On Dec 28, 2012, at 12:59 AM, Quincey Morris
wrote:
> On Dec 27, 2012, at 22:37, Rick Mann wrote:
>> So, the usual bit about references being weak by default in C structs (and
>> C++ classes) doesn't hold here.
>
> Huh? References aren't weak by default anywhere.
>
> Are you thinking of __usa
On Dec 28, 2012, at 0:59 , Quincey Morris
wrote:
> Huh? References aren't weak by default anywhere.
>
> Are you thinking of __usafe_unretained? That's not weak, just unmanaged by
> ARC, and it's the only ownership qualifier allowed by ARC for struct members
> that point to objects (and there
On Dec 27, 2012, at 22:37, Rick Mann wrote:
> So, the usual bit about references being weak by default in C structs (and
> C++ classes) doesn't hold here.
Huh? References aren't weak by default anywhere.
Are you thinking of __usafe_unretained? That's not weak, just unmanaged by ARC,
and it's
On Dec 27, 2012, at 22:50 , Ludovic Nicolle wrote:
> hey, is your project/target by any chance having NSZombieEnabled in its flags?
>
> there is/was a problem similar to this with zombies enabled so please make
> sure this aint the case.
My project was NOT NSZombieEnabled, but then it starte
On Dec 27, 2012, at 21:58 , Rick Aurbach wrote:
> Rick,
>
> I'm no Obj-C language expert; in fact I'm pretty much a newbie. But I have
> had some experience with smart pointers (boost, etc) from my C++ case.
>
> Here's how I think of strong and weak references in ARC. (If the following is
>
Rick,
I'm no Obj-C language expert; in fact I'm pretty much a newbie. But I have had
some experience with smart pointers (boost, etc) from my C++ case.
Here's how I think of strong and weak references in ARC. (If the following is
wrong, I'm sure someone more knowledgeable than myself will set
I just wrote some Objective-C++ code. It's basically just a regular Obj-C
class, but makes some C++-style calls in its methods.
Anyway, I added an NSMutableData* ivar to my class, and set it with
mInputBuffer = [NSMutableData data];
in one of the methods. Subsequent accesses, however, h