Hi Loic,

The Long polling transport method is the most "cross-browser" way to do
realtime request (unlike forever iframe, xhrstreaming, websocket, and so
forth).
This is what our framework do : http://www.ape-project.org/

APE is a client-side and server-side application. It can be easy
implemented to jpoker or another existing app. Sadly I didn't find the
time to work on a port of jpoker/APE, but I'm sure that someone can be
interested :-)

Anthony

Loic Dachary a écrit :
> Hi,
>
> I gave some thoughts to the implementation of long polling (
> http://en.wikipedia.org/wiki/Push_technology#Long_polling ) client
> side (jpoker & pok.me). At the moment, all XHR calls are queued in
> such a way that request sent at T+1 won't be sent until the request
> sent at T was acknowledged. If implementing long polling, the
> PacketPoll packet must be an exception to this rule. If there is a
> pending PacketPoll request, any other packet can be sent even if no
> answer was received.
>
> Server side it means that when a packet is received, if there is a
> pending PacketPoll it must return immediately, before the new packet
> is handled.
>
> I've not really thought this thru. I think this is all we need to
> implement long poll but I may have overlooked an important aspect of
> the problem.
>
> I'd like to hear what you think :-)
>

_______________________________________________
Pokersource-users mailing list
[email protected]
https://mail.gna.org/listinfo/pokersource-users


-- 
Anthony Catel - Développeur web & associé
Weelya - Improve the web
32 rue du faubourg boutonnet
34090 Montpellier
Tel : 04 67 169 778
http://www.weelya.com


_______________________________________________
Pokersource-users mailing list
[email protected]
https://mail.gna.org/listinfo/pokersource-users

Reply via email to