Re: nm_device_wifi_get_access_point returns an invalid number of APs.

[pan son <sonicpan sonicpan gmail com>]

Thanks Dan,

Nope, it does not confirm with the output of nm-tool. The NMA applet and nm-tool both behave same.

I call nm_device_wifi_get_access_points everytime I get a "access-point-added" or "access-point-removed" and I see a lots of APs with ssid /or bssid? as (hidden), which is different from the output of nm-tool.

My piece of code is pretty small, and it is straight forward to fetch the list of Access Points. 

Thanks,
Vink

On Wed, Aug 11, 2010 at 10:32 AM, Dan Williams <dcbw redhat com> wrote:

On Tue, 2010-08-10 at 20:22 -0700, pan son wrote:
> Hi All,
> I am writing my UI client to the Network Manager.
> I periodically call the nm_device_wifi_get_access_points after every
> 60 seconds. What I see is everytime the number of APs returned are
> more then the previous call and most of them have ssid as (hidden). I
> also see lots of signal "access-point-added", but I dont see anything
> happening on the NMA applet as the same time.
>
> I dont see the same thing on NMA applet. Do I need to ref/ unref the
> NMAccessPoints everytime I refresh my screen.
>
> I am newbie and am just getting the hold of glib-gobject system. Any
> help is very much appreciated.

The objects returned by that call are owned by libnm-glib, so you only
need to unreference them if you ever explicitly called g_object_ref() on
them at any point.

When you see the problem, does the #  of access points returned by
nm_device_wifi_get_access_points() correspond to the output of
'nm-tool'?  I took a quick look at the code and it seems to be doing the
right thing, provided that NetworkManager is correctly aging the access
points and sending the AccessPointRemoved signal when they are removed
from NM's internal list.

Dan