Hi Daniel,
On Fri, Jul 01, 2011 at 06:15:42PM +0200, Daniel Wagner wrote:
> From: Daniel Wagner <[email protected]>
>
> I don't think this is a good solution because the magic to choose the
> right network for the service shouldn't be part of device. Maybe
> passing the networks to service to select the right one would be
> better?
It would make more sense for the network selection logic to belong to
service.c, even though I don't see how we would do anything else but a
strength based selection algorithm.
Some more comments below:
> @@ -1119,6 +1121,32 @@ int connman_device_remove_network(struct
> connman_device *device,
>
> g_hash_table_remove(device->networks, identifier);
>
> + network = NULL;
> +
> + g_hash_table_iter_init(&iter, device->networks);
> +
> + while (g_hash_table_iter_next(&iter, &key, &value) == TRUE) {
> + struct connman_network *iter_network = value;
> + uint8_t strength, iter_strength;
> +
> + if (network == NULL) {
> + network = iter_network;
> + continue;
> + }
> +
> + strength = connman_network_get_strength(network);
> + iter_strength = connman_network_get_strength(iter_network);
> +
> + if (iter_strength > strength)
> + network = iter_network;
> + }
> +
> + if (network == NULL)
> + return 0;
> +
> + __connman_device_set_network(device, network);
> + __connman_service_update_from_network(network);
So here we probably should provide a service.c routine (e.g.
__connman_service_reset_from_networks(GHashTable *networks)) that would
basically select a network from a networks hash table and then call
__connman_service_create_from_network() with it. service_update_from_network
will not set the right service->network pointer for you.
This routine should also set the service->network pointer to NULL if the hash
table is empty.
Cheers,
Samuel.
--
Intel Open Source Technology Centre
http://oss.intel.com/
_______________________________________________
connman mailing list
[email protected]
http://lists.connman.net/listinfo/connman
