On 7 August 2011 12:49, George Kiagiadakis <kiagiadakis.george at gmail.com> wrote:
> On Sat, Aug 6, 2011 at 10:35 PM, David Edmundson
> <david at davidedmundson.co.uk> wrote:
>> Rather than talking, I've just attached a bodged header file for a
>> class that calculates this.
>>
>> /** Retrieves the current 'global presence' based on the current
>> status of all the accounts*/
>>
>> class GlobalPresence
>> {
>> public:
>> //FIXME account set or AccountManager?
>> GlobalPresence(const Tp::AccountSet &accounts);
>> virtual ~GlobalPresence();
>>
>> /** The current global presence, this will return the most
>> 'online' presence */
>> //online > away > extended away > busy > offline
>>
>> //FIXME maybe this will make more sense as a simplePresenceType?
>> Tp::Presence globalPresence() const;
>>
>> /** Returns Tp::Connecting if any account is connecting
>> * Otherwise Connected if any account is connected.
>> * Disconnected Otherwise
>> */
>> Tp::ConnectionStatus globalConnectionStatus() const;
>>
>> /** Try setting all accounts to the given presence
>> * This may fail, watch the status of account for details.
>> */
>> void setGlobalPresence(const Tp::Presence presence);
>>
>> signals:
>> globalPresenceChanged();
>> globalConnectionStatusChanged();
>> }
>
>
> Where is this code going? In the dataengine?, the contact list?, both
> perhaps?, in libktelepathy?...
>
> I was thinking that it would make a lot of sense to have something
> like that in tp-qt4/tp-glib and possibly mention this behaviour in the
> spec as well, so that other clients (empathy) have the same concept of
> global presence. Any thoughts?
Sounds like a good plan if upstream are also happy with it.
> Also, regarding the order: online > away > extended away > busy > offline
> I think that "busy" is more online than "away", as it indicates that
> you are there, probably working, in contrast to "away", which means
> you are not there.
Thinking about it some more, I think you are right here.
--
George