Hi all. I'm using SL 4 beta and my app needs to do a lot of small http requests to the server.

I believe that when exceeding the number of allowed concurrent requests, the subsequent requests are put in a queue. I am also aware that SL 4 has both a http browser stack and a http client stack, with both different limit in terms of the number of concurrent requests.

Let's say call those limits MAX_BROWSER and MAX_CLIENT.

Also I think I read somewhere that the number of concurrent requests is limited per domain, not overall. But I'm sure if this applies to both the http client stack. That means that you CAN have MAX_BROWSER requests to domain1.com AND MAX_BROWSER requests to domain2.com at the same time. And I even believe that sub domains are considered different so you can also have MAX_BROWSER requests to domain1.com AND MAX_BROWSER requests to sub.domain1.com at the same time.

I have ownership of the services and domain names so I could easily setup sub domains for my services.

Given those considerations I'm trying to optimize the number of concurrent http requests to my server. Here are few questions ?

Is is possible to use both stack at the same time ? Is the subdomain/domain story true for both stacks ? None ?

If so that would mean that I could potentially have a number of concurrent requests equal to : (MAX_BROWSER + MAX_CLIENT) * NUMBER_OF_DOMAINS which would be fairly good.

Is this correct ?

I'm kind of sharing my morning thoughts here, hoping somebody has experimented with those things.

Thank you.