Re: bug 52062

On Sun, 25 Dec 2011, Stas Malyshev wrote: > I see that test for bug 52062 (marked as fixed by this commit: > > http://svn.php.net/viewvc/?view=revision&revision=320481) > now fails on my > 32-bit system. Looking at the patch and the test, it can not actually succeed, > as the test expects this: > int(100000000000) > which is not possible on 32-bit system, and the code actually compares the > result to LONG_MAX which is 32-bit and returns false when it's bigger that > that. So I'd like to know what was the intent there: > > 1. Return false on 32-bit and the test should be fixed to account for 32-bit > and 64-bit? > 2. Return float on 32-bit? > 3. Do something else? > > Please advise. There are several issues here: 1. The test was wrong, the result for var_dump($d->getTimestamp()); should indeed have been bool: false; I've a patch for this. 2. DateInterval::format() didn't handle large ints succesfully; I have a patch for this too. 3. $d->setTimestamp(100000000000) is not ever going to work as zend_parse_parameters simply gives me long: 1215752192. This I find really strange actually. Any clue? cheers, Derick -- http://derickrethans.nl | http://xdebug.org Like Xdebug? Consider a donation: http://xdebug.org/donate.php twitter: @derickr and @xdebug