Re: [CsMain] [GSoC2011] Multi-Purpose UDP Networking System For Robust Client To Server Communication

Thanks again for feedback Christian!

On 4 April 2011 21:22, Christian Van Brussel
<[email protected]> wrote:
>
> That's it. And such a solution should be found and implemented for almost
> all propclasses (or at least for the most used ones).

Once I have done a couple, it should be easy to roll out to others.

>
> Note that iPcLinearMovement already implements dead reckoning, but this
> can probably be improved.
>

This is useful.

>
> But ENET will resend the same packet that was made for some time ago, and
> this won't include the changes that have happened in the game since then.
>

I believe this is not entirely true.

enet_packet_create(p_data, data_length, ENET_PACKET_FLAG_UNSEQUENCED);

Will create a packet that when sent, will never be resent, no
acknowledgement will be returned, etc. No quality of service control
what so ever.

http://enet.bespin.org/Features.html

"For unreliable packets, ENet will simply discard the lower sequence
number packet if a packet with a higher sequence number has already
been delivered. This allows the packets to be dispatched immediately
as they arrive, and reduce latency of unreliable packets to an
absolute minimum."

> A good UDP game layer should resend the packets delta compressed against
> the last time that was ackowledged, and this packet will be different each
> time it is re-sent since it will be kept up-to-date.

Agreed. this is the Quake 3 method I was referencing.

>
>> Can I ask who this is?
>
> You can usually find that kind of information in the header of the source
> files :p
>

haha you have been found out :p

Ok thanks for the feedback. I think a job for tomorrow is to revise my
proposal to focus on step 3.

Best wishes,

Dan

------------------------------------------------------------------------------
Xperia(TM) PLAY
It's a major breakthrough. An authentic gaming
smartphone on the nation's most reliable network.
And it wants your games.
http://p.sf.net/sfu/verizon-sfdev
_______________________________________________
Crystal-main mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/crystal-main
Unsubscribe: mailto:[email protected]?subject=unsubscribe