Re: squid-1.1.20, IE4 and HTTP 302 redirects

From memory, what all those 302 redirects from the .asp pages are doing are
interacting with the browser (gathering data in order to dynamically
generate the final page). What I figure is going on is that some of them are
getting cached, and MSIE which is expecting something different (more
in-sequence, or as part of an exchange protocol, perhaps), falls flat on
it's face.

But then, I have yet to go to a web-sit that uses .asp documents that
doesn't have a bunch of weird-ass bugs (It's okay boys and girls, I'm
australian. We're casually profane, by nature. It's considered polite here),
dead links and script-crashes. www.microsoft.com is a good example. (or a
bad one. Take your pick).

D

Tim Smith wrote:

> I have an interesting situation here, as follows
>
> NT Web Server(s) (IIS4) -> Squid 1.1.20 Accelerator -> MSIE4 browser
>
> (Please don't tell me not to use NT - I don't have a choice. I had to
> fight just to get a Sun box to front these monsters)
>
> There is a load of ASP stuff running on the NT servers, which issues
> HTTP code 302 redirects a lot. If the browser is anything other than
> MSIE4 (IE3 works fine, Netscape works, Lynx works, Mosaic works) then
> things go swimmingly. Also, when IE4 talks to the NT servers without
> going through Squid it works fine.
>
> However, in the situation above, with IE4 talking through the
> accelerator to the NT boxen, IE4 will *intermittently* fail to follow
> the 302 redirect. It gets it OK, the URL in the location bar changes to
> the new one, but it never fetches that new page, displaying "This page
> has moved..." as if it didn't support auto-redirects. The squid logs
> indicate that it never makes the followup fetch.
>
> I should probably describe it as intermittent success rather than
> intermittent failure, as once in about ten tries it will work.
>
> I strongly suspect a bug in IE4 (not that this was my first conclusion -
> honest!) as I've tried tracing the HTTP headers in squid and can't see a
> difference between what is returned to IE4 and what is returned to the
> other browsers, but I was wondering if anyone out there has seen this,
> and/or can suggest a workaround.
>
> Thanks for anything (I'm flummoxed).
>
> --
> Tim Smith
> (tim@zynet.net)

--
Did you read the documentation AND the FAQ?
If not, I'll probably still answer your question, but my patience will
be limited, and you take the risk of sarcasm and ridicule.