Re: Websockets on the IBMi

On Jul 11, 2012, at 12:38 PM, "Nathan Andelin" <nandelin@xxxxxxxxx> wrote:

There is big difference between a keepalive (apache) and a websocket.
A keepalive may die ...

I'm not just talking about keepalive "on". I'm talking about Timeout "nnn", and KeepAliveTimeout "nnn". The browser and the server will maintain persistent connections for the duration of "nnn", which can be as long as you want.

Yes, persistent connections, but not bi-directional. Additionally, you will be limited to the number of forked Apache processes. While it is no guarantee, most servers that handle WS will not limit you like that, instead, connecting to some kind of queue,

If a browser issues an asynchronous request via XHR, and the socket times out on the server, the browser will just get an empty response after minutes or hours of wait. Just code to cycle another request.

a WS remains open except in case of an error and is bi-diectional.

I still stand by my earlier assertion that WS is implementing a request-response cycle under the covers.

It is not. I built a setup last week using ActiveMQ, JMS and Stomp over websockets connecting to PHP over FastCGI. You are able to send data over WS whenever you like.

Consider the following proof of concept which simulates 4 chat clients establishing 4 persistent connections using XHR. If you use a tool like Fiddler to measure the interval between the 1st response and the 4th response, it often is withing a millisecond or two on a LAN.
http://www.radile.com/rdweb/temp/meet.html


-Nathan

--
This is the Web Enabling the AS400 / iSeries (WEB400) mailing list
To post a message email: WEB400@xxxxxxxxxxxx
To subscribe, unsubscribe, or change list options,
visit: http://lists.midrange.com/mailman/listinfo/web400
or email: WEB400-request@xxxxxxxxxxxx
Before posting, please take a moment to review the archives
at http://archive.midrange.com/web400.