[ENet-discuss] Broadcasting to multiple servers
goosey at gmail.com
Tue Feb 20 13:12:14 PST 2007
I ran into this same issue. Currently the solution I am using is to drop
down to low level UDP on the servers to broadcast their network information.
Clients listen to this and are able to form a list of the servers on the
network, and initiate connections to them in a non-broadcast manner.
Not sure how to have a more enet-integrated solution to this issue (which
would be nice), but the issue (as far as I can tell) is that my enet-clients
only allow 1 connection, and so all the servers 'jump on that connection'
and whoever jumps quickest gets the slot...
So random thought that might cure: allow multiple connections on the client
might be a way to make this work (although I have not tried this yet..)
Let us know how it goes! :)
On 2/20/07, Ben Moreno <ben at blitwise.com> wrote:
> Hello all,
> I'm running into a problem when trying to use ENet to find servers on
> a local network. I need to be able to be able to broadcast a packet
> such that any servers that are listening can either respond in kind
> or initiate a connection. As far as I can see, there is no way of
> being sure that you've connected to more than whichever server
> happens to reply first.
> The servers I'm using are simple ENet servers, essentially from the
> tutorial. The client (after initializing everything as usual),
> creates a peer with the address set to ENET_HOST_BROADCAST, then
> loops enet_host_service until it receives a connection event. After
> that, a second peer is created, which also broadcasts. This peer
> will usually end up connecting to the same server.
> Switching the order of calls around doesn't seem to help. For
> instance, waiting to run enet_host_service() until after both peers
> have been created gives the same behavior (usually).
> Am I missing something, or is connecting via ENET_HOST_BROADCAST
> limited to which server happens to respond first? Is there any way
> of reliably knowing how many servers responded and who they are?
> ENet-discuss mailing list
> ENet-discuss at cubik.org
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the ENet-discuss