Hi, Daniel
>
> I am not sure if this is what you want to do. In
> connman_network_create
> we do following:
>
> struct connman_network *connman_network_create(const char *identifier,
> enum
> connman_network_type type)
> {
> [...]
> if (identifier == NULL) {
> ident = g_strdup_printf("hidden_%d", hidden_counter++);
> network->hidden = TRUE;
> }
> [...]
> }
>
> I have not followed through the whole call stack from here, but I am
> assuming that in this case we don't need the network->hidden
> anymore and
> also the corresponding code in service.
>
I agree with you here, but I think we should remove the corresponding code in
network.c
instead of service, and I will submit a updated patch later to remove the code
related to
network->hidden in network.c, and I hope you and sameo can review it when you
have time.
Any suggestions, feel free to point them out, thanks.
> Samuel and I started to discuss this on IRC but we haven't found yet a
> solution. My guts feeling tells me that create_hidden_wifi() should be
> in the wifi plugin.
>
> Reasoning: Since all networks and providers are created and maintained
> now in the plugin layer it also should go there. If
> create_hidden_wife()
> is in wifi, we can easily identify our own hidden networks when the
> network_added() callback is running.
>
I am not sure whether create_hidden_wifi() will be moved to wifi plugin, and if
it would not be moved
in near future, I think maybe we can integrate this patch into connman to fix
the hidden wifi issue firstly.
What's about your thoughts?
> The downside of my proposal is that we would have to extend the struct
> connman_network_driver somehow and passing all the arguments to the
> plugin. Hmm, this wont be very generic then. Any better ideas?
>
_______________________________________________
connman mailing list
[email protected]
http://lists.connman.net/listinfo/connman
