Re: NMSupplicantInterface PROP_CONNECTION_STATE not installed

[Dan Williams <dcbw redhat com>]

On Fri, 2009-06-05 at 19:12 +0100, Daniel Drake wrote:
> On Fri, 2009-06-05 at 09:16 -0400, Dan Williams wrote:
> > How about using the 'scanning' property instead?  That's exactly what
> > the 'scanning' property is for, and connection-state doesn't always
> > reflect when the supplicant is actually scanning (ie, it won't back down
> > from CONNECTED -> SCANNING when it's associated but can't find the AP in
> > an unsolicited scan list, but will direct-scan again anyway), which is
> > why I added the 'scanning' property and associated patch to
> > wpa_supplicant.  Will that do what you need?
> 
> I don't know.
> 
> I need a "subscribable" way of monitoring "is this interface currently
> scanning for networks." I found the function
> nm_supplicant_interface_get_scanning() which reflects both the
> "scanning" property (which is subscribable, great), and the
> "connection-state" property (which only half exists).
> 
> If I'm to believe the logic in that function, I need to subscribe to
> both "scanning" and "connection-state". The function seems to suggest
> that sometimes, priv->scanning will be FALSE even though the connection
> state is NM_SUPPLICANT_INTERFACE_CON_STATE_SCANNING.

It really shouldn't; I was being conservative there.  The 'scanning'
property (assuming you have that supplicant patch which is now upstream
and also available at [1]) should reflect the scanning state internally
in the supplicant.

If you want, I'll take a patch that makes the supplicant interface's
'scanning' property derived from both wpa_supplicant and the connection
state, and then you could still just use g_signal_connect (sup_iface,
"notify::scanning", G_CALLBACK (my_callback), my_data);

Dan

[1] http://cvs.fedoraproject.org/viewvc/rpms/wpa_supplicant/F-11/wpa_supplicant-0.6.8-scanning-property.patch?view=log