[Bug 21787] Connection.Interface.ContactLists

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=21787

--- Comment #37 from Mikhail Zabaluev <mikhail.zabaluev at nokia.com> 2010-06-30 09:39:34 PDT ---
(In reply to comment #34)
> I think the timeout is a UI decision. For how long is it acceptable for your UI
> to display a "please wait" spinner before giving up in disgust - a minute, an
> hour, 24 hours? However long that is, that's the timeout you should have.

This is orthogonal to the way of signalling. But I realize that as a
non-modifying method with natural semantics of failure (timed out => no
contacts available yet), GCLA timing out is not such a big problem.

> I think a boolean property with change notification could be a good idea too,
> though. The change notification could be implicit (any ContactsChanged,
> GroupsChanged, etc. signal implicitly indicates the transition to
> ContactListRetrieved=TRUE), or explicit.
> 
> (In reply to comment #33)
> > (In reply to comment #32)
> > > Maybe a property/signal pair could be added
> > 
> > Also useful to expose the state when contact list retrieval fails.
> 
> Is your idea that this is state-recovery, something like this?
> 
> property ContactListFailure: s
>     The D-Bus error name with which retrieval of the contact list failed,
>     or the empty string if retrieval succeeded or is still in progress

I was thinking more about an enum ContactListState =
(Initial|Retrieving|Retrieved|Failed)

The error could go to the method return, if you want to keep GCLA as a
potentially slow roundtrip.

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