Why shoudn't we use the PumpIO API?

Diaspora* needs an API, it’s a fact. Why don’t use the PumpIO API as MediaGoblin is doing? I am not a developer, I have no idea about APIs and about Diaspora* code (I know D* is coded in Ruby, but nothing more).

I think it could be a good option, I mean, it’s finished and it could be a good choice for pumpIO/GNU Social Connection.

Please, excuse my english.

Note: This discussion was imported from Loomio. Click here to view the original discussion.

The question is not what kind of api-endpoints we should be using. The question is who implements it.

Also there are existing discussions on API, please use the search. Opening new discussions just creates noise and won’t make anything happen either.

@jonnehass I know, but they are nearly a year old.

But still active.

https://www.loomio.org/search?utf8=✓&commit=&search_form[query]=API which one is active? lastest posts have months since they were wrotten. I don’t consider them actives.

Whether or not you consider them active, they are active, even if there have been no contributions to them for a while.

Anyway, I didn’t find the proposal of using the pumpIO API in anyone of these posts, maybe is not too bad idea posting that here. Now please, let’s focus in the Pump API thing.

The pump.io API uses the oAuth 1 + automatic application registration model. Since oAuths trust model is designed around a central registry of applications, automatic registration pretty much defeats it.

Ok, do you think this feature is easy to change for a automatic registration?


what “central registry of applications” ?

the application is registered with the site you are connecting it to, i don’t see any other sites involved in that process.

maybe I’m not understanding what you mean by “automatic registration”.

I agree on implementing the PumpIO API.
I don’t know why developers are delaying so long to decide this.

@eduardo4 Because there are other existing standards out there; arguably some of them could be said to do things better than what Pump does, and implementing a new standard protocol requires a lot of internal refactoring for Diaspora.

2years after this https://joindiaspora.com/posts/1a773b6f9381744d and still no api? sean, i’m disappointed.

@harrytuttle https://joindiaspora.com/posts/3202188

If you look, you’ll see that D* as a project has made lots of progress since becoming a community-run project. However, developing an API for decentralized applications and getting everything right can be challenging, especially if you’re just a volunteer developing something.

The latest development on an API implementation can be found here.

Two years you are following the project and you never wrote code? @harrytuttle, I’m disappointed.

That’s just a joke, don’t take that seriously, but I start to become tired of “you should do that, I’m disappointed of what you are doing”. We do not need to be reminded everytime about the work to do, we know it, and want to see it done more than everyone else. We need help, that’s all.

^^ Exactly what Sean and Fla have said.