GSOC2016 - Idea #3: Implements SASL mechanisms about erlang-dbus HOT 5 CLOSED

I suppose the title sould be refined as, while D-Bus specifications tells every SASL mechanism should be supported, they are not all implemented by reference implementation (C one) and non-SASL authentication mechanism are implemented (nonce-TCP transport...).

In order to be fully compatible with D-Bus existing implementations, it would be preferable to:

1. identify all required mechanisms (SASL and non-SASL)
2. identify all required transports (I suppose transport and authentication mechanisms are not 100% decoupled)
3. Check which combination should be prioritized and doable in a GSoC.

I consider solving 1 and 2 is a prerequisite for being a good candidate for 3 ;)

from erlang-dbus.

Ah, i see, i thought the ideas were separate projects that could be picked independently. I find the whole project interesting but i have doubts on wether i could allocate enough time for it, what would be your estimate of the workload for the whole thing?

from erlang-dbus.

I don't have a precise idea of the workload as I have not yet identified what mechanisms are really required.

That's why I propose, as a first step, to identify these mechanisms (which shouldn't take more than a half day...) then discuss which ones are feasible as GSoC.

from erlang-dbus.

Got it, closing issue, hopefully will open a new one with a better analysis

from erlang-dbus.

Great, thanks !

from erlang-dbus.