[Bug 14540] Names interface - Aliasing replacement with separate nickname, local alias etc.

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=14540

--- Comment #32 from Simon McVittie <simon.mcvittie at collabora.co.uk> ---
(In reply to comment #28)
> Still wondering how this will affect gabble. If I understand correctly,
> gabble needs to fetch vCard to get the nickname, so it cache the value on
> the roster. So the nickname becomes a local-alias for all contacts after
> first time the account is connected using gabble?

Yes, but that's a bug. It is wrong for a CM to mix up local-alias (a name the
local user chose, which can be assumed to be trustworthy) and nickname (a name
the remote contact chose, which might be malicious).

> With my mixin, each time NICKNAME attribute is asked,
> request_nickname_async() is called, so that will do a vCard fetch all the
> time, right? How would a on-disk cache in mixin avoid that? Should mixin
> avoid calling request_nickname_async() if it already has a local-alias?

I think the only solution to this is to distinguish between "get" and
"request", have "get" read from cache if possible or kick off a request if not
(with either the mixin or the CM implementing a cache), and have "request"
always round-trip to the server.

I vaguely feel as though Gabble caching vCards might be better than tp-glib
caching nicknames, but perhaps I'm wrong about that.

-- 
You are receiving this mail because:
You are the QA Contact for the bug.