the problem with the headers such as expire and no-cache is that they just don't work erliably since many browsers ignore tham and there are intermediate places such as ISPs that sometimes cache.
When I have this problem I make data be displayed by a script such as:
showresults.cgi
That script also uses one additional parameter, not otherwise needed, I call it onetime. So the invocation of the script is something like:
showresults.cgi?onetime=10
Now, the value 106.... is actually the time that the page making the invocation was created with the process id added to it. In this manner I can ensure that no two users ever ask for the same page.
Main Topics
Browse All Topics
by: kanduraPosted on 2003-11-05 at 01:23:44ID: 9685367
you could add an 'expires' header to the response, by using this:
print header( -expires => '-1d' );
But are you sure it is a problem in the first place? I don't think client A would have the report for client B in his cache in the first place. Not unless he requested it before.
It may be just you who is having the problem.
It's not obvious from your code, but shouldn't there be a parameter in there that tells you which client is making the request? How do you know which report to generate?