[Lift] Re: (Maybe bug)comet long polling into javascript loop?

2009-12-18 Thread Marius
This is not a bug. When Lift detects multiple requests for the same session it breaks out the Comet request. This is a safety mechanism in order to avoid HTTP connection starvation on the client side. This is because browsers have a very limited connection pools and the number of parallel HTTP

Re: [Lift] Re: (Maybe bug)comet long polling into javascript loop?

2009-12-18 Thread Andrea Peruffo
And this does not affect performances? or server side computations? Andrea Il Friday 18 December 2009 10:50:17 Marius ha scritto: This is not a bug. When Lift detects multiple requests for the same session it breaks out the Comet request. This is a safety mechanism in order to avoid HTTP

[Lift] Re: (Maybe bug)comet long polling into javascript loop?

2009-12-18 Thread Marius
Not sure what you mean ... what concerns do you have about performance? ... If you are referring to this mechanism built in Lift, there is virtually no performance impact. This is very lightweight. If you are referring to something else, please provide more details about your concerns. Br's,

Re: [Lift] Re: (Maybe bug)comet long polling into javascript loop?

2009-12-18 Thread David Pollak
On Fri, Dec 18, 2009 at 1:58 AM, Andrea Peruffo a.peru...@mcmspa.it wrote: And this does not affect performances? or server side computations? The short answer is that it does increase the amount of HTTP traffic, but with HTTP keep-alive, that's not a lot of new TCP/IP connections. On the