On Thu, 08 Dec 2011 20:31:51 +0100, Mikkel Islay
said:
>A twist on the tale, is that the observer wasn't the UIViewController, but a
>helper object subclassed from NSObject. Hence view-unloading was one step
>further removed from deallocation of the observer-instance.
I have *never* added or
Thanks for the insightful reply, and of course to Jonny Yu and Scott Anguish as
well.
On 8 Dec 2011, at 17:36, Matt Neuburg wrote:
>> You can certainly make assumptions about when dealloc is called - it is
>> extremely deterministic, actually. See Greg Parker's discussion (in the WWDC
>> 2011
On Wed, 07 Dec 2011 21:56:56 +0100, Mikkel Islay
said:
>Dear list,
>
>A question regarding the proper use of dealloc in ARC-environments (under iOS):
>I have a need to unregister NSNotification observers during exchange of
>UiViewControllers. Is implementing - (void)dealloc considered a proper u
My understanding is, due to auto release semantics, your object might be auto
released again during current pool draining. So it will be put into outer
Autorelease Pool, that's to say the dealloc are not guaranteed to be executed
at specific timing. ARC still relies on retain/release/autorelease
Dear list,
A question regarding the proper use of dealloc in ARC-environments (under iOS):
I have a need to unregister NSNotification observers during exchange of
UiViewControllers. Is implementing - (void)dealloc considered a proper use for
this purpose? Is it safe to make assumptions regarding
