[ENet-discuss] Only next to last packet out of order?

[Pablo de Heras Ciechomski]

Yes I am curious about the behaviour itself, as it's so consistent almost
algorithmic and looks like an off-by-one error.

I would not have blinked to a completely shuffled packet order but this
seems, synthetic.

Pablo

On Fri, Apr 3, 2015 at 7:03 PM, Laftur Jeremy <laftur.jeremy at gmail.com>
wrote:

> I suggest attempting to replicate this behavior with using UDP only.
> Success should indicate that Enet is not responsible. I assume you
> don't expect unreliable packets to be received in any particular
> order, and are just curious of this consistent behavior.
>
> On Fri, Apr 3, 2015 at 12:07 PM, Pablo de Heras Ciechomski
> <pablo.deheras at gmail.com> wrote:
> > Hi,
> >
> > 1) I am sending with unreliable packets < MTU
> > 2) I am sending packets in bursts (30 or more but less than 70 in a short
> > time)
> > 3) I have tested with the server and the client on the same machine
> > 4) I have tested with the server and the client on the different machines
> >
> > Now the weird part.
> >
> > I send the packet bursts like this
> >
> > 0 1 2 3 4 .... n
> >
> > and received them as
> >
> > n-1 0 1 2 4 ... n
> >
> > Meaning the next to last packet goes out first on the line. I know that I
> > send them in order because if I change to reliable sending the order is
> > correct 0 to n.
> >
> > Is this due to ENet internal buffers or due to some network drivers?
> >
> > I found this while trying to home brew a packet loss formula
> > (maxPacketIndex-minPacketIndex kind of thing) and saw how packet loss
> peaked
> > and fell in periods. This is due to the first packet being much larger
> index
> > indicating a fake packet loss with this strategy.
> >
> > Best regards,
> >
> > Pablo
> >
> > _______________________________________________
> > ENet-discuss mailing list
> > ENet-discuss at cubik.org
> > http://lists.cubik.org/mailman/listinfo/enet-discuss
> >
> _______________________________________________
> ENet-discuss mailing list
> ENet-discuss at cubik.org
> http://lists.cubik.org/mailman/listinfo/enet-discuss
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.cubik.org/pipermail/enet-discuss/attachments/20150403/000e36ca/attachment.html>