52635 - Always wants to resubmit form if going back, even with cache-control

Status: VERIFIED DUPLICATE of bug 55055

(Core :: Networking: Cache, defect, P3)

Description

[Darwin O'Connor]

I am a programmer on a web application where the users go from page to page by
POSTing forms. The problem is, when I use Mozilla or Netscape 6 to use the
application and I press the back button it always want to rePOST the page
instead of using the cached version. I have tried to convince Mozilla to use the
cache by using the http headers "Expires: [Some time in the near future]" and
"Cache-Control: max-age=1200" and they didn't help.

I have set up a some pages show the problem. Go to
http://www.reamined.on.ca/doconnor/test.htm and you can do a POST to the
dynamically generated page which contains Cache-Control: max-age=120 in the HTTP
headers. You can then go to another page and the go Back and Mozilla will ask
you if you want to repost the form data. Which ever one you choose Mozilla will
still reload the page instead of using the cached version.

References:

http://www.w3.org/Protocols/rfc2616/rfc2616-sec14.html#sec14.9.3
   The first paragraph describes what the Cache-Control: max-age should do.

http://www.deja.com/=dnc/threadmsg_ct.xp?AN=645017742.1
   In answering a question similar to mine "Matthew `mpt' Thomas, Mozilla user
interface QA" says Mozilla will follow the caching information.

Comment 1

[R.K.Aa.]

*** Bug 52702 has been marked as a duplicate of this bug. ***

Comment 2

[Andreas M. "Clarence" Schneider]

Going back to a 2xx result of a HTTP POST _in session history_ should always
redisplay the page without reposting, even if it has a "Cache-Control: no-cache" 
header. See RFC 2616 and my latest comment in bug 43123.

But this bug differs from 43123: If a "Cache-Control: max-age=x", x > 0, header
or something similar is included, the page should make it into the normal cache
and be used on every subsequent request until it expires.

Comment 3

[Gervase Markham [:gerv]]

Confirming based on my assessment of the sanity of the reporter :-)

Gerv

Comment 4

[Scott Gifford (old account)]

Is this a dup of bug 55055?

Comment 5

[Darwin O'Connor]

Actually bug 55055 is a duplicate of this bug, but I will yield to it, as that 
bug is actually getting some attention.

One wonders how this problem as been around for so long, but never really got 
any attention until it was too late to properly fix it before Netscape 6 was 
released. I noticed it a year ago, but figured it was so obvious it would have 
been fixed without me having to figure out the whole process of posting a bug. 
Maybe it just shows how horribly wrong Gervase Markham's assessment of my sanity 
was. :-)

*** This bug has been marked as a duplicate of 55055 ***

Comment 6

[Stephen Walker]

vrfy dup