flat assembler
Message board for the users of flat assembler.
![]() Goto page Previous 1, 2, 3 Next |
Author |
|
pal
Yeah. The _sid and _data cookies are for encrypted log in information and "remember me" type things.
Not sure what the exact purpose of _t is but mine is now: 644 (143 bytes bigger also). Correlation? In 20 minutes it will have been 24 since the 501 I posted before? |
|||
![]() |
|
revolution
Another 24 hours and another 113 bytes.
_t cookie: 1849 bytes. And after reading all the messages: _t cookie: 1936 bytes. +87 bytes just to mark 8 messages as read. |
|||
![]() |
|
revolution
Now "_t"=1993 bytes (the other cookies don't change)
I have confirmed that there is no proxy in my connection so I will be able to test this all the way to the server without any disturbance. I had some connection problems today - the occasional blank response page from the forum - but they seem to have cleared now. Normally this is where I would kill all the cookies and start over, but for the sake of testing I will persevere for a while longer. |
|||
![]() |
|
LocoDelAssembly
Quote:
How do you confirmed that? |
|||
![]() |
|
revolution
I started using the https connection to powweb.
|
|||
![]() |
|
LocoDelAssembly
hehe, good plan!
|
|||
![]() |
|
revolution
My test had to stop about 12 hours ago when Firefox decided not to run anymore. I am not sure if it was a cookie problem or not that caused it. FF just closed without any error notification and all my cookies were gone when I started it up again. So I guess the monitoring starts again. If FF crashes again after 1993 byte cookies then perhaps there is an issue with it? Not sure, time will tell but the cookie spec allows for up to 4k.
|
|||
![]() |
|
LocoDelAssembly
Quote:
RFC 2109 - HTTP State Management Mechanism wrote: 6.3 Implementation Limits |
|||
![]() |
|
LocoDelAssembly
OK, I have installed the following php on my Apache:
Code: <?php if (!isset($_COOKIE['testCookie'])) { $cookie = 'A'; $browserCookieSize = "Cookie not present"; } else { $cookie = $_COOKIE['testCookie']; $browserCookieSize = strlen($cookie); } if (isset($_GET['increase'])){ // No, I'm not validating the data type of 'increase', this is just a test you know, not a production software... for ($i = 0; $i < $_GET['increase']; $i++) $cookie = $cookie . 'A'; } setcookie('testCookie', $cookie, time()+60*60); ?> <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> <html xmlns="http://www.w3.org/1999/xhtml"> <head> <title>Cookie Maximum Length Test</title> <script type="text/javascript"> function load(){ var div = document.getElementById("lengthDiv"); var start = -1, end = -1; if ((start = document.cookie.indexOf("testCookie=")) != -1){ if ((end = document.cookie.indexOf(";", start)) == -1) end = document.cookie.length; div.innerHTML = end - start - "testCookie=".length + " (not including 'testCookie=')"; } else div.innerHTML = "Cookie not found!"; } </script> </head> <body onload="load()"> <div id="lengthDiv"></div> <br /> <div>Cookie size recieved by server: <?php echo $browserCookieSize ?></div> <div>Cookie size sent by server (may be wrong): <?php echo strlen($cookie); ?></div> </body> </html> Results were: IE 7 wrote: Cookie not found! Firefox 3.0.10 wrote: 4086 (not including 'testCookie=') Opera 9.64 wrote: 4046 (not including 'testCookie=') Note that although the server side measure of the cookie size sent may be wrong, I have confirmed by sniffing that at least +8K bytes can be sent. Looks like all the browsers when receive a cookie bigger than their maximum supported size they just ignore it and use the previous value of it instead (so using a increase of 6000 the first time will make all browser not set the cookie). There are some more experiments to do but I have no more time for this now. PS: All the results correspond to /cookie-test.php?increase=1 after being repeated several times with no perceivable increment. |
|||
![]() |
|
Madis731
|
|||
![]() |
|
revolution
My _t cookie is now 2376 bytes. No noticeable access problems right now, working fine with https.
|
|||
![]() |
|
revolution
Okay, I have brought my _t cookie to 3667 bytes. Let's see how it goes from here. Only ~400 bytes till the FF max is achieved.
|
|||
![]() |
|
Mac2004
pal wrote: I have just updated to 3.0.10 now (since you said it was out - my FF doesn't download automatically). At the time I was using 3.0.8. I check FF updates manually due to the same fact. FF automatic update doesn't work all the time. Annoyance caused by this is not very big though. |
|||
![]() |
|
revolution
_t = 3897.
What will happen at ~4096? |
|||
![]() |
|
Borsuc
stop before you'll create a black hole!
|
|||
![]() |
|
revolution
Borsuc wrote: stop before you'll create a black hole! 4011 bytes. |
|||
![]() |
|
revolution
My _t cookie is now stuck at 4068 bytes. Nothing is updating, all the unread threads remain unread.
[edit] It is not all the threads remain unread, it is new threads that remain unread. Threads that I have previously viewed can update the unread flag for new messages. |
|||
![]() |
|
LocoDelAssembly
With both, HTTP and HTTPS?
|
|||
![]() |
|
revolution
My results are with HTTPS.
If I try HTTP then I get a blank page and cannot use the board at all with my giant cookie. |
|||
![]() |
|
Goto page Previous 1, 2, 3 Next < Last Thread | Next Thread > |
Forum Rules:
|
Copyright © 1999-2020, Tomasz Grysztar. Also on GitHub, YouTube, Twitter.
Website powered by rwasa.