[erlang-questions] bug in timer module in R14B02?
Mon Jun 20 04:02:18 CEST 2011
I ran into a lot of problems with the timer module on R14B02 and eventually
switched to using erlang:send_after. This particular section in the erlang
doc also convinced me to switch over to erlang:send_after and maintain my
I was running on a 64-bit system (CentOS 5.5), so, as Dale mentioned, the
problem may be fixed.
On Sun, Jun 19, 2011 at 5:00 PM, Dale Harvey <dale@REDACTED> wrote:
> I believe there was an integer overflow bug on 64 bit systems that affect
> ordered ets tables (and therefore caused timers not to trigger) in R14B02
> That may be your problem?
> On 19 June 2011 03:43, Zsolt Keszthelyi <zsolt.erl@REDACTED> wrote:
>> Dear All,
>> I ran into something in R14B02.
>> It looks like if I start a timer with a long interval then all previously
>> started timers stop.
>> The timer server is still running and the timers are in the ETS table.
>> No error returned. Nothing crashes.
>> This was working on R13B03 and is not working on R14B02
>> The long interval is exactly 2^31/1000 ms.
>> Is it a bug or an intentional change?
>> Maybe the timer module was never meant to be used with long intervals but
>> shouldn't it give at least an error msg.?
>> %% works on R13B03
>> %% doesn't work on R14B02
>> timer:apply_interval(1000, io, format, ["Hello!\n"]),
>> timer:apply_interval(40*60***1000, io, format, ["No more hello.\n"]).
>> erlang-questions mailing list
> erlang-questions mailing list
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the erlang-questions