Our aim ATM is for the new ‘dev facing’ API to be a much friendlier / simpler one. Advanced functionality should still exist, but hopefully we can cover enough with the basic API that there wont be a need to go deeper very often.
Our initial work on this is the API we’re using for the CLI. Built first in rust, you can see what’s expected re: params etc there. And here we have a couple of ‘layers’, one doing the abstract data stuff, eg: putting files on to the network (and producing
FilesMaps etc.). And these APIs use the lower (though still available), safe client libs wrapper APIs which offer functionality like [“put sequenced append only data”](https://github.com/maidsafe/safe-cli/blob/master/src/api/safe_client_libs.rs#L250 (which is still a wrapper on the core SCL functionality).
So most of this functionality should be what we’re looking to expose. The bredth of that, and how soon we expose those lower functions (vs building out the missing functionality for other data types etc), remains to be seen.
And then on top of that we have to build the FFI to all of this.
So what can you get your hands dirty with right now ? If you made a nodejs wrapper functions for CLI calls, the data you require for these function calls should be very similar to the final APIs eg.
I won’t say it’s fixed yet, but I’d be surprised if a call to put a file or ID changed much from our current implementation… (Though feedback on that if it needs changing from a dev/user perspective is invaluable!)
Hope that helps @happybeing!