My replies aren't showing up

Author	Post
Mister Fahrenheit All-Star Member Meh 4655 Posts 10/13	Posted - Feb 8 2016 : 12:56PM I've made three replies this morning, all in the Polls section. They were in the, "Last Non-Porn Movie You Watched," "Last Thing You Drank," and "Last Thing You Ate" threads. I typed out my post, submitted it, got the confirmation page that my reply was submitted, and then it's not there at all on the thread. Maybe I'm just not seeing it on my end? [Link]
Flash Golden Age Classic 11018 Posts 5/01	Posted - Feb 8 2016 : 1:21PM I can see your replies in those threads [Link]
Flash Golden Age Classic 11018 Posts 5/01	Posted - Feb 8 2016 : 1:28PM However I can see what you are talking about, my replies in both Firefox and Chrome aren't showing up immediately in that browser but show immediately through a different browser. In other words, I can see the replies I just did in Firefox in Chrome but not Firefox and the ones I just did in Chrome in Firefox but not Chrome. [Link]
Flash Golden Age Classic 11018 Posts 5/01	Posted - Feb 8 2016 : 1:35PM It has to have something to do with the caching. Going to IE which I wasn't logged in to and looking at this thread then logging in and returning showed me as not logged in again and unable to post. But when I clicked on the login button it stated I was logged in. [Link]
Jacco All-Star Member "liable to deprave and corrupt" 2673 Posts 5/04	Posted - Feb 8 2016 : 1:48PM ^Clearly some caching issue, I had a reply not show up too, but it showed when I refreshed the thread page. Jacco [Link]

Drew Black broken crankshaft 8009 Posts 9/99	Posted - Feb 8 2016 : 3:56PM Should be better now [Link]
Mister Fahrenheit All-Star Member Meh 4655 Posts 10/13	Posted - Feb 8 2016 : 3:56PM ^ Seems like it. Thanks Drew! [Link]
Drew Black broken crankshaft 8009 Posts 9/99	Posted - Feb 8 2016 : 4:41PM I've been doing a lot of work behind the scenes to make better use of browser and intermediary caches in the forum. I wasn't properly setting the Last-Modified header in the server response so your browser wasn't able to correctly ask if a new version of a page was available. [Link]
Drew Black broken crankshaft 8009 Posts 9/99	Posted - Feb 8 2016 : 7:03PM reply [Link]
Jacco All-Star Member "liable to deprave and corrupt" 2673 Posts 5/04	Posted - Feb 8 2016 : 7:58PM Another reply Nope, still had to refresh the page for it to show up. Jacco Edited by - jacco on 2/8/2016 7:58:52 PM [Link]
Mister Fahrenheit All-Star Member Meh 4655 Posts 10/13	Posted - Feb 8 2016 : 8:09PM I'm using Chrome and my replies have all been showing up since Drew's first fix. [Link]
Drew Black broken crankshaft 8009 Posts 9/99	Posted - Feb 8 2016 : 8:14PM Testing [Link]
Goldstein All-Star Member "You have sacrificed nothing and no one." 3689 Posts 8/10	Posted - Feb 8 2016 : 8:22PM I just posted, and my reply didn't show up like it normally would. So, I refreshed the page myself, and it appeared. Firefox v44.0 [Link]
Drew Black broken crankshaft 8009 Posts 9/99	Posted - Feb 8 2016 : 8:59PM Thanks. I'll do a little more testing. Seems like FF doesn't like the last modified header or something [Link]
Drew Black broken crankshaft 8009 Posts 9/99	Posted - Feb 8 2016 : 9:10PM reply with FF 44 to test header trace [Link]

Drew Black broken crankshaft 8009 Posts 9/99	Posted - Feb 8 2016 : 9:26PM Firefox 44 is ignoring the must-revalidate directive in the Cache-Control header. I'll see if I can work around it. It does the validation with a refresh but works out of its own local cache otherwise. [Link]
Flash Golden Age Classic 11018 Posts 5/01	Posted - Feb 8 2016 : 9:54PM Drew, I don't envy you one bit. [Link]
Drew Black broken crankshaft 8009 Posts 9/99	Posted - Feb 8 2016 : 9:55PM Flash wrote: Drew, I don't envy you one bit. Click to expand just trying to speed things up a bit by poking a stick in the spokes testing to see if an edit appears too in FF44 Edited by - Drew Black on 2/8/2016 9:55:43 PM [Link]
Flash Golden Age Classic 11018 Posts 5/01	Posted - Feb 8 2016 : 10:02PM It appears to me to be fixed here using Firefox 44.0.1 [Link]
Drew Black broken crankshaft 8009 Posts 9/99	Posted - Feb 8 2016 : 10:04PM Yeah, I think I got it right this time. And I learned something in the process that will be very helpful with some other work I'll be doing on the site soon. FireFox caches everything (even items explicitly labeled as do-not-cache) and ignores some cache-control directives. [Link]
Flash Golden Age Classic 11018 Posts 5/01	Posted - Feb 9 2016 : 6:10PM It is occurring intermittently (two of the last three posts I made) again. Firefox 44.0.1. [Link]
Drew Black broken crankshaft 8009 Posts 9/99	Posted - Feb 9 2016 : 6:15PM Strange. I'll watching things a bit on my end. [Link]
Flash Golden Age Classic 11018 Posts 5/01	Posted - Feb 9 2016 : 6:24PM I can see the replies doing a forced refresh but for some reason the browser shows what it saw prior to posting when coming back. When I came here to see what you wrote, both yours and mine was missing but doing a forced refresh showed them. Edited to add, but it is intermittent because this reply showed up like normal. Edited by - flash on 2/9/2016 6:25:10 PM [Link]
Drew Black broken crankshaft 8009 Posts 9/99	Posted - Feb 9 2016 : 10:16PM I've been doing HTTP traces with Firefox. It does some very strange things and appears to ignore the "must-revalidate" HTTP cache control header. So it's saving a copy of the page it stored locally and it's supposed to ask the server every time it loads a page if the copy it has is the most recent. We provide 2 pieces of data that both can be used to establish the validity of a cached page. But FF isn't revalidating and instead using it's own local copy. I haven't figured out why. Up until a few days ago our server generated and sent over the wire a fresh copy of the entire page for every request. Stupid, I know. That's what caching is for. If a page hasn't been modified then don't send the data over the wire. It speeds things up by orders of magnitude. I have been reading about the FF cache handling and they flat-out state that they will cache things locally that are even marked non-cacheable. So it's going to take a little more research. [Link]
Drew Black broken crankshaft 8009 Posts 9/99	Posted - Feb 9 2016 : 10:19PM I will change the way that the post page redirects back to the topic page to try to force a refresh there. It uses a very antiquated way of redirecting a page that may no longer be supported and certainly doesn't force a refresh of the cached content. I'll work on it on Thurs. [Link]

Goldstein All-Star Member "You have sacrificed nothing and no one." 3689 Posts 8/10	Posted - Feb 10 2016 : 12:20AM FWIW... Regarding Firefox 44.0.1, I had difficulty updating it on a second machine tonight (from 44.0.0 to 44.0.1). Specifically, it would not complete the update to 44.0.1 when you do the mandatory "restart" of Firefox to complete the update. I tried three times. So, after re-booting the entire computer, I did the following: (1) Exported a backup of my bookmarks. (2) Did a complete uninstall of Firefox, using Revo Uninstaller, so that it would get rid of all traces. (3) Did a fresh install of Firefox (using the 44.0.0 stub, previously downloaded), which installed a complete current version of 44.0.1. Program works perfectly. Important: Time to do another Flash Player update to v 20.0.0.306 So, the reason for this post is to make everyone aware that Firefox may be having some issues of its own, at the moment. That you need to get updated to v 44.0.1 And, that you may want to do a complete fresh install to 44.0.1 as I discussed above. [Link]
Goldstein All-Star Member "You have sacrificed nothing and no one." 3689 Posts 8/10	Posted - Feb 10 2016 : 12:21AM ^ And, the above post appeared immediately, with the automatic page refresh after posting, just like the good old days [Link]
Drew Black broken crankshaft 8009 Posts 9/99	Posted - Feb 10 2016 : 8:44AM Thank you for the debugging efforts. It certainly helps me diagnose this situation [Link]
Drew Black broken crankshaft 8009 Posts 9/99	Posted - Feb 11 2016 : 9:47AM i changed the way the page redirects back to the topic upon posting. Seems like this may force FF to at least ask the server if there's a fresh copy. testing an edit in an existing post. Edited by - Drew Black on 2/11/2016 9:54:23 AM [Link]
Harri Patel Senior Member #SDNY he's all yours. -- Michael Steele 5110 Posts 3/03	Posted - Feb 11 2016 : 10:15PM I'm on Firefox 45.0b4 and my replies are not showing up unless i manually refresh the page. [Link]
Drew Black broken crankshaft 8009 Posts 9/99	Posted - Feb 12 2016 : 7:22AM I am going to remove the last modified HTTP header and see if that helps. [Link]
Drew Black broken crankshaft 8009 Posts 9/99	Posted - Feb 12 2016 : 8:11AM I dropped the last-modified header. I think this will help. There must be a bug in the way Firefox handles its local cache. Here's their documentation on the issue: What other factors influence revalidation? Revalidation is triggered when the user presses the reload button. It is also triggered under normal browsing if the cached response includes the "Cache-control: must-revalidate" header. Another factor is the cache validation preferences in the Advanced->Cache preferences panel. There is an option to force a validation each time a document is loaded. Click to expand The code here provides that header but FF is ignoring it. Must-revalidate is a server instruction that tells the browser to check back with the server every time before using page content from its local cache. It appears that FF is placing a higher priority on their internally-computed expiration time. How are expiration times calculated (since not every response includes an Expires header)? Mozilla strives to implement RFC 2616 (see in particular section 13). The following response headers generate an expiration time that is always in the past (given the value "00:00:00 UTC, January 1, 1970"): Cache-control: no-cache Cache-control: no-store Pragma: no-cache Expires: 0 It is interesting to note that "Expires: 0" and "Pragma: no-cache" are technically invalid response headers. If none of these headers are present, then the expiration calculation is essentially the algorithm described in RFC 2616 section 13.2. We estimate the current age of the response and the freshness lifetime based on available information. The current age is usually close to zero, but is influenced by the presence of an Age header, which proxy caches may add to indicate the length of time a document has been sitting in its cache. The precise algorithm, which attempts avoid error resulting from clock skew, is described in RFC 2616 section 13.2.3. The freshness lifetime is calculated based on several headers. If a "Cache-control: max-age=N" header is specified, then the freshness lifetime is equal to N. If this header is not present, which is very often the case, then we look for an"Expires" header. If an "Expires" header exists, then its value minus the value of the "Date" header determines the freshness lifetime. Finally, if neither header is present, then we look for a "Last-Modified" header. If this header is present, then the cache's freshness lifetime is equal to the value of the "Date" header minus the value of the "Last-modified" header divided by 10. This is the simplified heuristic algorithm suggested in RFC 2616 section 13.2.4. The expiration time is computed as follows: expirationTime = responseTime + freshnessLifetime - currentAge where responseTime is the time at which the response was received according to the browser. Click to expand If Firefox was actually doing a revalidation request it wouldn't be falling back on this calculated expiration. Since I'm no longer providing a Last-modified header it will have nothing to calculate, which appears to trigger the desired revalidation. [Link]
Goldstein All-Star Member "You have sacrificed nothing and no one." 3689 Posts 8/10	Posted - Feb 12 2016 : 9:12AM Now Firefox update is 44.0.2 They're obviously tweaking "something" Updated my first machine to 44.0.2 without a problem. [Link]
Flash Golden Age Classic 11018 Posts 5/01	Posted - Feb 12 2016 : 11:06AM Both of my machines did 44.0.1 and 44.0.2 without issue. .2 says it fixes "same-origin-policy violation using Service Workers with plugins." [Link]
Drew Black broken crankshaft 8009 Posts 9/99	Posted - Feb 12 2016 : 3:47PM This problem should be dealt with for good at this point. I've put some workarounds in place to support Firefox's unique caching scheme. [Link]

Mister Fahrenheit All-Star Member Meh 4655 Posts 10/13	Posted - Feb 14 2016 : 1:14AM Sent you a PM Drew. [Link]
Porn Scout Deactivated User 3735 Posts 10/02	Posted - Feb 18 2016 : 10:28PM I'm seeing this issue. I've seen it twice today. My last reply was to the Jules Jordan Upcoming Releases thread and I'm not seeing it there. Maybe it'll show up in a few minutes. [Link]
Porn Scout Deactivated User 3735 Posts 10/02	Posted - Feb 18 2016 : 10:30PM It just happened on this thread as well. The comment seems to show up after a refresh though. [Link]