I have a timer that counts down every second. It works great until the user opens up 3 or 4 tabs of the my site, at which point the newest tab's timer goes double or triple speed. I can currently only reproduce the bug in IE8. I was previously using setInterval, and could reproduce the bug in Firefox as well.
I'm actually using FBJS (Facebook's Javascript), so I'll just give some pseudocode.
function countDown() {
...
setTimeout(function() { countDown() }, 1000);
}
countDown();
However, what I'm really looking for is more theoretical. I know browsers can try and play "catch up" with setInterval, but how can multiple tabs cause this behaviour for setTimeout?
Browsers including Internet Explorer, Chrome, Safari, and Firefox store the delay as a 32-bit signed integer internally. This causes an integer overflow when using delays larger than 2,147,483,647 ms (about 24.8 days), resulting in the timeout being executed immediately.
Set timeout would not cause stack overflow, because it is asynchronous. It will just put the callback to the event queue and not block the execution.
To cancel a setTimeout() method from running, you need to use the clearTimeout() method, passing the ID value returned when you call the setTimeout() method.
No significant effect at all, setTimeout runs in an event loop, it doesn't block or harm execution.
That whole situation is very odd. The only scenario that's coming to mind where it makes any sense is one where the browser is trying to "smooth" the period of self-reinstating setTimeouts, same as for setInterval, and the code that's doing that actually confuses timers in different windows with each other.
I don't know if it's feasible, especially with Facebook involved, but an interesting test would be to give each instance a randomized name for the countDown
function and see if that makes any difference, like:
<?php $timerTag = rand(1, 1000); ?>
function countDown<?php echo $timerTag ?>() {
...
setTimeout(function() { countDown<? php echo $timerTag ?>() }, 1000);
}
countDown<?php echo $timerTag ?>();
If this changes the observed behavior, that argues for the scenario I have in mind. (And possibly provides a way of addressing the problem.)
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With