In Apple's delightful documentation of Chris Kane's Receptionist pattern (which
I think of as a trampoline), the key method looks like this:
- (void)observeValueForKeyPath:(NSString *)keyPath ofObject:(id)object
change:(NSDictionary *)change context:(void *)context {
[queue addOperati
This might be another way of phrasing the same question: why does the new iOS 5
method +[NSURLConnection sendAsynchronousRequest:queue:completionHandler:] want
an NSOperationQueue as its queue rather than a dispatch queue? m.
--
matt neuburg, phd = m...@tidbits.com, http://www.apeth.net/matt/
pa
On Feb 15, 2012, at 11:19 AM, Matt Neuburg wrote:
> My question is simply this: in a world that has GCD, is there any reason why
> we wouldn't have "queue" be a GCD dispatch queue (probably the main queue)
> and call dispatch_asynch?
I don't believe there's any difference in this example.
>
On Wed, 15 Feb 2012 11:19:49 -0800, Matt Neuburg said:
>My question is simply this: in a world that has GCD, is there any reason
>why we wouldn't have "queue" be a GCD dispatch queue (probably the main
>queue) and call dispatch_asynch? Putting it even more generally (and
>this is probably my *real
On Feb 15, 2012, at 12:53 PM, Sean McBride wrote:
> Apple's always saying to use the highest level API available
We must be watching different versions of the WWDC videos... :)))
Backwards compatibility is a good reason on the desktop, but I'm on iOS and on
iOS if you've got blocks you've got
Because it's a higher-level API answers pretty much all your questions. You
have a little more easy flexibility with what you can do to an NSOperationQueue
compared to raw GCD.
To put it another way, what downside are you seeing to using NSOperationQueue?
On 15 Feb 2012, at 19:32, Matt Neuburg
On Wed, Feb 15, 2012 at 11:19 AM, Matt Neuburg wrote:
> My question is simply this: in a world that has GCD, is there any reason why
> we wouldn't have "queue" be a GCD dispatch queue (probably the main queue)
> and call dispatch_asynch? Putting it even more generally (and this is
> probably my
On Feb 15, 2012, at 12:32 PM, Matt Neuburg wrote:
> This might be another way of phrasing the same question: why does the new iOS
> 5 method +[NSURLConnection sendAsynchronousRequest:queue:completionHandler:]
> want an NSOperationQueue as its queue rather than a dispatch queue? m.
In OS X 10.6
On Wed, 15 Feb 2012 13:56:14 -0800, Kyle Sluder said:
>On Wed, Feb 15, 2012 at 11:19 AM, Matt Neuburg wrote:
>> My question is simply this: in a world that has GCD, is there any reason why
>> we wouldn't have "queue" be a GCD dispatch queue (probably the main queue)
>> and call dispatch_asynch?
> On iOS, NSOperationQueue doesn't use GCD at all.
Actually, posts by Apple people in the Developer Forums and the
documentation differs on this:
http://developer.apple.com/library/ios/#documentation/Cocoa/Reference/NSOper
ationQueue_class/Reference/Reference.html
"Note: In iOS 4 and later, oper
On Feb 16, 2012, at 8:09 AM, Matt Neuburg wrote:
> On Wed, 15 Feb 2012 13:56:14 -0800, Kyle Sluder said:
>>
>> On iOS, NSOperationQueue doesn't use GCD at all. On Mac OS X 10.6 and
>> greater, block-based NSOperations are submitted to global
>> (asynchronous) dispatch queue.
>
> Very useful; t
On Feb 16, 2012, at 9:34 AM, Kyle Sluder wrote:
> I must correct myself: NSOperationQueue uses GCD on iOS 4 and higher
Good; I thought so, but you made me doubt it... :) m.
--
matt neuburg, phd = m...@tidbits.com, http://www.apeth.net/matt/
pantes anthropoi tou eidenai oregontai phusei
Programm
12 matches
Mail list logo