On Sep 7, 2011, at 10:06 , Sean McBride wrote:
> Not to speak for the OP, but I am in the same situation as him.
If the "situation" you mean is what you describe in the rest of your post,
you're not in the same situation. The OP's very first sentence was:
> On Sep 6, 2011, at 03:33 , Oleg Krupnov wrote:
>
>> I'm implementing a new Lion's API, namely the resume.
That is, the OP was opting in, not opting out.
> It's not a case of "trying to take advantage of 10.7 features", it's a case
> of 10.7's "Resume" feature breaking *binary* compatibility of apps built with
> 10.6 deployment & 10.6 SDK.
>
> My app, and others I've used, and probably the OP's, were broken by Resume
> (basically because it's opt-out, not opt-in).
Well, if that's so, consider what you're saying: Lion *broke binary
compatibility with 10.6 applications*. If true, that's not at all what this
thread is about. It's also a *much* more serious problem. (Technically serious,
I mean. The practical effect is a different question.)
> To "fix" this fast for a dot release (that's built on 10.6-based buildbots)
> the easiest thing to do is to opt out of Resume, and to do that one must call
> setRestorationClass:, setRestorable:, etc.
>
> If Apple had provided the ability to opt out using an Info.plist key, as was
> suggested by many during the seed phase <rdar://9589219>, we would not be
> having this discussion.
>
> Another point: upping your min SDK requirement means that all your coworkers,
> build machines, etc. must also update. If you deploy to 10.6, it means you
> sometimes need to debug on 10.6, and doing that is a PITA if you can't tweak
> code and rebuild. If you require 10.7 SDK, how do you do that?!
>
> My solution is as others have said, namely:
>
> // If the SDK being compiled against is older than the 10.7 SDK...
> #if MAC_OS_X_VERSION_MAX_ALLOWED < 1070
>
> @interface NSWindow (RRRestorationHackForOldSDKs)
> - (void)setRestorationClass:(Class)restorationClass;
> - (void)setRestorable:(BOOL)flag ;
> - (void)invalidateRestorableState;
> @end
>
> #endif
>
> if ([window respondsToSelector:@selector(setRestorationClass:)]
> &&
> [window respondsToSelector:@selector(setRestorable:)] &&
> [window
> respondsToSelector:@selector(invalidateRestorableState)])
> {
> [window setRestorationClass:Nil];
> [window setRestorable:NO];
> [window invalidateRestorableState];
> }
Again, the point of this thread is that the above general technique is not the
correct way to opt *in* to a Lion feature. However, if you're trying to opt
*out* in order to work around the claimed Lion bug, then I don't see how there
can be any pragmatic objection to your taking this approach.
_______________________________________________
Cocoa-dev mailing list ([email protected])
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]
