On 01/02/2013, at 4:08 PM, Kyle Sluder <k...@ksluder.com> wrote: > But CFTypes *are* NSObjects. So I'm not sure why your case isn't working. Can > you be more specific?
This is why the result surprised me - I have "known" this for years, and assumed that there wouldn't be a problem with this. But, this is the first time I've ever actually tried it, and what I found is that if I have a property, e,g @property (nonatomic, assign) CGColorRef myColor; and I ask for this using [someObject valueForKey:@"myColor"]; then instead of returning the object, it hits -valueForUndefinedKey: instead. While that gives me an opportunity to do something to wrap or convert the object, I must admit it wasn't the expected behaviour. I have been through all the KVC documentation and I could find nothing that stated explicitly whether CFTypes can be accessed via KVC; the lack of mention could imply that they can't be, or alternatively, that they can be... ;-) Incidentally, while I declare these properties as 'assign', they do in fact retain, it's just that the property synthesizing mechanism doesn't allow me to specify retain for these types. Instead I have to implement the setter manually and perform the necessary retains/releases as needed. Perhaps this is a clue? Again, I would have thought that the system would 'know' that a CFType is an NSObject, and permit the use of retain in the property declaration. --Graham _______________________________________________ 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: https://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com This email sent to arch...@mail-archive.com
