Did sleep() *EVER* take a non-integer argument?

Did sleep() EVER take a non-integer argument?

dalecosp

What it says. I found this today:

//did not get the lock, sleep for a little then try again.
sleep(.1);

NOT my code, but some the company purchased like, years ago and is still in use. And it sucks. This is from a caching routine, inside a finite loop ($x > 15 ), and since the argument would evaluate to "0/no sleep at all", it would explain why a lot of cache files NEVER get written or used.

I guess they never heard of [man]usleep/man ... but that's been around since PHP4 16 years ago!!!

Grr...

NogDog

Don't think it ever did -- at least I vaguely remember discovering usleep() at some point for something I was trying to do (and wondering now why the heck I'd need to sleep for less than a second -- or even sleep at all :rolleyes: ).

Bonesnap

sleep() can be very unpredictable. Seems like most times I try to use it an unhandled CatWantsHisFoodGoddammitException is thrown and I am forced to stop.

dalecosp

NogDog;11054027 wrote:
Don't think it ever did -- at least I vaguely remember discovering usleep() at some point for something I was trying to do (and wondering now why the heck I'd need to sleep for less than a second -- or even sleep at all :rolleyes: ).

I just find it to be a little strange that you can pay several hundred dollars to a company for software written by people that don't appear to know that sleep only takes an integer argument. Strange little bug; I'm not really sure how we can determine whether or not we were actually affected by it, but in theory, it could cause their caching system to be quite a bit less efficient than they would lead you to believe if you ask them ...

I gave them some notification about their bug, just to see if maybe it was still there 7 years ex post facto. I'm not holding my breath to hear from them ;-)

Bonesnap wrote:
sleep() can be very unpredictable. Seems like most times I try to use it an unhandled CatWantsHisFoodGoddammitException is thrown and I am forced to stop.

I really don't like cats too much, and that Exception, along with CatOnTheTableTearingIntoBreadBag, CatPukedInLivingRoom, and CatProtestsLitterBoxByTakingDumpNearby are the chief reasons why....

Of course, I wonder if I'd be a better hacker if I were a cat person? 😉

laserlight

It could well have been a typo (one character difference, after all), though it should have been caught during review or testing.

dalecosp

laserlight;11054057 wrote:
It could well have been a typo (one character difference, after all), though it should have been caught during review or testing.

Nice thought, though not likely; there are 3 identical occurrences of the call. I suppose maybe a cut/paste thing, but to cut/paste a typo in 2 additional locations seems almost as bad as creating the bug in the first place.... :eek:

dalecosp

NogDog;11054027 wrote:
(and wondering now why the heck I'd need to sleep for less than a second -- or even sleep at all :rolleyes: ).

Typically for some (sense of, at least) protection against resource contention. I use sleep(), for example, in an "hourly" cron routine to make sure we don't jump in on another instance of ourself, or the "daily" routine which is more important:

//check for the lockfile, sleep if exists
$lockfile = "/tmp/cron.lock";

// if we're already running (since an hour ago!!!) don't run again
if (trim(@file_get_contents($lockfile)) == "hourly" )
{
  exit;
}

while (@file_exists($lockfile))
{
   sleep(60);
}

//create our instance of the lockfile
file_put_contents($lockfile, "hourly");

NogDog

Yeah, I just always feel like in that situation there should be something better to do -- not that there is, it just feels wrong. (Maybe an event-listener like some languages support would be "cleaner"? { Decided to google a bit, look what I found: http://php.net/manual/en/class.eventlistener.php } )

Bonesnap

dalecosp;11054055 wrote:
I really don't like cats too much, and that Exception, along with CatOnTheTableTearingIntoBreadBag, CatPukedInLivingRoom, and CatProtestsLitterBoxByTakingDumpNearby are the chief reasons why....

These exceptions are not necessarily exclusive to Cats, and CatProtestsLitterBoxByTakingDumpNearby is substituted with DogGoesOutsideAndSpinsAroundForTenMinutesAndMightTakeADumpIfYouAreLucky or worse: DogGoesOutsideAndDoesntDoAnythingButComesInsideAndDumpsImmediately. Winter is the worst for that one. There is also DogBarksBecauseSomethingMovedOrMadeANoiseOrJustBecause, as well.

NogDog

Bonesnap;11054095 wrote:
These exceptions are not necessarily exclusive to Cats, and CatProtestsLitterBoxByTakingDumpNearby is substituted with DogGoesOutsideAndSpinsAroundForTenMinutesAndMightTakeADumpIfYouAreLucky or worse: DogGoesOutsideAndDoesntDoAnythingButComesInsideAndDumpsImmediately. Winter is the worst for that one. There is also DogBarksBecauseSomethingMovedOrMadeANoiseOrJustBecause, as well.

If it was raining out, I couldn't depend on Noggin to "do his business" unless I went out with him. Otherwise he might just huddle miserably under a bush or something until I let him back in. But, if I was outside with him, he'd often go about his business and mostly ignore the rain, unless it was a deluge. (shrug)