date:20081220



On 20 Dec 2008, at 8:05 pm, Andrew Farmer wrote:

However, I think you've overengineering things here. The standard  
runtime introspection methods, like [NSObject class] and [NSObject  
isKindOfClass:], should be more than enough to implement this sort  
of functionality.



You might be right about the overengineering, but -class and - 
isKindOfClass: can't be used on methods, only on objects you already  
have. If you have an object that has a property setter, e.g.


- (void)setColor:(NSColor*) aColor;

and you have a variable selector that can refer to that (but also  
potentially many other) methods, and an object that might, or might  
not, be an NSColor*, how do you make sure you don't end up sending  
that method a string?


KVV works pretty well for this - it also solves another related  
situation where I can usefully convert objects to other types in some  
cases - but what I still haven't quite worked out yet is how to  
*prevent* the situation arising in the first place, in other words,  
given a list of methods, and an object class, how to filter out those  
methods/properties from a list that can't accept the object.


--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:
http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com

This email sent to arch...@mail-archive.com

Re: Newbie Question: implementing a stack

2008-12-20 Thread Dave DeLong

Sorry to jump in to this a little late, but a guy in our CocoaHeads  
group wrote a framework as part of his Master's thesis work.  It's a  
datastructure framework and contains a bunch of datastructures not  
available (publicly) in Cocoa, such as stacks, queues, dequeues, avl/ 
rb/aa-trees, treaps, etc.  It's open sourced under the LGPL license,  
and you can check it out of svn from our CocoaHeads page on it:


http://cocoaheads.byu.edu/code/CHDataStructures

Cheers,

Dave

On 19 Dec, 2008, at 12:22 PM, Steve Wetzel wrote:


Hi All,

I am new to this list and new to mac programming as well.  I am  
working on implementing a stack as a category of NSMutableArray.  I  
want this stack to be able to store objects.  This is what I have so  
far:


//
//  Stack.h
//  Stack implements a basic stack object in an NSMutableArray object.

#import 

@interface NSMutableArray (Stack)
-(void)push:(id)obj;//push obj of on the stack
-(id)pop;   //pop top item off the 
stack
-(id)peek;  //look at the top item 
on the stack
-(void)clear;   //remove all objects from the 
stack
-(NSUInteger)size;  //return the number of items on 
the stack
@end

//
//  Stack.m
//  Stack


#import "Stack.h"


@implementation NSMutableArray (Stack)
-(void)push:(id)obj
{
[self addObject:obj];
}

-(id)pop
{
id popedObj;
if([self count]) {
popedObj = [[[self lastObject] retain] autorelease];
[self removeLastObject];
} else {
popedObj = nil;
}
return popedObj;
}


-(id)peek;  //look at the top item 
on the stack
{
return [self lastObject];
}

-(void)clear;   //remove all objects from the 
stack
{
[self removeAllObjects];
}

-(NSUInteger)size;  //return the number of items on 
the stack
{
return [self count];
}
@end

My question involves storing objects on the stack.  As I understand  
it when I store an object on the stack, I am simply storing a  
pointer to the object on the stack.  If I create an NSString object  
called foo and do the following:


NSString *foo = [[NSString alloc] initWithString:@"foo text"];
[FractionStack push:foo];

I have put a copy of the pointer to foo on the stack.  If I now  
change foo, the top element on the stack changes also.  I do not  
want this to happen but I am not sure as to the best approach to  
make sure this does not happen.


Would it be best to copy the object to another object - say fooCopy  
then push fooCopy on the stack? Or would it be better to copy the  
object inside of the push method?  And either way, I am struggling  
with memory management.  If I first copy foo to fooCopy before  
placing it on the stack then I release fooCopy, will the element on  
the stack still be able to reference it?  do I need to retain it in  
my push method?


Also, I am thinking I likely will need to release these objects when  
I pop them and when I clear the stack.  Is that correct?


I am really trying to get my head wrapped around this but I have  
been struggling with this for some time.  If, in addition to some  
words of wisdom, you can point me in the direction of how I could  
create a simple test of this in my code to work things out on my own  
that would be appreciated also.


As I said I am new to this so if my post is not in keeping with the  
way things work on this list please let me know.


Thanks in advance,

Steve

___

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/davedelong%40me.com

This email sent to davedel...@me.com


___

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 arch...@mail-archive.com

Re: Question Regarding the Memory Address of Objects

2008-12-20 Thread Dong Feng

I have a concern that the direct use of a pointer as a memory address
may have a problem in the case of GC. In Java, object reference is not
a simple memory address because objects may be moved by GC. Does Cocoa
has a GC implementation avoid it?



2008/12/19 Carter R. Harrison :
>
> On Dec 18, 2008, at 3:44 PM, David Duncan wrote:
>
>> On Dec 18, 2008, at 11:57 AM, Carter R. Harrison wrote:
>>
>>>[dict setValue:@"button1" forKey:[NSString stringWithFormat:@"%x",
>>> &button1]]; // "button1" is one of my IBOutlets.
>>>NSString *value = [dict valueForKey:[NSString
>>> stringWithFormat:@"%x", &sender]];
>>>
>>> It seems like the memory address of the sender is different than what it
>>> should be.  I did some debugging and the address of the sender is always
>>> bfffe2e8 which really shouldn't be possible at all.
>>
>>
>> Your confusing the pointer with the object itself. button1 is a pointer to
>> a button. sender is a pointer to the object that sent the message. When you
>> take the address of either pointer, you get the location of that pointer in
>> memory - not the location of the object. The pointer itself contains the
>> location of the object in memory.
>> --
>> David Duncan
>> Apple DTS Animation and Printing
>>
>
> David,
>
> Once again you have saved me from pulling my hair out.  Thank you once again
> for your assistance.  For the rest of the folks on the thread the solution
> was to remove the "&" from the front of all my pointers.
>
> Regards,
> Carter
> ___
>
> 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/middle.fengdong%40gmail.com
>
> This email sent to middle.fengd...@gmail.com
>
___

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 arch...@mail-archive.com

Re: Question Regarding the Memory Address of Objects

2008-12-20 Thread Clark S. Cox III

The objective-c garbage collector does not move objects, so that is  
not an issue with using raw pointers.


Sent from my iPhone

On Dec 20, 2008, at 6:50, "Dong Feng"  wrote:


I have a concern that the direct use of a pointer as a memory address
may have a problem in the case of GC. In Java, object reference is not
a simple memory address because objects may be moved by GC. Does Cocoa
has a GC implementation avoid it?



2008/12/19 Carter R. Harrison :


On Dec 18, 2008, at 3:44 PM, David Duncan wrote:


On Dec 18, 2008, at 11:57 AM, Carter R. Harrison wrote:

  [dict setValue:@"button1" forKey:[NSString  
stringWithFormat:@"%x",

&button1]]; // "button1" is one of my IBOutlets.
  NSString *value = [dict valueForKey:[NSString
stringWithFormat:@"%x", &sender]];

It seems like the memory address of the sender is different than  
what it
should be.  I did some debugging and the address of the sender is  
always

bfffe2e8 which really shouldn't be possible at all.



Your confusing the pointer with the object itself. button1 is a  
pointer to
a button. sender is a pointer to the object that sent the message.  
When you
take the address of either pointer, you get the location of that  
pointer in
memory - not the location of the object. The pointer itself  
contains the

location of the object in memory.
--
David Duncan
Apple DTS Animation and Printing



David,

Once again you have saved me from pulling my hair out.  Thank you  
once again
for your assistance.  For the rest of the folks on the thread the  
solution

was to remove the "&" from the front of all my pointers.

Regards,
Carter
___

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/middle.fengdong%40gmail.com

This email sent to middle.fengd...@gmail.com


___

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/clarkcox3%40gmail.com

This email sent to clarkc...@gmail.com

___

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 arch...@mail-archive.com

Sorting an NSMutableArray

2008-12-20 Thread Jacob Rhoden

How do you sort an NSMutableArray, when the difference between the two 
objects is determined by information not contained completely within the 
objects themselves. ie in Java you can implement a Comparator that takes 
two objects to compare then. Is this similar in Cocoa?


ie in this case I need to apply a mathematical algorithm using 
information from each object to determine which one is greater than the 
other.


Thanks!
Jacob


http://jacobrhoden.com
___

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 arch...@mail-archive.com

Re: Sorting an NSMutableArray

2008-12-20 Thread Ricky Sharp



On Dec 20, 2008, at 6:15 AM, Jacob Rhoden wrote:

How do you sort an NSMutableArray, when the difference between the  
two objects is determined by information not contained completely  
within the objects themselves. ie in Java you can implement a  
Comparator that takes two objects to compare then. Is this similar  
in Cocoa?


ie in this case I need to apply a mathematical algorithm using  
information from each object to determine which one is greater than  
the other.



See sortUsingSelector:


What I then personally do is to create a "Sorting" category on  
whatever object it is that the collection will contain.  For example:


@interface MyObject (Sorting)
- (NSComparisonResult)myCustomCompare:(MyObject*)rhs;
@end

Then, LHS is self and RHS is the param coming in.


___
Ricky A. Sharp mailto:rsh...@instantinteractive.com
Instant Interactive(tm)   http://www.instantinteractive.com



___

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 arch...@mail-archive.com

Re: Listening for changes

Thanks for the idea Mike.  Not sure how I could implement this without  
sending the notification in every setter of the class though...



Andre Masse

On Dec 19, 2008, at 19:24, Mike Abdullah wrote:

Since you apparently want to just know that one of the fields  
changed, but not which one specifically, why not just have the class  
post an NSNotification called MyClassDidChange or similar?


Mike.


___

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 arch...@mail-archive.com

Re: Optimizing NSRectFill

2008-12-20 Thread Oleg Krupnov

Here is how I arrived at the conclusion that NSRectFill is the
bottleneck. (I cannot show the code because it's spread over objects).

The slowness of redraw is only noticeable when I drag the mouse, e.g.
to resize an object in the custom view. I see that update of the
resized object is always late after the actual position of the mouse
pointer, and the latency is sometimes pretty great to not be excused.
Especially when the size of the rectangle to be updated is large.

My custom view supports zooming, so I call NSRectFill after setting an
NSAffineTransform. Larger zooms, as I said, seem to cause greater
latency. Sometimes I also use clipping ([NSBezierPath addClip]) before
calling NSRectFill.

I never call -[NSView displayIfNeeded] or use other ways of forcing
the view to update itself. I always rely on -[NSView
setNeedsDisplayInRect:] to invalidate the rectangle(s) that need to be
updated and then wait until the next refresh. What is true, that
-[NSView setNeedsDisplayInRect:] are often called multiple times for
the same rectangles, but I don't think it matters.

I run my app with the Shark tool and drag the mouse to resize an
object, to and fro, for few seconds, then check the sample. It always
shows that 20-30% of time is taken by the NSRectFill function, which
is called only once to simply draw the background of the view with
solid white! This is ridiculous. I tried to put an NSLog  near the
NSRectFill call to check if it is called extra times, but it seems
that it is being called exactly one time per single mouse drag
message. When I comment out the call to NSRectFill, then the refresh
becomes noticeably smoother, and the Shark gives leadership to other
bottlenecks, in my case the NSBezierPath stroke and fill messages, for
the objects drawn in the view.

Any ideas?

On Sat, Dec 20, 2008 at 1:45 AM, Peter Ammon  wrote:
>
> On Dec 19, 2008, at 7:37 AM, Oleg Krupnov wrote:
>
>> I'm developing a custom view and noticed that drawRect is becoming
>> increasingly slow. The Shark has pointed out that the bottleneck is
>> the NSRectFill function (that calls CGContextFillRect under the hood)
>> that I use to draw the background of the view.
>
> A common cause of apparently excessive time in graphics related ops is that
> you are running afoul of coalesced updates by attempting to refresh faster
> than 60 times a second.
>
> See
> http://developer.apple.com/documentation/Performance/Conceptual/Drawing/Articles/FlushingContent.html
>  .
>  Try disabling coalesced updates in Quartz Debug and see if it speeds up
> your drawing.  If so, ensure your app does not refresh faster than 60 Hz.
>
> -Peter
>
>
___

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 arch...@mail-archive.com

Re: Optimizing NSRectFill

Le 20 déc. 08 à 15:22, Oleg Krupnov a écrit :

Here is how I arrived at the conclusion that NSRectFill is the
bottleneck. (I cannot show the code because it's spread over objects).

The slowness of redraw is only noticeable when I drag the mouse, e.g.
to resize an object in the custom view. I see that update of the
resized object is always late after the actual position of the mouse
pointer, and the latency is sometimes pretty great to not be excused.
Especially when the size of the rectangle to be updated is large.

My custom view supports zooming, so I call NSRectFill after setting an
NSAffineTransform. Larger zooms, as I said, seem to cause greater
latency. Sometimes I also use clipping ([NSBezierPath addClip]) before
calling NSRectFill.

I never call -[NSView displayIfNeeded] or use other ways of forcing
the view to update itself. I always rely on -[NSView
setNeedsDisplayInRect:] to invalidate the rectangle(s) that need to be
updated and then wait until the next refresh. What is true, that
-[NSView setNeedsDisplayInRect:] are often called multiple times for
the same rectangles, but I don't think it matters.

I run my app with the Shark tool and drag the mouse to resize an
object, to and fro, for few seconds, then check the sample. It always
shows that 20-30% of time is taken by the NSRectFill function, which
is called only once to simply draw the background of the view with
solid white! This is ridiculous. I tried to put an NSLog near the
NSRectFill call to check if it is called extra times, but it seems
that it is being called exactly one time per single mouse drag
message. When I comment out the call to NSRectFill, then the refresh
becomes noticeably smoother, and the Shark gives leadership to other
bottlenecks, in my case the NSBezierPath stroke and fill messages, for
the objects drawn in the view.

Any ideas?

On Sat, Dec 20, 2008 at 1:45 AM, Peter Ammon wrote:

On Dec 19, 2008, at 7:37 AM, Oleg Krupnov wrote:

I'm developing a custom view and noticed that drawRect is becoming
increasingly slow. The Shark has pointed out that the bottleneck is
the NSRectFill function (that calls CGContextFillRect under the
hood)

that I use to draw the background of the view.

A common cause of apparently excessive time in graphics related ops
is that
you are running afoul of coalesced updates by attempting to refresh
faster

than 60 times a second.

See
http://developer.apple.com/documentation/Performance/Conceptual/Drawing/Articles/FlushingContent.html
.
Try disabling coalesced updates in Quartz Debug and see if it
speeds up
your drawing. If so, ensure your app does not refresh faster than
60 Hz.

-Peter

___

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/devlists%40shadowlab.org

This email sent to devli...@shadowlab.org

___

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 arch...@mail-archive.com

Re: Optimizing NSRectFill

Sorry for the blank previous mail.

Try to call NSFillRect with the rect pased as parameter in drawRect:
and before applying the scaling method.
As the rect parameter contains the invalid rect in screen coordinates,
you don't have to transform it before filling the rect with white.

Le 20 déc. 08 à 15:22, Oleg Krupnov a écrit :

Here is how I arrived at the conclusion that NSRectFill is the
bottleneck. (I cannot show the code because it's spread over objects).

Any ideas?

On Sat, Dec 20, 2008 at 1:45 AM, Peter Ammon wrote:

On Dec 19, 2008, at 7:37 AM, Oleg Krupnov wrote:

that I use to draw the background of the view.

A common cause of apparently excessive time in graphics related ops
is that
you are running afoul of coalesced updates by attempting to refresh
faster

than 60 times a second.

-Peter

___

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/devlists%40shadowlab.org

This email sent to devli...@shadowlab.org

___

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 arch...@mail-archive.com

Re: Get notified of fullscreen switch

2008-12-20 Thread Joachim Deelen


Hi Eric,

thanks for the quick response.

"Listening" to the Carbon Event you described did it. The only  
"disadvantage"  is: Now the App depends on Carbon.framework so far it  
was Cocoa, Quartz, QuartzCore and ScriptingBridge only


But, between the lines of your reply I could read that there'll be a  
pure Cocoa solution in the future.


regards,
Joachim

Am 17.12.2008 um 18:59 schrieb Eric Schlegel:



On Dec 17, 2008, at 9:26 AM, Joachim Deelen wrote:

Is there a way, that my application (Playwatch), which is an iTunes  
Controller running in the menu bar, gets notified, when another  
Application, like DVD-Player, switches to full screen (kiosk) mode?


At the moment, the recommended way to do this is to use a Carbon  
event handler for the kEventAppSystemUIModeChanged Carbon event.


-eric



___

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 arch...@mail-archive.com

Re: Optimizing NSRectFill

Le 20 déc. 08 à 15:33, Jean-Daniel Dupas a écrit :

Sorry for the blank previous mail.

Try to call NSFillRect with the rect pased as parameter in drawRect:
and before applying the scaling method.
As the rect parameter contains the invalid rect in screen
coordinates, you don't have to transform it before filling the rect
with white.

To be exact, the rect parameter is in screen coordinates. So you have
to convert it to your view coordinates before drawing.
And so, you have to apply to it the reverse of your "view -> screen"
transform (that you apply to the context).

And so, this is what append:
- You apply scaling to the context.
- You apply reverse scalling to the rect.
- You call Fill Rect =>The system applies the context's transformation
to your rect (and revert it back to screen coord) and fills it.

That's why I suggest you can fill it directly before applying the
transform to the context.
But for other drawing, you should use the transformed rect to compute
what should be refreshed.

Le 20 déc. 08 à 15:22, Oleg Krupnov a écrit :

Here is how I arrived at the conclusion that NSRectFill is the
bottleneck. (I cannot show the code because it's spread over
objects).

My custom view supports zooming, so I call NSRectFill after setting
an

NSAffineTransform. Larger zooms, as I said, seem to cause greater
latency. Sometimes I also use clipping ([NSBezierPath addClip])
before

calling NSRectFill.

I never call -[NSView displayIfNeeded] or use other ways of forcing
the view to update itself. I always rely on -[NSView
setNeedsDisplayInRect:] to invalidate the rectangle(s) that need to
be

updated and then wait until the next refresh. What is true, that
-[NSView setNeedsDisplayInRect:] are often called multiple times for
the same rectangles, but I don't think it matters.

the objects drawn in the view.

Any ideas?

On Sat, Dec 20, 2008 at 1:45 AM, Peter Ammon
wrote:

On Dec 19, 2008, at 7:37 AM, Oleg Krupnov wrote:

that I use to draw the background of the view.

A common cause of apparently excessive time in graphics related
ops is that
you are running afoul of coalesced updates by attempting to
refresh faster

than 60 times a second.

-Peter

___

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/devlists%40shadowlab.org

This email sent to devli...@shadowlab.org

___

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/devlists%40shadowlab.org

This email sent to devli...@shadowlab.org

___

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 arch...@mail-archive.com

Trackpad multi touch events in Mac OSX

2008-12-20 Thread Andreas Eriksson

Hi,
is there a public API to catch the new trackpad multi touch events in Mac
OSX?

Regards,

Andreas
___

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 arch...@mail-archive.com

MathPaper (Chapter 10) fails

2008-12-20 Thread Richard S. French

MathPaper project  chapter 10 in Oreilly¹s ³Building Cocoa Applications²
fails with the following error:
³2008-12-20 11:23:59.664 MathPaper[517] *** -[NSCFDictionary
setObject:forKey:]: attempt to insert nil value²

Note that I am using Xcode 2.4.1. I also have the evaluator copied under
MathPaper in targets.
I have searched the archives and found others have had this problem but I
did not find a resolution.
Evaluator fails to launch?

Code is as follows:
- (void)windowDidLoad
{
NSString *path=0;

[super windowDidLoad];
[[self window] makeFirstResponder:theText];

path = [[NSBundle mainBundle] pathForAuxiliaryExecutable:@"Evaluator"];
NSLog(@"path set");

toPipe = [NSPipe pipe];
fromPipe = [NSPipe pipe];

toEvaluator = [toPipe fileHandleForWriting];
fromEvaluator = [fromPipe fileHandleForReading];
NSLog(@"pipe set");

evaluator = [[NSTask alloc] init];

[evaluator setLaunchPath:path];
[evaluator setStandardOutput:fromPipe];
[evaluator setStandardInput:toPipe];
[evaluator launch];
NSLog(@"Evaluator launched");

[[NSNotificationCenter defaultCenter]
addObserver:self
   selector:@selector(gotData:)
   name:NSFileHandleReadCompletionNotification
 object:fromEvaluator];

[fromEvaluator readInBackgroundAndNotify];

}

___

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 arch...@mail-archive.com

How to use animation for CALayer frame.size attribute?

2008-12-20 Thread Alex . Wang

Hi everyone.I am trying to create some animation when to change the size of
a CALayer object.
My code looks like this:
CABasicAnimation * _animation =[CABasicAnimation animation].
[_animation setKeyPath:@"frame.size"];
NSSize _size;

*_size.**width** = 400**;*

*_size.**height** = 300**;*

*_animation.fromValue = [NSValue valueWithSize:_size];*

*_size.**width** = 200**;*

*_size.**height** = 150**;*
*_animation.toValue = [**NSValue** **valueWithSize**:_size];*

*[_layer **addAnimation**:_animation*

*  **forKey**:_animationKey];*


However, the above code doesn't work at all. Can anyone tell me what is
wrong with that? Or did I miss anything?

Thank you very much for any help or guidance.

Good luck.


-- 
Best regards.
___

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 arch...@mail-archive.com

Re: Using runtime functions to match up object types safely when setting properties

On Sat, Dec 20, 2008 at 5:16 AM, Graham Cox  wrote:
> KVV works pretty well for this - it also solves another related situation
> where I can usefully convert objects to other types in some cases - but what
> I still haven't quite worked out yet is how to *prevent* the situation
> arising in the first place, in other words, given a list of methods, and an
> object class, how to filter out those methods/properties from a list that
> can't accept the object.

There is no way to prevent the situation in the first place using
standard techniques. All that static typing information gets thrown
away when you compile. At runtime all that survives is the knowledge
that your method takes an object pointer of some kind.

You could build a table, either manually or perhaps with the help of
some sort of crazy macro. But there is absolutely no way to get the
NSColor out of -(void)setColor:(NSColor*)c at runtime because that
info simply isn't there.

Mike
___

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 arch...@mail-archive.com

Re: Question Regarding the Memory Address of Objects

On Sat, Dec 20, 2008 at 7:00 AM, Clark S. Cox III  wrote:
> The objective-c garbage collector does not move objects, so that is not an
> issue with using raw pointers.

But note that storing a pointer value in a string like this is still
problematic under garbage collection. It will not act like a strong
reference, so the object may be destroyed while still being
"referenced" by the string. But it will not act like a zeroing weak
reference either, so you'll still be able to get the pointer even
after it's been destroyed, which will (probably) lead to a crash.

An NSMapTable can be configured to give either reasonable behavior
(strong or zeroing weak reference) so use that instead.

Mike
___

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 arch...@mail-archive.com

Re: Optimizing NSRectFill

On Sat, Dec 20, 2008 at 9:22 AM, Oleg Krupnov  wrote:
> Here is how I arrived at the conclusion that NSRectFill is the
> bottleneck. (I cannot show the code because it's spread over objects).
>
> The slowness of redraw is only noticeable when I drag the mouse, e.g.
> to resize an object in the custom view. I see that update of the
> resized object is always late after the actual position of the mouse
> pointer, and the latency is sometimes pretty great to not be excused.
> Especially when the size of the rectangle to be updated is large.

When doing things like this you always need to coalesce events before
drawing. If you're triggering a redraw for every event then you always
leave yourself open to problems. The moment that your events come in
faster than your redraws can occur, you lose. This doesn't necessarily
only happen when your code is really slow. Maybe you find a user on a
really slow computer. Maybe the computer starts to swap. Maybe the
computer is running 300 CPU intensive processes simultaneously. Maybe
the mouse is sending events at 120Hz and the screen is only refreshing
at 60Hz.

A simple way to do this is to avoid calling setNeedsDisplay: (or its
friends) directly from your event handler. Instead start a timer with
some suitably small interval, and set a flag. If the flag is already
set, don't start the timer, it's already been started. When the timer
fires, invalidate the view and clear the flag.

Once you've done this, if it's still too slow then it will manifest as
simply being unsmooth, but it will still follow your mouse movements
in realtime, which is a substantial improvement. If it's still too
slow at that point then you can see what else might be going on and
try to fix it from a better position.

Mike
___

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 arch...@mail-archive.com

toolbar to view gradient displaying correctly only sometimes?

2008-12-20 Thread Chunk 1978

i have a window with a toolbar, and instead of having a sharp
separation between the toolbar and window, i've created a gradient
from toolbar color to window color (which is simply a custom view
placed over the window).  this gradient looks correct on my iMac 24",
and correct on other MacBooks.  however, i'm now on my dad's 20" iMac
and the gradient isn't smooth at all.  it is clearly a gradient band
that is separating the toolbar from the window's view area, instead of
a proper gradient.

i'm assuming this is a problem with the screen/color on my dad's 20"
iMac (Alu 2007), but i can't be sure.

it would be great for others to look at this attached image and post
back whether the gradient is proper, or if it is clearly a band
because the gradient's colors are darker than they should be.  also,
eventhough this isn't a cocoa dev question, is there a reason why my
dad's 20 alu iMac 2007 would be displaying this gradient improperly?
<>___

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 arch...@mail-archive.com

Re: Optimizing NSRectFill

2008-12-20 Thread Bill Bumgarner


On Dec 20, 2008, at 9:18 AM, Michael Ash wrote:

When doing things like this you always need to coalesce events before
drawing. If you're triggering a redraw for every event then you always
leave yourself open to problems. The moment that your events come in
faster than your redraws can occur, you lose. This doesn't necessarily
only happen when your code is really slow. Maybe you find a user on a
really slow computer. Maybe the computer starts to swap. Maybe the
computer is running 300 CPU intensive processes simultaneously. Maybe
the mouse is sending events at 120Hz and the screen is only refreshing
at 60Hz.


And... maybe the test case involves grabbing something w/the mouse and  
then rapidly waving the object back and forth over the view to  
"quickly" grab some data to sample, thus making the original problem  
seem much, much, worse.


(No -- I'm not saying the problem doesn't exist... just that we  
developers tend toward spastic monkey behavior when optimizing  
something that a real user would perform with directness.)


b.bum
___

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 arch...@mail-archive.com

Re: Question Regarding the Memory Address of Objects

2008-12-20 Thread Scott Ribe

> An NSMapTable can be configured to give either reasonable behavior
> (strong or zeroing weak reference) so use that instead.

Or, in this special case, simply make sure that the dictionary is only used
while the controller is live--especially easy if the dictionary is a private
field of the controller...

I forget the details of the original message, and whether the code has to
dynamically adapt to different sets of outlets. If not, then consider:

enum { kButton1 = 0, kButton2, kText1, ..., kNumOfControls};

id controls[kNumOfControls];

controls[kButton1] = button1;
...

For (int i = 0; i < kNumOfControls; ++i)
  [controls[i] doSomething: ...];


-- 
Scott Ribe
scott_r...@killerbytes.com
http://www.killerbytes.com/
(303) 722-0567 voice


___

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 arch...@mail-archive.com

selectionShouldChangeInTableView called twice


Hi,

I'm using a master/detail view in my application. If the user select  
another row in the master table and the detail has been modified, I  
want to present an alert to save, cancel or return to detail. I'm not  
using binding here, only datasource and delegate methods. So in - 
selectionShouldChangeInTableView: I return YES if the user saved or  
cancelled and NO if he choose the third option.


Problem is that selectionShouldChangeInTableView is called twice when  
returning NO. Having googled it, I found a couple of matches on this  
list and no replies... To make sure this is a bug, I made a test  
project available here and will radar it if it has no workaround:


http://idisk.mac.com/miyano/Public/TestRowChanges.zip

Now, since I can't rely on -selectionShouldChangeInTableView:, how do  
you guys handle this?


Thanks,

Andre Masse
___

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 arch...@mail-archive.com

NSBezierPath

2008-12-20 Thread Amr Nashaat


Hi,

does anyone knows how can I use the NSBezierPath function to create a  
circle?


thanx.
___

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 arch...@mail-archive.com

Re: NSBezierPath

2008-12-20 Thread Randall Meadows


On Dec 20, 2008, at 11:35 AM, Amr Nashaat wrote:


does anyone knows how can I use the NSBezierPath function


NSBezierPath is not a "function", but anyway...


to create a circle?


+ (NSBezierPath *)bezierPathWithOvalInRect:(NSRect)aRect

where aRect describes a square.
___

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 arch...@mail-archive.com

Re: Optimizing NSRectFill

On Sat, Dec 20, 2008 at 12:18 PM, Michael Ash  wrote:
> A simple way to do this is to avoid calling setNeedsDisplay: (or its
> friends) directly from your event handler. Instead start a timer with
> some suitably small interval, and set a flag. If the flag is already
> set, don't start the timer, it's already been started. When the timer
> fires, invalidate the view and clear the flag.

Alright, perhaps I don't completely understand how the drawing model
works at a low enough level, but why would this be helpful?  Doesn't
AppKit just coalesce updates received by -setNeedsDisplayInRect: until
Quartz has finished pushing pixels?  So if you are accumulating dirty
rects in your event handler, your drawing should still be limited to
min(as fast as possible, 60hz).

Or am I thinking of the Quartz GL pipeline?

--Kyle Sluder
___

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 arch...@mail-archive.com

Re: toolbar to view gradient displaying correctly only sometimes?

2008-12-20 Thread Benjamin Dobson

It certainly doesn't blend the toolbar to the toolbar to the window,  
but I don't see anything else wrong with the gradient.


Why do you want to do this anyway? Note that the gradient will always  
look silly on Tiger.


On 20 Dec 2008, at 17:39:16, Chunk 1978 wrote:


i have a window with a toolbar, and instead of having a sharp
separation between the toolbar and window, i've created a gradient
from toolbar color to window color (which is simply a custom view
placed over the window).  this gradient looks correct on my iMac 24",
and correct on other MacBooks.  however, i'm now on my dad's 20" iMac
and the gradient isn't smooth at all.  it is clearly a gradient band
that is separating the toolbar from the window's view area, instead of
a proper gradient.

i'm assuming this is a problem with the screen/color on my dad's 20"
iMac (Alu 2007), but i can't be sure.

it would be great for others to look at this attached image and post
back whether the gradient is proper, or if it is clearly a band
because the gradient's colors are darker than they should be.  also,
eventhough this isn't a cocoa dev question, is there a reason why my
dad's 20 alu iMac 2007 would be displaying this gradient improperly?
___

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/importedfromspace%40googlemail.com

This email sent to importedfromsp...@googlemail.com


___

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 arch...@mail-archive.com

Find Panel Stopped Working (maybe)

2008-12-20 Thread John Nairn

I have the "Uses Find Panel" checked for the only NSTextView in a  
window, but it does not use a Find panel. I think it used to use the  
Find Panel before (but I am not sure). Is there something more that  
is needed besides checking that box?


Some other possibilities:

1. I noticed the items in the Find menu were not connected to any  
actions. I tried connecting them all to performFindPanelAction: (I  
could not find any other options that make sense), but that did not  
make it work.


2. The NSTextView is in an NSScrollField which is in an NSSplitView.  
I used to used a standard NSTextView but recently subclassed it for  
some extra features. The subclass does very little. I tried  
commenting out all its methods (so it does nothing more than an  
NSTextView) and that did not make the Find Panel work either.


---
John Nairn
GEDitCOM - Genealogy Software for the Macintosh
http://www.geditcom.com


___

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 arch...@mail-archive.com

Re: MathPaper (Chapter 10) fails

On Sat, Dec 20, 2008 at 11:28 AM, Richard S. French
 wrote:
> Note that I am using Xcode 2.4.1. I also have the evaluator copied under
> MathPaper in targets.

Where is it being copied to?  (It is under a "Copy Files" phase, correct?).

Instead of just printing "path set", try logging the value of path to
make sure that -[NSBundle pathForAuxiliaryExecutable:] was actually
able to find the path of your Evaluator binary.  Or, add breakpoints
on objc_exception_throw and -[NSException raise] and examine the value
in the debugger.

--Kyle Sluder
___

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 arch...@mail-archive.com

Re: Get notified of fullscreen switch

2008-12-20 Thread Eric Schlegel



On Dec 20, 2008, at 6:49 AM, Joachim Deelen wrote:


Hi Eric,

thanks for the quick response.

"Listening" to the Carbon Event you described did it. The only  
"disadvantage"  is: Now the App depends on Carbon.framework so far  
it was Cocoa, Quartz, QuartzCore and ScriptingBridge only


But, between the lines of your reply I could read that there'll be a  
pure Cocoa solution in the future.


The AppKit team is aware of the desire for a Cocoa-only solution here  
and they have some plans to address it. That's all I can say for now. :)


-eric

___

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 arch...@mail-archive.com

Re: Find Panel Stopped Working (maybe)

On Sat, Dec 20, 2008 at 2:10 PM, John Nairn  wrote:
> 1. I noticed the items in the Find menu were not connected to any actions. I
> tried connecting them all to performFindPanelAction: (I could not find any
> other options that make sense), but that did not make it work.

Looking at TextEdit's MainMenu.nib, Find, Find Next, Find Previous,
and Use Selection for Find are all wired up to First Responder's
-performFindPanelAction:.  Find is tag 1, Find Next=2, Find
Previous=3, and Use Selection=7.  Jump to Selection is wired up to
First Responder's -centerSelectionInVisibleArea:.

Maybe you've wired up your actions to the wrong object?

--Kyle Sluder
___

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 arch...@mail-archive.com

Re: How to use animation for CALayer frame.size attribute?

2008-12-20 Thread Matt Long

I'm hoping that all of the asterisks are just artifacts from a copy/ 
paste maneuver or something, however, if they are not then you're  
missing some serious fundamentals. This code will never compile.


Let's just assume a copy/paste issue. If you are wanting to animate  
the size of the frame, use the bounds property instead of frame.


http://developer.apple.com/qa/qa2008/qa1620.html

-Matt



On Dec 20, 2008, at 10:04 AM, Alex.Wang wrote:

Hi everyone.I am trying to create some animation when to change the  
size of

a CALayer object.
My code looks like this:
CABasicAnimation * _animation =[CABasicAnimation animation].
[_animation setKeyPath:@"frame.size"];
NSSize _size;

*_size.**width** = 400**;*

*_size.**height** = 300**;*

*_animation.fromValue = [NSValue valueWithSize:_size];*

*_size.**width** = 200**;*

*_size.**height** = 150**;*
*_animation.toValue = [**NSValue** **valueWithSize**:_size];*

*[_layer **addAnimation**:_animation*

*  **forKey**:_animationKey];*


However, the above code doesn't work at all. Can anyone tell me what  
is

wrong with that? Or did I miss anything?

Thank you very much for any help or guidance.

Good luck.


--
Best regards.
___


___

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 arch...@mail-archive.com

Re: Find Panel Stopped Working (maybe)

2008-12-20 Thread John Nairn

Thanks Kyle. That did solve it. I am guessing it never worked before  
and that my original nib file was created before 10.3 and thus did  
not use the required performFindPanelAction tags.


I did wire them to the correct actions, by I assume NSTextView  
validated according to those tags and since they were wrong, the menu  
commands were always disabled.


On Dec 20, 2008, at 11:18 AM, Kyle Sluder wrote:


On Sat, Dec 20, 2008 at 2:10 PM, John Nairn  wrote:
1. I noticed the items in the Find menu were not connected to any  
actions. I
tried connecting them all to performFindPanelAction: (I could not  
find any

other options that make sense), but that did not make it work.


Looking at TextEdit's MainMenu.nib, Find, Find Next, Find Previous,
and Use Selection for Find are all wired up to First Responder's
-performFindPanelAction:.  Find is tag 1, Find Next=2, Find
Previous=3, and Use Selection=7.  Jump to Selection is wired up to
First Responder's -centerSelectionInVisibleArea:.

Maybe you've wired up your actions to the wrong object?

--Kyle Sluder


---
John Nairn
GEDitCOM - Genealogy Software for the Macintosh
http://www.geditcom.com


___

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 arch...@mail-archive.com

Re: Trackpad multi touch events in Mac OSX

2008-12-20 Thread Scott Anguish



On 20-Dec-08, at 10:56 AM, Andreas Eriksson wrote:


Hi,
is there a public API to catch the new trackpad multi touch events  
in Mac

OSX?


No.
___

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 arch...@mail-archive.com

Re: NSBezierPath

On Sat, Dec 20, 2008 at 1:35 PM, Amr Nashaat  wrote:
> Hi,
>
> does anyone knows how can I use the NSBezierPath function to create a
> circle?

Have you considered searching the internet? If I take your question
and put it verbatim into Google, the very first hit that comes back
has sample code that exactly answers your question

Mike
___

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 arch...@mail-archive.com

Re: Optimizing NSRectFill

On Sat, Dec 20, 2008 at 1:51 PM, Kyle Sluder  wrote:
> On Sat, Dec 20, 2008 at 12:18 PM, Michael Ash  wrote:
>> A simple way to do this is to avoid calling setNeedsDisplay: (or its
>> friends) directly from your event handler. Instead start a timer with
>> some suitably small interval, and set a flag. If the flag is already
>> set, don't start the timer, it's already been started. When the timer
>> fires, invalidate the view and clear the flag.
>
> Alright, perhaps I don't completely understand how the drawing model
> works at a low enough level, but why would this be helpful?  Doesn't
> AppKit just coalesce updates received by -setNeedsDisplayInRect: until
> Quartz has finished pushing pixels?  So if you are accumulating dirty
> rects in your event handler, your drawing should still be limited to
> min(as fast as possible, 60hz).
>
> Or am I thinking of the Quartz GL pipeline?

My understanding is that everything is sychronous. So if you call
setNeedsDisplay: in your event handler, then the sequence will always
look like this:

event handler
   mark view invalid
check for invalidated views
   redraw invalidated view
   flush results to screen
event handler
   mark view invalid
check for invalidated views
   redraw invalidated view
   flush results to screen

And so forth. I could be wrong about this, and if anyone knows better
please pipe up!

Mike
___

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 arch...@mail-archive.com

Re: Sorting an NSMutableArray

2008-12-20 Thread Rob Rix

See -sortUsingFunction:context: in NSMutableArray’s documentation. As  
Ricky Sharp has already mentioned, there’s also -sortUsingSelector:,  
however if you’re comparing partly with external data you might like  
having the void * context provided by this method.


Sincerely,
Rob

On 20-Dec-08, at 7:15 AM, Jacob Rhoden wrote:

How do you sort an NSMutableArray, when the difference between the  
two objects is determined by information not contained completely  
within the objects themselves. ie in Java you can implement a  
Comparator that takes two objects to compare then. Is this similar  
in Cocoa?


ie in this case I need to apply a mathematical algorithm using  
information from each object to determine which one is greater than  
the other.


Thanks!
Jacob


http://jacobrhoden.com
___

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/rix.rob%40gmail.com

This email sent to rix@gmail.com


___

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 arch...@mail-archive.com

Re: Optimizing NSRectFill



Le 20 déc. 08 à 21:14, Michael Ash a écrit :

On Sat, Dec 20, 2008 at 1:51 PM, Kyle Sluder   
wrote:
On Sat, Dec 20, 2008 at 12:18 PM, Michael Ash  
 wrote:

A simple way to do this is to avoid calling setNeedsDisplay: (or its
friends) directly from your event handler. Instead start a timer  
with

some suitably small interval, and set a flag. If the flag is already
set, don't start the timer, it's already been started. When the  
timer

fires, invalidate the view and clear the flag.


Alright, perhaps I don't completely understand how the drawing model
works at a low enough level, but why would this be helpful?  Doesn't
AppKit just coalesce updates received by -setNeedsDisplayInRect:  
until

Quartz has finished pushing pixels?  So if you are accumulating dirty
rects in your event handler, your drawing should still be limited to
min(as fast as possible, 60hz).

Or am I thinking of the Quartz GL pipeline?


My understanding is that everything is sychronous. So if you call
setNeedsDisplay: in your event handler, then the sequence will always
look like this:

event handler
  mark view invalid
check for invalidated views
  redraw invalidated view
  flush results to screen
event handler
  mark view invalid
check for invalidated views
  redraw invalidated view
  flush results to screen

And so forth. I could be wrong about this, and if anyone knows better
please pipe up!


I did a quick test.
I have a 1ms timer which just do -setNeedDisplay: on a small view.
And the result is that my timer callback is called 60 times per seconds.
If I remove the seetNeedDisplay: call, it is called 1000 times (as  
expected).


So I think that you're right (and that everything is synchronous) but  
I also think that the 'flush results to screen' blocks until the  
screen is ready and so limit the number of redraw to the refresh rate  
of the screen.
If this is effectively the case, your timer trick will not be very  
helpful.






___

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 arch...@mail-archive.com

Re: Optimizing NSRectFill



Le 20 déc. 08 à 22:47, Jean-Daniel Dupas a écrit :



Le 20 déc. 08 à 21:14, Michael Ash a écrit :

On Sat, Dec 20, 2008 at 1:51 PM, Kyle Sluder  
 wrote:
On Sat, Dec 20, 2008 at 12:18 PM, Michael Ash  
 wrote:
A simple way to do this is to avoid calling setNeedsDisplay: (or  
its
friends) directly from your event handler. Instead start a timer  
with
some suitably small interval, and set a flag. If the flag is  
already
set, don't start the timer, it's already been started. When the  
timer

fires, invalidate the view and clear the flag.


Alright, perhaps I don't completely understand how the drawing model
works at a low enough level, but why would this be helpful?  Doesn't
AppKit just coalesce updates received by -setNeedsDisplayInRect:  
until
Quartz has finished pushing pixels?  So if you are accumulating  
dirty

rects in your event handler, your drawing should still be limited to
min(as fast as possible, 60hz).

Or am I thinking of the Quartz GL pipeline?


My understanding is that everything is sychronous. So if you call
setNeedsDisplay: in your event handler, then the sequence will always
look like this:

event handler
 mark view invalid
check for invalidated views
 redraw invalidated view
 flush results to screen
event handler
 mark view invalid
check for invalidated views
 redraw invalidated view
 flush results to screen

And so forth. I could be wrong about this, and if anyone knows better
please pipe up!


I did a quick test.
I have a 1ms timer which just do -setNeedDisplay: on a small view.
And the result is that my timer callback is called 60 times per  
seconds.
If I remove the seetNeedDisplay: call, it is called 1000 times (as  
expected).


So I think that you're right (and that everything is synchronous)  
but I also think that the 'flush results to screen' blocks until the  
screen is ready and so limit the number of redraw to the refresh  
rate of the screen.
If this is effectively the case, your timer trick will not be very  
helpful.




How, and I found that too:

http://developer.apple.com/technotes/tn2005/tn2133.html






___

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 arch...@mail-archive.com

Re: Text track in QTMovie

2008-12-20 Thread Joe Turner

Okay, I took your suggestions. It seems that my array of characters is  
the issue. My ByteCount s and other variables have seemingly correct  
values, but the array of chars' value does not change from before it  
gets written to. So, I tried a few more ways and no luck.


I tried using a UInt8 *chars = malloc(size)
char *chars = malloc(size). But still no luck.

I have a feeling that it has to do with how I initialize it.

K&R is coming this week, so, that should help me too (thanks for the  
recommendation!!).


Thanks a ton!

Joe
On Dec 16, 2008, at 1:28 PM, douglas welton wrote:


Hi Joe,

My suggestions:

1)  write a small application that test your routine before you give  
it someone else to use.  given the code you already have doing this  
should take all of 5 minutes.  Once you do this, you can use the  
debugger and find your boo-boos.  I suspect they will be quite  
obvious.


2) GetMediaSample2 returns an error code.  However, the code in your  
email does not check the returned value.  Making an assumption that  
the function has completed without an error may not be in your term  
best interest ;^}


3) chars[] is not a C function call.  K&R is your friend!

regards,

douglas


On Dec 16, 2008, at 1:40 PM, Joe Turner wrote:


Sorry for taking so long.

Okay, what I am looking for is the contents of the text track.  
Like, for a movie, it would be the chapter names and times.


The issue is that I am making this for someone who is using  
AppleScript, and is just using a call method on my method to do  
this. So, the Debugger will not stop at my breakpoints (I thought  
it did, even if it is being called by a call method).


Anyways, I guess what I am looking for is an error in the code I  
have created. Like, am I passing something wrong? Am I trying to  
parse something as a different format than it is? I can create a  
new project and do this and tell you the debugger contents too  
though, if needed.


I am used to objective-c, and some of these C calls can confuse me  
sometimes. Like, is chars[] the same thing as a UInt8?


Thanks for your help,

Joe
On Dec 14, 2008, at 8:54 PM, douglas welton wrote:


when you say it won't show you the "track names" do you mean:

a) your NSLog() function is failing and nothing is printed
b) NSLog() prints something, but the value of string appears to be  
either nil or empty (which one?)
c)  you really want the name of the track (i.e., the thing  
displayed in the QT Player) and not the text in each sample. If  
this is the case then we have to do something entirely different ;^}
d) all of the function calls succeed and when examined in the  
debugger you can see the values for size and chars[] are valid but  
your NSString cannot be properly constructed.


Can you clarify?  BTW, what values are you seeing in the debugger?

On Dec 14, 2008, at 8:33 PM, Joe Turner wrote:

Okay, so I did what you said, but yet, it still won't show me the  
track names. Here is my code:


Media media = [movieView movie]  
tracksOfMediaType:QTMediaTypeText] objectAtIndex:0] media]  
quickTimeMedia];


TimeValue64 timeValue = 0;
TimeValue64 duration = 0;
	GetMediaNextInterestingDecodeTime(media, nextTimeMediaSample +  
nextTimeEdgeOK, timeValue, fixed1, &timeValue, &duration);


ByteCount size;
	GetMediaSample2(media, NULL, 0, &size, timeValue, NULL, NULL,  
NULL, nil, nil, 1, NULL, NULL);


char chars[size];

	GetMediaSample2(media, &chars, size, NULL, timeValue, NULL,  
NULL, NULL, nil, nil, 1, NULL, NULL);
	NSString *string = [[NSString alloc] initWithBytes:chars  
length:size encoding:NSUTF8StringEncoding];

NSLog(@"String: %@", string);

Thanks,

Joe
On Dec 14, 2008, at 7:03 PM, douglas welton wrote:


Hi Joe,

Since you didn't define "easy" or tell me what you'd tried, then  
I'll have to make some assumptions about what you know.  I will  
assume that you have read the documentation on QT media samples.


The basic steps:

Get the text track.
Get the track media.
	loop thru the media samples using a function like  
GetMediaNextInterestingDecodeTime() with the parameters of your  
choice
		Use GetMediaSample2() to get the data/length associated with  
the text sample.
		Create an NSString with the bytes/length returned in the  
previous call


regards,

douglas

On Dec 14, 2008, at 7:26 PM, Joe Turner wrote:

I have a QTMovie (that is valid), and it has a text track. I am  
wondering if there is any easy way to get the contents of the  
text track into an NSString.


Thanks,

Joe










___

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 arch...@mail-archive.c

Re: toolbar to view gradient displaying correctly only sometimes?



On 21 Dec 2008, at 4:39 am, Chunk 1978 wrote:


is there a reason why my
dad's 20 alu iMac 2007 would be displaying this gradient improperly?



Many Macs have 6-bit LCD displays, making them incapable of displaying  
32-bit colour, which requires 8-bit pixels. Could that be it?


You might also find that settings such as the screen's gamma can  
strongly affect the appearance of gradients, etc. Try examining the  
settings in the Display System Prefs. In fact if your UI is sensitive  
to such things, it's probably best to try something else.


--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:
http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com

This email sent to arch...@mail-archive.com

Re: Using runtime functions to match up object types safely when setting properties



On 21 Dec 2008, at 4:07 am, Michael Ash wrote:


You could build a table, either manually or perhaps with the help of
some sort of crazy macro. But there is absolutely no way to get the
NSColor out of -(void)setColor:(NSColor*)c at runtime because that
info simply isn't there.



Yep, I've come to understand that.

It looks as if I'll have to extend the scheme where my objects publish  
the properties that are exposed in the UI to also include acceptable  
type information, so the UI has something to go on when disabling  
improper combinations. A bit of a pain but not all that unreasonable.


thanks everyone for helping out,

cheers, 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:
http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com

This email sent to arch...@mail-archive.com

Re: selectionShouldChangeInTableView called twice



On 21 Dec 2008, at 5:23 am, Andre Masse wrote:

I'm using a master/detail view in my application. If the user select  
another row in the master table and the detail has been modified, I  
want to present an alert to save, cancel or return to detail.



Just my 2¢ worth, but this sounds like a horrible UI. If I get  
interrupted by an alert every time I change a table selection, I'm  
going to get pretty irritated by that quite quickly. Normally table  
selections changes are considered "cheap", and the user shouldn't be  
punished for browsing.


Instead why not just keep a change flag for each modified object, then  
at some suitable time when the user is moving to a task that needs to  
use the modified data or else revert to the earlier state, check the  
flags and only present the alert once. This is how Mail's preferences  
settings work for example - I can change lots of different settings  
but only when I close the window am I prompted to save any changes I  
made. It's much less intrusive.


--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:
http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com

This email sent to arch...@mail-archive.com

Re: Newbie Question: implementing a stack

2008-12-20 Thread Steve Wetzel



On Dec 20, 2008, at Dec 20:12:08 AM, Graham Cox wrote:



On 20 Dec 2008, at 4:52 pm, Graham Cox wrote:



On 20 Dec 2008, at 3:15 pm, Steve Wetzel wrote:

Regarding memory management - does it make more sense to copy the  
object to be pushed from within the Stack object rather then  
copying it externally before the push call?  I am thinking that it  
does because then that object is encapsulated which is how a stack  
should really work.  If I do this I realize I will need to  
implement a copy method within any object that I want to place on  
the stack if it does not already have one.



My own view is that the stack object shouldn't copy the object.



Of course, a very easy, and generally useful solution, is to provide  
two methods:


- (void)push:(id) obj;
- (void)pushCopy:(id) obj;

G.


I do see your point Graham but what I am trying to understand is how  
to work with the stack if I don't copy the object to put on it.  If I  
simply push the pointer on the stack, it seems that I have to make a  
lot of objects in the code that handles the stack object.  If I have  
an class MyObject I wish to push on the stack:


MyObject *myObj1 = [[MyObject alloc] init];
MyObject *myObj2 = [[MyObject alloc] init];
...
MyObject *myObj10 = [[MyObject alloc] init];


[stack push:myObj1];
[stack push:myObj2];
...
[stack push:myObj10];

Each time I want to push a MyObject onto the stack I need to create a  
new MyObject.  If I copy I can do:


MyObject *myObj = [[MyObject alloc] init];

[stack push:myObj];

[stack push:myObj];

...


I guess can simply assign the pointer, but if I do, it seems to me I  
will need an NSMutableArray to hold myObj1... myObj10 (or more).  If I  
do that, what benefit is the stack?


Steve
___

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 arch...@mail-archive.com

Copying Managed Objects from App to Doc MOContext

2008-12-20 Thread Jerry Krinock

My app maintains in its managed object context an array of, say,  
Potato objects.  The potatoes in this central managed object context  
come and go occasionally -- someone might throw in a new one, or eat  
one.  An archived potato is low in calories, consisting of a half  
dozen numbers or short strings.


My persistent documents can also contain several potatoes.  To set  
one, user clicks a popup menu in the document window.  The popup is  
populated by the potatoes currently available in the central managed  
object context.


Because the potatoes in this popup are inserted into the central  
managed object context, however, I cannot simply "set" one of them  
into my document when the user makes their seledtion.  I implement a  
custom setter which inserts a ^new^ potato into the document's managed  
object context, and copies to it all of the attributes from the  
selected potato.


Looks OK to me, but having never seen anything like this, I was  
wondering if I'm seeing OK today.


Thank you,

Jerry Krinock


Here are a couple of (not yet fully tested) methods which can be  
implemented in a category or subclass of NSManagedObject to help with  
the all-attributes/value copying:



 .h *

/*!
 @briefA dictionary containing all attribute keys of the
 receiver for which current values are not nil, and their values.

 @details  Does not give relationships.  The output
 returned is mutable as a convenience because, in some
 applications, key/value pairs will be immediately
 added or removed.

 Tip: If you want to copy both attribues and properties, use
 -[NSKeyValueCodingProtocol dictionaryWithValuesForKeys:].
 */
- (NSMutableDictionary*)attributesDictionary ;

/*!
 @briefSets the receiver's attributes from a dictionary

 @details  All keys in 'attributes' are attempted
 to be set using setValue:forKey:.  Thus, setValue:forUndefinedKey:
 will be invoked (and possibly an exception raised) if 'attributes'
 contains keys for which the receiver is not KVC compliant.

 attributes The dictionary of attributes to be set.
 */
- (void)setAttributes:(NSDictionary*)attributes ;


 .m *

- (NSMutableDictionary*)attributesDictionary {
NSEntityDescription* entity = [self entity] ;
NSDictionary* attributes = [entity attributesByName] ;
/* The following one-liner will almost work:
 return [self dictionaryWithValuesForKeys:attributes] ;
 Unfortunately, instead of omitting pairs for which the values  
are nil,
 it inserts keys with values of NSNull.  So, I use my own  
implementation:

 */

NSMutableDictionary* attributesDic = [[NSMutableDictionary alloc]  
init] ;

for (id attributeName in attributes) {
id value = [self valueForKey:attributeName] ;
if (value) {

[attributesDic setValue:value
 forKey:attributeName] ;
}
}

return [attributesDic autorelease] ;
}

- (void)setAttributes:(NSDictionary*)attributes {
for (id attributeName in attributes) {
[self setValue:[attributes valueForKey:attributeName]
forKey:attributeName] ;
}
}


___

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 arch...@mail-archive.com

Re: selectionShouldChangeInTableView called twice

2008-12-20 Thread Rob Rix

I'm using a master/detail view in my application. If the user  
select another row in the master table and the detail has been  
modified, I want to present an alert to save, cancel or return to  
detail.



Just my 2¢ worth, but this sounds like a horrible UI. If I get  
interrupted by an alert every time I change a table selection, I'm  
going to get pretty irritated by that quite quickly. Normally table  
selections changes are considered "cheap", and the user shouldn't be  
punished for browsing.


Instead why not just keep a change flag for each modified object,  
then at some suitable time when the user is moving to a task that  
needs to use the modified data or else revert to the earlier state,  
check the flags and only present the alert once. This is how Mail's  
preferences settings work for example - I can change lots of  
different settings but only when I close the window am I prompted to  
save any changes I made. It's much less intrusive.


Not to be picky, but Mail’s accounts pane actually works the way he  
describes. If I select an account, uncheck “Include when automatically  
checking for new mail” and then select a different account, it asks me  
if I want to save the changes before updating the detail view with the  
contents of the newly-selected account.


So there is precedent, although it is a bit annoying.

Rob___

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 arch...@mail-archive.com

Re: Newbie Question: implementing a stack



On 21 Dec 2008, at 10:52 am, Steve Wetzel wrote:

I guess can simply assign the pointer, but if I do, it seems to me I  
will need an NSMutableArray to hold myObj1... myObj10 (or more).  If  
I do that, what benefit is the stack?



I don't really follow your argument. Presumably you decided you need a  
stack for some reason to do with your application's logic - so I can't  
answer "what benefit is the stack?" - that's up to you!


A stack in general terms doesn't inherently copy items placed on it,  
there's nothing in the definition of 'stack' that implies a copy. It's  
just storage, with a LIFO semantic.


I don't really see what creating objects has do with it - the stack is  
to do with organising objects, not creating them.


--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:
http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com

This email sent to arch...@mail-archive.com

Re: selectionShouldChangeInTableView called twice



On 21 Dec 2008, at 11:03 am, Rob Rix wrote:
Not to be picky, but Mail’s accounts pane actually works the way he  
describes. If I select an account, uncheck “Include when  
automatically checking for new mail” and then select a different  
account, it asks me if I want to save the changes before updating  
the detail view with the contents of the newly-selected account.


So there is precedent, although it is a bit annoying.

Rob



Yeah, maybe that's not the best example. In fact just trying it now it  
doesn't always even work - changing a minor setting such as time to  
leave mail on server, it causes the save changes sheet to show, then  
hitting Cancel doesn't revert the setting!


On the other hand buttons in a preferences panel are not quite the  
same thing as a selection in a table - the user is unlikely to expect  
(or forgive) a UI that interrupts them constantly on a mere selection  
change.


--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:
http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com

This email sent to arch...@mail-archive.com

(no subject)

2008-12-20 Thread John Nairn

I had a similar problem and there were serious performance issues  
with the built-in sorting. I could make it work with  
sortUsingFunction:context:, but the performance depends significantly  
on the efficiency of that method (i.e., how much work it takes to  
interact with that external data). In my case, when used for sorting  
a column with a few thousand lines, it got very slow. I was able to  
get more much faster sorting (at least 100x) by not relying on built- 
in tools.


If you are interested, here is the process I used:

1. Building a c++ array of integers with indices to the new sort order
2. Precalculate all crucial sort terms (that depend on external data)  
for the objects to be sorted.
3. Sort the index array manually (e.g., a shell sort from Numerical  
Recipes in C, pg 244 (my edition))
4. Once done, build a new array based on the sorted indices, and free  
any allocated memory


Here is partial code. You would have to fill in the objects array and  
the method (BOOL)checkObj1:(id)obje1 vsObj2:(id)obj2 simply returns  
YES if the objects need rearranging (i.e., currently in the wrong  
order).


// sorting objects in array order
int *ind=(int *)malloc(([order count]+1)*sizeof(int));
	NSMutableArray *objects=[[NSMutableArray alloc] initWithCapacity: 
[order count]];

for(i=0;i<[order count];i++)
{   ind[i+1]=i;
		// fill index i of objects with information needed to sort object  
at item i in order

}
int n2,n=[order count];
int lognb2=0,j,m,nn,ii,ij;
n2=n>>1;
while(n2>0)
{   lognb2++;
n2>>=1;
}
m=n;
for(nn=1;nn<=lognb2;nn++)
{   m>>=1;
for(j=m+1;j<=n;j++)
{   ii=j-m;
ij=ind[j];
			while(ii>=1 && [self checkObj1:[objects objectAtIndex:ind[ii]]  
vsObj2:[objects objectAtIndex:ij]])

{   ind[ii+m]=ind[ii];
ii-=m;
}
ind[ii+m]=ij;
}
}
	// build new array from sorted indices in ind[], i.e. index i of new  
array is index ind[i] from original array



It will all depend on how efficient you can make the function in the  
built-in sorting function. For me, the difference was dramatic.



See -sortUsingFunction:context: in NSMutableArrays documentation. As
Ricky Sharp has already mentioned, theres also -sortUsingSelector:,
however if youre comparing partly with external data you might like
having the void * context provided by this method.

Sincerely,
Rob

On 20-Dec-08, at 7:15 AM, Jacob Rhoden wrote:


How do you sort an NSMutableArray, when the difference between the
two objects is determined by information not contained completely
within the objects themselves. ie in Java you can implement a
Comparator that takes two objects to compare then. Is this similar
in Cocoa?

ie in this case I need to apply a mathematical algorithm using
information from each object to determine which one is greater than
the other.

Thanks!
Jacob


---
John Nairn
GEDitCOM - Genealogy Software for the Macintosh
http://www.geditcom.com


___

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 arch...@mail-archive.com

Re: selectionShouldChangeInTableView called twice

2008-12-20 Thread Jerry Krinock



On 2008 Dec, 20, at 10:23, Andre Masse wrote:

I'm using a master/detail view in my application. If the user select  
another row in the master table and the detail has been modified, I  
want to present an alert to save, cancel or return to detail.


Back to the original question, I always thought that the desired  
behavior in this case would be to simply set the change in the data  
model without asking.  You mentioned "save".  Indeed, I would not  
"save" in the sense of saveDocument:.  But like Graham said, a table  
entry is "cheap" [in user effort] to restore the old value if it was a  
mistake.  And the proper, easy way for the user to restore it would be  
-- drum roll -- Undo -- cowbell  :)


___

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 arch...@mail-archive.com

Re: Copying Managed Objects from App to Doc MOContext

On Sat, Dec 20, 2008 at 6:58 PM, Jerry Krinock  wrote:
> My app maintains in its managed object context an array of, say, Potato
> objects.  The potatoes in this central managed object context come and go
> occasionally -- someone might throw in a new one, or eat one.  An archived
> potato is low in calories, consisting of a half dozen numbers or short
> strings.

This doesn't make sense... you don't store arrays in managed object
contexts.  There's a set of Potato managed objects in the context,
though.  So did you mean that, or are you storing some sort of
ArrayOfPotatoes managed object somewhere?

> My persistent documents can also contain several potatoes.  To set one, user
> clicks a popup menu in the document window.  The popup is populated by the
> potatoes currently available in the central managed object context.

So, there's a separate app-wide persistent store, or just a separate MOC?

> Looks OK to me, but having never seen anything like this, I was wondering if
> I'm seeing OK today.

It depends on the semantics you're trying to achieve.  If what you
really want is to store a reference to the central Potato, then store
its UUID.  If you want to store a copy of the potato, keep doing what
you're doing.

--Kyle Sluder
___

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 arch...@mail-archive.com

Re: Optimizing NSRectFill

On Sat, Dec 20, 2008 at 4:47 PM, Jean-Daniel Dupas
 wrote:
> So I think that you're right (and that everything is synchronous) but I also
> think that the 'flush results to screen' blocks until the screen is ready
> and so limit the number of redraw to the refresh rate of the screen.
> If this is effectively the case, your timer trick will not be very helpful.

As long as the timer's time interval is less than the screen refresh
interval (and the technote says to simply assume 60Hz and that most
drawing doesn't need more than 30Hz, so that's easy enough to assure)
then it will work. And even if the timer is more, it will depend on
the behavior of the event loop. If the event loop is smart it will
process all queued events before running the next timer, in which case
the technique will work even for a zero-interval timer, but I don't
know if it actually works this way.

Mike
___

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 arch...@mail-archive.com

Re: Newbie Question: implementing a stack

On Sat, Dec 20, 2008 at 6:52 PM, Steve Wetzel  wrote:
> I do see your point Graham but what I am trying to understand is how to work
> with the stack if I don't copy the object to put on it.  If I simply push
> the pointer on the stack, it seems that I have to make a lot of objects in
> the code that handles the stack object.  If I have an class MyObject I wish
> to push on the stack:
>
> MyObject *myObj1 = [[MyObject alloc] init];
> MyObject *myObj2 = [[MyObject alloc] init];
> ...
> MyObject *myObj10 = [[MyObject alloc] init];
>
>
> [stack push:myObj1];
> [stack push:myObj2];
> ...
> [stack push:myObj10];
>
> Each time I want to push a MyObject onto the stack I need to create a new
> MyObject.  If I copy I can do:
>
> MyObject *myObj = [[MyObject alloc] init];
>
> [stack push:myObj];
> 
> [stack push:myObj];
> 
> ...
>
>
> I guess can simply assign the pointer, but if I do, it seems to me I will
> need an NSMutableArray to hold myObj1... myObj10 (or more).  If I do that,
> what benefit is the stack?

You're very confused. When you assign something new to myObj, you're
only affecting that one pointer. You don't affect anything else that
has a reference to the original object. For example:

NSString *str = @"hello";
[stack push:str];
str = @"world";
[stack push:str];

Your stack now contains @"hello" and @"world". It does not matter what
its copying behavior is, this is *always* true. (Try it with an
NSMutableArray and see; NSMutableArray does not copy the objects it
receives.) When you write str = @"world" you just put a new address
into the str pointer. But the stack isn't holding your str variable.
It's simply holding the value that str held originally. So there's no
problem.

What copying saves you from is stuff like this:

NSMutableString *str = [NSMutableString stringWithString:@"hello"];
[stack push:str];
[str setString:@"world"];
[stack push:str];

In this case the stack ends up holding two copies of a single mutable
string which contains the text @"world". Here we aren't mutating
"str", we're mutating the *object* that it points to. If the stack
copied the object then we wouldn't mutate the copy and so the
-setString: call wouldn't change its contents.

This usually isn't much of a problem. And to the extent that it is,
being able to push non-copyable objects or referenced mutable objects
onto a stack will generally far outweigh them. Follow Cocoa's lead:
with the exception of dictionary keys, Cocoa does not copy objects
that are put into collections.

Mike
___

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 arch...@mail-archive.com

Re: Copying Managed Objects from App to Doc MOContext

2008-12-20 Thread Jerry Krinock



On 2008 Dec, 20, at 17:55, Kyle Sluder wrote:


This doesn't make sense... you don't store arrays in managed object
contexts.


Whoops.  Indeed, that does not make sense.  I should have said that  
there are potatoes inserted into the app's central managed object  
context.  They're not in an array and not related to any other entity;  
they just hang loose in there, fetched when needed, which is not very  
often.


So, there's a separate app-wide persistent store, or just a separate  
MOC?


Both.  The app-wide MOC has its PSC set to the app-wide PSC, which in  
turn has a persistent store in the Application Support folder.


Looks OK to me, but having never seen anything like this, I was  
wondering if

I'm seeing OK today.


It depends on the semantics you're trying to achieve.  If what you
really want is to store a reference to the central Potato, then store
its UUID.  If you want to store a copy of the potato, keep doing what
you're doing.


By UUID, I believe you mean [[[self objectID] URIRepresentation]  
absoluteString].


I'd considered that, but since it's possible that the source potato in  
the central store might go away, indeed I want a copy.  I was thinking  
that, for portability and robustness, in most cases one would not want  
a document to reference an object in an app's Application Support.


Thank you, Kyle -- I'm starting to see a clearer picture what I'm  
doing now.


___

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 arch...@mail-archive.com

Re: NSTimer help

2008-12-20 Thread Rob Keniger



On 18/12/2008, at 2:20 PM, Andrew Merenbach wrote:

Is there a reason to use it in the case that one might instead use - 
isEqual: or, as was discussed in a recent thread, -isEqualToString:  
(which was stated, if I recall correctly, to be optimized for speed  
in the case of strings)?


Apologies once again for the tangent, but I am rather curious.  I am  
extremely grateful for any insight.



Xcode's autocomplete has sometimes popped up with -isEqualTo: for me,  
which I guess you might not notice if you're unfamiliar with Cocoa.


--
Rob Keniger



___

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 arch...@mail-archive.com

Re: selectionShouldChangeInTableView called twice



On 21 Dec 2008, at 12:26 pm, Jerry Krinock wrote:



On 2008 Dec, 20, at 10:23, Andre Masse wrote:

I'm using a master/detail view in my application. If the user  
select another row in the master table and the detail has been  
modified, I want to present an alert to save, cancel or return to  
detail.


Back to the original question, I always thought that the desired  
behavior in this case would be to simply set the change in the data  
model without asking.  You mentioned "save".  Indeed, I would not  
"save" in the sense of saveDocument:.  But like Graham said, a table  
entry is "cheap" [in user effort] to restore the old value if it was  
a mistake.  And the proper, easy way for the user to restore it  
would be -- drum roll -- Undo -- cowbell  :)



I agree with this.

If the UI is part of an inspector-type interface, i.e. modeless, then  
just make the change but do it undoably. Performing the Undo naturally  
updates any visible UI that matters (KVO/Bindings makes this trivial).


The main case where a Save/Cancel approach makes much sense is in a  
modal situation, in which case the OK/Cancel buttons already imply the  
answer - Save on OK, revert on Cancel. Or, if you prefer for a  
modeless dialog, present an alert on window close, though I dislike  
this myself - a modeless window might as well use Undo, unless it's  
editing a very specific aspect of the data model, such as user  
defaults, where it's not usually appropriate to use Undo. Bear in mind  
you can create an undo manager at any time.


--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:
http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com

This email sent to arch...@mail-archive.com

Re: selectionShouldChangeInTableView called twice


Thanks for your replies guys.

This application is a front end to a database with a lot of data  
entries and deals with legal data. There's a lot of validations going  
on and data integrity is vital. Normally, the user will save or cancel  
by clicking a button in the detail view and changes will be committed  
or rolled back to the database. The alert is for handling the cases  
where the user select another row in the master table after having  
done some modifications to a record and "forgot" to save or cancel  
(most document based application display a similar alert when the  
window is closed without saving). This case is exceptional. As an  
analogy, let's say the master table is a list of clients. A typical  
use case for data entry goes like this:


1- search for a client
2- select the "good" one from the list (if a match is found, if not  
add a new one)

3- make some modifications in the detail view
4- save
repeat

Suppose the user begin editing a client then decide this is not the  
record he wanted to modify and simply select another row in the master  
table. Or, the master table contains the next client he has to  
modified and after finishing editing the first, he simply select the  
next. How can I know what to do without asking him what he want? In  
the first case I should cancel the record and in the second I should  
save it. Now, suppose the user somehow move the focus from the detail  
view (by switching application and coming back, accidently clicking in  
the master view or whatever). This is a third case, he may have not  
completed its modifications and should be able to continue editing the  
record.


I agree that this is a restrictive approach and I'll think how I could  
handle all this with Undo. This has a lot of implications on my  
application design which rely on being in a database transaction for  
editing, will never be sold as it's an in house application for 3  
users (me, my wife and my son) and has been working this way for years  
using 4th Dimension on a PC...


Food for thoughts :-)

Thanks,

Andre Masse



On Dec 20, 2008, at 20:26, Jerry Krinock wrote:



On 2008 Dec, 20, at 10:23, Andre Masse wrote:

I'm using a master/detail view in my application. If the user  
select another row in the master table and the detail has been  
modified, I want to present an alert to save, cancel or return to  
detail.


Back to the original question, I always thought that the desired  
behavior in this case would be to simply set the change in the data  
model without asking.  You mentioned "save".  Indeed, I would not  
"save" in the sense of saveDocument:.  But like Graham said, a table  
entry is "cheap" [in user effort] to restore the old value if it was  
a mistake.  And the proper, easy way for the user to restore it  
would be -- drum roll -- Undo -- cowbell  :)

___

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 arch...@mail-archive.com

Re: Copying Managed Objects from App to Doc MOContext

2008-12-20 Thread Dave Fernandes

I am doing something very similar to you. I have an application-based  
library of potatoes, and each document has a copy of some of those  
potatoes. I did this because the document may be opened on another  
computer that does not have access to the same library (or as you  
mention, the library can change over time, but the document should  
not change unless someone has explicitly changed it.)


An added complication in my case is that when one opens the document,  
I need to compare the document's potatoes to the library's potatoes  
and give the user the opportunity to update the document, the  
library, or both if they are not in sync. This means comparing all  
the potato attributes for each potato in the document against those  
in the library. This can be very time consuming if it is not  
optimized well, but it is certainly doable in Core Data.


Another set of complications is: What happens if the user changes the  
application library while a document is open? Will they expect the  
document to be automatically updated? If so, should the document be  
updated immediately, or only after saving changes to the application  
library?


I came up with a concept of "live" versus "archived" documents. If a  
document is new or dirty, then it is live and it updates its potatoes  
in sync with the persistent store of the application library. If the  
document is archived (i.e. it has been opened, but not touched), then  
it ignores changes to the library.


But all this should be based on how your users will expect the  
application to behave.


Dave

On Dec 20, 2008, at 9:29 PM, Jerry Krinock wrote:



On 2008 Dec, 20, at 17:55, Kyle Sluder wrote:


This doesn't make sense... you don't store arrays in managed object
contexts.


Whoops.  Indeed, that does not make sense.  I should have said that  
there are potatoes inserted into the app's central managed object  
context.  They're not in an array and not related to any other  
entity; they just hang loose in there, fetched when needed, which  
is not very often.


So, there's a separate app-wide persistent store, or just a  
separate MOC?


Both.  The app-wide MOC has its PSC set to the app-wide PSC, which  
in turn has a persistent store in the Application Support folder.


Looks OK to me, but having never seen anything like this, I was  
wondering if

I'm seeing OK today.


It depends on the semantics you're trying to achieve.  If what you
really want is to store a reference to the central Potato, then store
its UUID.  If you want to store a copy of the potato, keep doing what
you're doing.


By UUID, I believe you mean [[[self objectID] URIRepresentation]  
absoluteString].


I'd considered that, but since it's possible that the source potato  
in the central store might go away, indeed I want a copy.  I was  
thinking that, for portability and robustness, in most cases one  
would not want a document to reference an object in an app's  
Application Support.


Thank you, Kyle -- I'm starting to see a clearer picture what I'm  
doing now.


___

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/dave.fernandes% 
40utoronto.ca


This email sent to dave.fernan...@utoronto.ca


___

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 arch...@mail-archive.com

Re: selectionShouldChangeInTableView called twice


On Dec 20, 2008, at 22:00, Graham Cox wrote:

If the UI is part of an inspector-type interface, i.e. modeless,  
then just make the change but do it undoably. Performing the Undo  
naturally updates any visible UI that matters (KVO/Bindings makes  
this trivial).


The main case where a Save/Cancel approach makes much sense is in a  
modal situation, in which case the OK/Cancel buttons already imply  
the answer - Save on OK, revert on Cancel. Or, if you prefer for a  
modeless dialog, present an alert on window close, though I dislike  
this myself - a modeless window might as well use Undo, unless it's  
editing a very specific aspect of the data model, such as user  
defaults, where it's not usually appropriate to use Undo. Bear in  
mind you can create an undo manager at any time.


The UI is using a single window (per database relations) with the  
master table on the left and detail on the right. I've struggle a lot  
with this as the detail view in the principal table is pretty complex:  
25 fields (although many are generated) and 2 table views (to many  
relationships). In a typical session, the window will be closed at the  
end of the day. In the present application (built with 4th Dimension)  
the window contains the list and users must double-clic on a row to  
edit a record. Since they often need to check the detail view for a  
couple of rows before making a decision, I thought it would be great  
to use a master/detail design. This has some side effects though,  
hence my reason for using the alert...


I will think on the implications of using Undo as there's some limit  
to this design. The first thing I must resolve is Undoing changes  
hours or days after entries... Using (database row) versions is  
expansive as there's more than a million rows spread over many  
database tables. Not counting the legal implications when saving data  
to the wrong "client". The present design doesn't always prevent this  
but gives a more immediate feedback to the user, as "annoying" as it  
may sounds...


Thanks for the comments, I will sleep on this :-)

Andre Masse
___

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 arch...@mail-archive.com

Re: Copying Managed Objects from App to Doc MOContext