Independent server and client about hydrus

[Mec-iS]

@pchampin Yeah, that is the idea. I started from a functional approach, so ideally even single features/functions of each of them (server and client) can be though as independent (different midllewares in the same server/pipe). The idea was to have a distributed architecture of different services using queueing and/or multistreaming. Let's strike the right balance for integration.

For the server we have a pretty defined starting point with the spec and Levanzo. For the client it will be probably an even more complicated task (Space Is Hard! (: ) and I have already some ideas ready to make it integrated with prediction-based and, in general, machine-learning libraries.

For now I think, to make everybody comfortable with the spec, we should focus on the server implementation. This way we can fully understand the features we are going to leverage, and when GSOC starts we will have all the material ready to start also with the client.

[Mec-iS]

[USEFUL LINK] For a very good implementation of a Web server for APIs, take a look to HUG

[pchampin]

Very nice, indeed! Having HUG automatically generate a Hydra documentation would be really cool...

…

On 24 February 2017 at 18:25, Lorenzo ***@***.***> wrote: [USEFUL LINK] For a very good implementation of a Web server for APIs, take a look to HUG <https://github.com/timothycrosley/hug> — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#5 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AASl5Pu_dKvoXMwDkX6Wc2S4FOyYh92Zks5rfxKagaJpZM4MLCuA> .