[thelist] Is server-side browser-sniffing a bad idea?

deke web at master.gen.in.us
Fri Dec 7 06:25:25 CST 2001


On 6 Dec 2001, at 17:37, Craig Saila posted a message which said:

> Despite the fact it cancels out the benefits of caching (i.e., faster
> download times for the user), wouldn't putting:
>    <meta http-equiv="Pragma" content="no-cache">
> in the head solve these server-side issues?

That's a misuse of the "pragma" header - it's supposed to be for 
sending messages *to* the server, not from the server to the client.

And the biggest single chunk of the internet is AOL, and their 
caching firewall does *not* look at web pages to see if you have meta 
tags. You have to have *real* headers to play in their ballpark.

Most corporate caching firewalls, as well as the MSNTV one, will not 
cache a page that sets cookies in the headers, a page that is the 
result of a POST operation, or pages that have an https:// URL.  I 
can't think of a better reason to avoid javascript cookies than that.

If you use the 
     Cache-Control:private
header, that tells the browser it's OK to cache the page, but firewalls 
aren't supposed to save it to a shared cache.  That depends on 
everybody knowing and understanding HTTP/1.1, though.

deke




--------
We are the parents our people warned us about....





More information about the thelist mailing list