On 31 January 2013 19:31, Nick Jennings
<address@hidden> wrote:
Hi Melvin,
On Thu, Jan 31, 2013 at 5:53 PM, Melvin Carvalho
<
address@hidden> wrote:
>
> Why limit this to activity streams over JSON? Surely the system should be
> independent of the serialization for a true polyglot approach.
streams, I'm not sure they could be considered 100% activity stream
compliant. We just adopted the base-level properties "actor" "verb"
"target" "object".
There is no marriage to activity streams, and other data transfer
formats could be used ( code submissions welcome. *cough* :), but the
point here is to provide a simple, and easy to use API for developers,
this is not a 'web-facing' API, but more a library for developers (the
library just happens to communicate via WebSockets).
Sounds good.
What will the API look like?
[ Identifier A ] sends message to [ Identifier B ]
For example?
Even tho the web has been around for 2 decades, we are nowhere close to even solving the simple use case. There are many competing protocols that have balkanized the social web. I'm interested to see the approach that will be taken here. Also dont forget that the web was designed to be social from the start, and almost everything talks HTTP these days. To date, only facebook have really leveraged this, we really need something we designed under FLOSS in this space!
> Also note that there are many serializations that are patent and royalty
> free that can do this job, activity streams is not one of them.
Could you provide some links? Activity streams involve royalties? This
is news to me.
All standards under the IETF and W3C are patent and royalty free. I dont have a problem with activity streams, in fact my name is on the spec. But AS is outside of the recognized standards process, and more business oriented.