[ENet-discuss] I am puzzled with the flow of send buffer
jaxecn at gmail.com
Mon Jul 30 00:39:09 PDT 2012
correct my last mail : Why the content which is send to peer A
is broadcasted to all peers .Why not only directly to send to the peer A.
2012/7/30 JiangHao <jaxecn at gmail.com>
> thanks for your reply .
> my point is the same bufffers is send to the different address of the
> peers.because the function 'enet_socket_send' is called in a loop of the
> host's peers and the buffer argument of the function is allways
> why the content is send to all other peer which is only send to peer A
> 2012/7/30 Lee Salzman <lsalzman at gmail.com>
>> The packet gets sent to the address of the peer, so it is thus received
>> by the remote host. The packet is tagged with the id of the destination
>> peer, so the remote host only has to index into its array of peers with
>> that id. After, it compares the address in the peer with the address of the
>> packet's sender to verify the packet came from the right source. There's no
>> real performance penalty for this.
>> On Mon, Jul 30, 2012 at 8:35 AM, JiangHao <jaxecn at gmail.com> wrote:
>>> hello ,
>>> I'm leaning the enet libaray,and i am puzzled with the flow of send
>>> buffer when i reading
>>> the implementation of enet_protocol_send_outgoing_commands,
>>> here is: sentLength = enet_socket_send (host -> socket, &
>>> currentPeer -> address, host -> buffers, host -> bufferCount);
>>> Is I made a mistake or you have it in the design? will it send
>>> the same buffers to all the peers of one host and filtered by
>>> remote host when it has receved these buffers.?
>>> the performance is not affected if it is send to all peers.isit?
>>> forgive my poor english .and thanks for answer.
>>> ENet-discuss mailing list
>>> ENet-discuss at cubik.org
>> ENet-discuss mailing list
>> ENet-discuss at cubik.org
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the ENet-discuss