[ale] Web Socket Implementations

[Richard Bronosky]

Your biggest issue with websockets is concurrency. At Cox we use a node.js
implementation because nothing else comes close to having a similar
connection capacity with the same hardware resources. We are in a unique
situation of needing to handle hundreds of thousands of concurrent
connections.


On Tue, Mar 19, 2013 at 11:44 AM, Hendry, Chris <Chris.Hendry at turner.com>wrote:

> Thanks Scott, the slide show was supportive.
>
> I lean on a server side implementation that would allow me to user either
> C,C++, or java. This would give me flexibility  I feel comfortable with.
> Implementations that only allow using python, node.js, or even erlang  I
> find too restrictive.
>
> For the client, do not see a way to keep way from using branded libraries,
> they just provide so much helpful tools, that goes for the server side as
> well. I like the models that jWebSocket provides:
>
> jWebSocket Channels concept:
> http://jwebsocket.org/devguide/dg_channels.htm
> jWebSocket Events model http://jwebsocket.org/devguide/dg_events.htm
> jWebSocket Token concept http://jwebsocket.org/devguide/dg_tokens.htm
> jWebSocket Remote Procedure Calls (RPC)
> http://jwebsocket.org/devguide/dg_rpc.htm
>
>
> Again, thanks
>
> Chris
>
>
>
> _______________________________________________
> Ale mailing list
> Ale at ale.org
> http://mail.ale.org/mailman/listinfo/ale
> See JOBS, ANNOUNCE and SCHOOLS lists at
> http://mail.ale.org/mailman/listinfo
>



-- 
.!# RichardBronosky #!.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.ale.org/pipermail/ale/attachments/20130320/e0acf894/attachment.html>