Closed Bug 192178 Opened 23 years ago Closed 23 years ago

an expires header with a date too far in the future is not handled correctly

Categories

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

defect

Tracking

()

RESOLVED WORKSFORME
mozilla1.4alpha

People

(Reporter: darin.moz, Assigned: skasinathan)

References

()

Details

(Keywords: topembed+)

an expires header with a date too far in the future is not handled correctly. mozilla fails to parse this expires header correctly: Expires: Wed, 17 Jul 2050 20:29:17 GMT the expiration time is computed to be zero causing a fresh GET (or at least a fresh conditional GET) each time the page is viewed. IE and 4.x handle this case fine, and I suspect it is not uncommon for websites to use such a date to force a cache entry to remain in the client's cache for as long as possible.
Status: NEW → ASSIGNED
Priority: -- → P3
Target Milestone: --- → mozilla1.4alpha
Keywords: topembedtopembed+
-> suresh
Assignee: darin → suresh
Status: ASSIGNED → NEW
Status: NEW → ASSIGNED
>>the expiration time is computed to be zero I think the expiration time is set correctly in this testcase. nsHttpChannel::UpdateExpirationTime() seems to set it fine. And about:cache shows the right expiry date. Key: http://unagi/bugs/bug_expires/nph-test.cgi Data size: 199760 Bytes Fetch count: 6 Last Modified: 02/25/03 10:47:32 Expires: 07/17/50 12:29:02 When I try to load the page next time, the call rv = mCacheEntry->GetMetaDataElement("request-method", getter_Copies(buf)); fails :( and hence (I think) we never check the expiration time from cache.
interesting
this testcase started to work fine for me from yesterday evening. We always load the content from cache. (tried both in new profile and in profile that i duplicated yesterday).
ok, i changed the Expires headers to several different future/past dates. Everything works fine as expected. For future expiry dates, the image is loaded from cache for the second and subsequent load. Darin, when you get a chance can you try this in your system? Thanks!!!
hmm.. WFM too [linux 2003030308 trunk]. i'm not sure what fixed this.
Status: ASSIGNED → RESOLVED
Closed: 23 years ago
Resolution: --- → WORKSFORME
Darin, your metadata changes are only on the minimo branch, right? What should the 'request-method' metadata have been? What was the error? Anything interesting?
gordon: actually my metadata changes haven't landed anywhere yet. haven't had a chance to polish off the patch after your review :(
You need to log in before you can comment on or make changes to this bug.