could let each other know through the Twister network that they are live (and regularly send eachother a ping), and also send this list with known live twister-endpoints to the end user application, which might cache it to ensure that when one of the endpoints is down, the data can still be received from one of possibly many alternative sources.
I'm going to explore this possibility, as soon as I get the TwisterRPC connection to work...
Op donderdag 9 april 2015 10:54:44 UTC+2 schreef Arco Mul: