[Bug 27622] Pluggable SASL authentication

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=27622

--- Comment #23 from Sjoerd Simons <sjoerd at luon.net> 2010-05-17 04:47:12 PDT ---
(In reply to comment #22)
> Alright,
> 
> This branch deviated from it's name, but it is basically a set of changes that
> is required for pluggable SASL as done in Gabble. Any chance of review soonish?

Sure :)

b1283b8e6bd5703f378b65f216e4ddde9e076a5e

This patch actually does 3 things (add virtual functions, makes handler and
their list public and adds success functions). Would have been nice to split
these up in seperate patches for easier review.

Anyway, i'm unsure why you made the handler(s) public? rationale there would be
good..

For the virtualizing stuff, it's customary to let the actual implemention
allocate the GAsyncResult as it can choose not to use a GSimpleAsyncResult.
Also it should be possible to also override the finish method to not force the
Result to safe the data in a certain way. I'm ok with having the default one do
something sane, but in that case it will be an API guarantee and should be
properly documented

For the success function, should that be async ? If so the finish function
seems missing and also there is no definition of a finish method or a
definition of _success_async in the header

61d7de141597d74bd003e959b3d23fb33983b848

Might be nice to document that wocky_auth_registry_start_data_new takes
ownership of the initial_response so you don't have to do one extra copy. 

c0bcb52118db08eb76752b2a48df6a786f19e3a4 &&
cfeb9702d8695fa99469b01ad36081150e83870f

look fine :)

91690551e627ede09c673ce6405e4f5284ad1b71

looks ok, might be nice to document which error domains and also how to
distinguish between fatal errors and temporary errors (as in did the server say
failed or did it close the stream)

-- 
Configure bugmail: https://bugs.freedesktop.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the QA contact for the bug.