Michael:

We actually use the following formula to determine if the action of 
hitting the back button should query the server:

  hit_server_on_back_forward = (https && no-cache) || no-store

So, for normal HTTP content, you'd have to send "Cache-control: 
no-store" to bypass the cache when loading pages from session history 
(via back/forward buttons).  For HTTPS content, you can either specify 
"Cache-control: no-store" or "Cache-control: no-cache" or "Pragma: 
no-cache".  The distinction between HTTP and HTTPS content is made for 
compatibility reasons.  Idealy, we should only bypass the cache when 
presented with a "no-store" header, but sadly too many online banks (and 
other secure sites) depend on the "no-cache" headers triggering this 
behavior.

If you check out RFC2616 
(http://www.w3.org/Protocols/rfc2616/rfc2616.html), you'll see that 
mozilla is in compliance with the HTTP/1.1 spec.

Darin



Gagan Saksena wrote:

> I would imagine that a "Cache-Control: no-store, no-cache" header 
> should do the trick. If not you may have found a bug. Cc'ing Gordon 
> and Darin.
>
> -Gagan
>
> Michael Kinsley wrote:
>
>> Hello,
>>
>> I am sorry to bug, but I have searched through FAQ's and newsgroups 
>> to no
>> avail.
>>
>> Here's the problem:
>>
>> When using Mozilla 1.0 and a user hits the BACK button the page 
>> reloads from
>> cache no matter what!  I have set the HEADERS telling the browser to 
>> reload
>> the page every time.
>>
>> Even checking the status via about:cache claims that the page won't be
>> reloaded from cache - but it IS!!!   There is no problem in IE5.X, it
>> reloads the page whenever the back button is hit.
>>
>> How do I tell mozilla to reload a page even if the back button is hit?
>>
>> thank you
>>
>> Michael Kinsley
>>  
>>



Reply via email to