va*****@agilis.st.com.sg wrote:
I am developing a Web Interface for a embedded target using CGI-C. I
have a requirement, which requires a HTML page to be continuously
updated by the server.ie.,I am dynamically updating the fields of the
HTML page using Javascript. I generate the javascript code from "C".
I need to do the above operation as long as the user stays in the
Current Page.
You can't detect whether the user is 'in the current page', since
'current page' has no meaning to the server. I guess you you mean
'whether the user is still displaying the last page that their UA
requested from this server in one of their currently open windows or
tabs'.
All you can tell is whether a script you have installed in a page is
still polling or not (as you appear to be doing already), you can't
detect the 'current page' from the server, nor can you detect the URI
of one page or tab from another (except perhaps in the very limited
case of child windows opened by one of your pages).
If the user has clicked some other page, i want to stop
the operation in the server. Is there a way to notify the server
process when the user clicks some other page.
Only if that 'other page' is on your server and the user clicks on
your page to get there.
Even if there was some event that on page unload or window/tab close
would send some undisclosed information to the server, the user can
kill their browser or stop communication with the server a number of
ways that prevent the message ever getting through.
To that extent, your attempt to create a session would be extremely
unreliable to the point of being useless.
I dont want implement this as client pull operation, since i had to do
some other processing in the server
From the HTTP 1.1 specification:
Abstract
The Hypertext Transfer Protocol (HTTP) is an application-level
protocol for distributed, collaborative, hypermedia information
systems. It is a generic, stateless, protocol which can be used for
many tasks beyond its use for hypertext,
....
1.4 Overall Operation
The HTTP protocol is a request/response protocol. A client sends a
request to the server...
<URL:ftp://ftp.isi.edu/in-notes/rfc2616.txt>
In brief, HTTP is stateless, so the client and server only
communicate when asked to - there is no maintenance of a connection
or concept of a 'session'. HTTP is based on a client requesting
information, not the server deciding to sent it.
You may be able to use a script that is called at regular intervals
to poll the server and retrieve content at some specified interval,
but you can't do the reverse, i.e. get the server to poll the client
to see if it wants more information.
If your server does not get a request within some multiple of the
interval, you may presume that the user has 'left the page' and
terminate their session (as is done by many secure sites like banks
and stock brokers).
--
Fred
