Re: upper limit on retain count

FROM : Axel Andersson
DATE : Tue Apr 19 18:44:14 2005

On Apr 18, 2005, at 22:43, Ivan S. Kourtev wrote:

> (1) is there a documented limit to the retain count?
> (2) is there a documented behavior once the retain count max is
> reached?


In the interest of nothing very much at all, I investigated what would
happen when a retain count overflows:

2005-04-19 18:31:33.610 retain[10574] after another one: 4294967292
2005-04-19 18:31:33.610 retain[10574] after another one: 4294967293
2005-04-19 18:31:33.610 retain[10574] after another one: 4294967294
2005-04-19 18:31:33.610 retain[10574] after another one: 4294967295
2005-04-19 18:31:33.610 retain[10574] after another one: 0
2005-04-19 18:31:33.610 retain[10574] after another one: 0
2005-04-19 18:31:33.610 retain[10574] after another one: 0
2005-04-19 18:31:33.611 retain[10574] after another one: 0
2005-04-19 18:31:33.611 retain[10574] after 1 release: 0

So apparently the overflow is controlled. Furthermore, I seem to have
created an immortal object as -release has no effect on it. It took my
dual 2.5 GHz G5 11 minutes to reach this conclusion (sending the
messages, I only actually logged the last 8 retains), so I tend to
agree with Joar that this is not a great idea. Maybe if you explained
in more detail what you're trying to accomplish, we could give you
better answers.

Cheers,

--
Axel Andersson
<email_removed>
http://www.zankasoftware.com/