On Jul 19, 2008, at 10:52 AM, Uli Kusterer wrote:
On 19.07.2008, at 05:52, Andy Lee wrote:
It's only when the ownership metaphor was applied that I questioned
whether the abstraction holds up.
It's a metaphor, an abstraction. As those things go, they're bound
to break down at some point.
Certainly, of course. But I'm sure I'm not the only person here who
cares *where* a metaphor breaks down.
"ownership" works, because it's a term already taught in programming
classes for more low-level languages (and even in good for higher-
level garbage collected languages, because even there destruction
order can matter).
I was a bit red-faced when Michael Ash pointed out that the term is
commonly used to mean general responsibility for freeing memory. I'd
honestly thought that Apple had made it up. Serves me right for not
paying attention in class.
It may feel a bit odd, but thats because it works by analogy.
It only feels odd to me because the metaphor requires more stretching
than in the case of, for example, the C++ String example. This is
related to my discomfort with calling a retain count a reference count.
"autoreleased" in the context where an object is not added to a
pool, however, leads to stale pointers and memory bugs, and may even
lead to memory corruption and/or crashes. So I think I'll take an
analogy hobbling in on its last leg over a label that leads to wrong
code any day.
Yes, I've already admitted my suggestion was crappy. For the record,
again, I've never used the term that way.
I rarely answer memory management questions, but I may answer some in
the future just to get used to saying "you own it" or "you don't own
it."
--Andy
_______________________________________________
Cocoa-dev mailing list (Cocoa-dev@lists.apple.com)
Please do not post admin requests or moderator comments to the list.
Contact the moderators at cocoa-dev-admins(at)lists.apple.com
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com
This email sent to [EMAIL PROTECTED]
