Re: Is HTTP Stateless ?
From: Eric Biggs (e)
Date: 03/20/04
- Next message: Sebastien Lambla: "Re: Application configuration"
- Previous message: Teemu Keiski: "Re: Field Validation - Can you tell me what I'm doing wrong?"
- In reply to: Sami Vaaraniemi: "Re: Is HTTP Stateless ?"
- Next in thread: Polaris: "Re: Is HTTP Stateless ?"
- Reply: Polaris: "Re: Is HTTP Stateless ?"
- Messages sorted by: [ date ] [ thread ]
Date: Sat, 20 Mar 2004 12:29:44 -0800
Right, and the main reason for creating a tcp connection upon each request
and response is for reliability reasons... the connection is mostly used for
resending dropped packets etc.
"Sami Vaaraniemi" <samivawantsnospam@jippii.fi> wrote in message
news:c3i6n6$nq5$1@phys-news1.kolumbus.fi...
> Yes HTTP is stateless even if it runs on top of TCP.
>
> A HTTP conversation essentially consists of a single request and a single
> response. After that the "session" is usually gone. Two subsequent
requests
> from one client can come through two separate TCP connections. Web servers
> do optimize things by keeping the connection alive for a while in case the
> client sends another request, but basically all that HTTP sees is one
> request and a response.
>
> In order to maintain session in the web application, programmers need to
> associate seemingly unrelated HTTP requests from one client somehow. In
> practice this is done with cookies or storing cookie-like information in
the
> URL.
>
> Sami
> www.capehill.net
>
> "Polaris" <etpolaris@hotmail.com> wrote in message
> news:urAhF0qDEHA.2768@tk2msftngp13.phx.gbl...
> > Read a few books on ASP.NET and seems to me all of them say that HTTP is
> > stateless (as a result, it is a complex job to maintain the session
state
> of
> > a web application ,,, blaa, blaaaa...). I just wonder if this statement
is
> > correct? HTTP runs on top of TCP; and TCP is a connection-oriented
> protocol
> > in which each conversation (session) is established through a 3-way
> > handshaking and the session state is maintained by the TCP stack until
the
> > client signals to end the conversation (in ASP.NET's web application,
this
> > action could be the user's explicit "logout", or his session expires).
> >
> > ANY thoughts on this is appreciated.
> >
> > Thanks
> >
> >
> >
>
>
- Next message: Sebastien Lambla: "Re: Application configuration"
- Previous message: Teemu Keiski: "Re: Field Validation - Can you tell me what I'm doing wrong?"
- In reply to: Sami Vaaraniemi: "Re: Is HTTP Stateless ?"
- Next in thread: Polaris: "Re: Is HTTP Stateless ?"
- Reply: Polaris: "Re: Is HTTP Stateless ?"
- Messages sorted by: [ date ] [ thread ]
Relevant Pages
|
