[ENet-discuss] Optimizing ENet Protocol ?
manu.n02 at laposte.net
Thu May 24 03:57:32 PDT 2012
1) maybe peerID could be used for anti packet spoofing, but as it's
none essential, it'd be better to have it optionally there
2) dataLength can likely be set by the decompressor, it must know how
much data it expanded ! ;-)
3) I rechecked and reliableSequenceNumber is actually set in
enet_peer_setup_outgoing_command , I had found 1st a
"reliableSequenceNumber = 0" but it wasn't the correct one...
:-S But I checked the reception code and couldn't figure out where
the reliableSequenceNumber was used for unsequenced messages ; it may
be easier to follow in the debugger once I'll have something running
At 17:41 24/05/2012, you wrote:
>Interesting thoughts! I'm not worried at all about the overhead in
>ENet, but things like peerID have caught my attention earlier as I
>didn't understand why it was there.
>I still don't understand why it's there as (like you said) we could
>just map each Peer to the source address and be fine with it - but
>maybe there's more to it than that that I haven't noticed yet.
>dataLenght I'm not so sure about - what if we compress the data? We
>need to keep the datalength for the uncompressed size in ENet.
>Are you sure that reliableSequenceNumber always is 0?
More information about the ENet-discuss