1. cocoa@apple-dev.groups.io
  2. Messages
   Date   
Re: Verbose warning splurged to the log with Secure Coding (dearchiving) -- can they be serious?

Graham Cox
 

This sounds so right. But unfortunately doesn’t work for __NSCFConstantString, which returns that, and not NSString. I haven’t yet tried on other classes, but most of the time it’ll be a string...

In respect of this particular issue, I think if you call `classForCoder` rather than `class` on the object of interest, you'll get the portable class name ("NSString") you're looking for.

-ben
Re: Verbose warning splurged to the log with Secure Coding (dearchiving) -- can they be serious?

Quincey Morris
 

I think you’re Doing It Wrong™. Your solution is effectively disabling secure coding, which solves your run-time error, but leaves you with the exact security vulnerability that secure coding is supposed to protect against.

Your goal of “I want it to be as flexible as possible” without specifying classes in advance isn’t compatible with secure decoding. 

Let’s stick with your NSDictionary example. There are 3 cases:

1. You (try to) decode the dictionary and specify class NSDictionary. This is insecure, because decoding the dictionary also decodes arbitrary classes contained in the dictionary.

2. The dictionary contains only objects of “plist” classes such as NSNumber, NSDate and standard NS collection classes. There are a finite number of these, so you can always list them all in `decodeObjectOfClasses:forKey:`, and this kind of dictionary is useful for a lot of things without getting into custom classes.

3. The dictionary contains your custom classes and/or client code's custom classes as well. Again, you must use `decodeObjectOfClasses:forKey:` to specify *all* of the classes that might be in the dictionary, but you don’t know them all at compile time.

You can’t do #1, so you must do #2 or #3. Either way, use `decodeObjectOfClasses:forKey:` for a heterogenous collection or hierarchy.

Note that your set of classes for `decodeObjectOfClasses:forKey:` can be built at run-time, but you should avoid the temptation to encode that set into the archive, since that also compromises security. (An attacker can simply ask your code to unarchive data containing malicious classes, and naming their classes!)

If you want to make things flexible for client code, I guess the way to do it would be to have the client tell you, on unarchiving, the list of its custom classes that it allows in the archive. You’d augment that set with the additional classes your own wrapper code needs, and specify that for `decodeObjectOfClasses:forKey:`.

On Feb 24, 2022, at 03:35, Graham Cox <graham@...> wrote:

Well, I figured out a solution, though I’m not sure — it seems slightly smelly, even though it looks neat from the outside and is easy to use. Would appreciate any feedback.

Basically, I walk up the class hierarchy until I hit NSObject, then return the class once removed from that. Then I added a category on NSCoder that uses this to encode that class name in the archive. Using these methods in NSCoding is easy and hides the minor pain involved.
Re: Verbose warning splurged to the log with Secure Coding (dearchiving) -- can they be serious?

Ben Kennedy
 

On 24 Feb 2022, at 2:07 am, Graham Cox <graham@mapdiva.com> wrote:

The actual class is __NSCFConstantString, rather than just a generic NSString, that seems to be what it is complaining about — if I change it to use [NSString class] it’s fine.

Since the archive originally wrote the result of NSStringFromClass(), it will of course write the explicit class rather than the more general umbrella class of a cluster. Can you think of a simple solution for this, in general?
In respect of this particular issue, I think if you call `classForCoder` rather than `class` on the object of interest, you'll get the portable class name ("NSString") you're looking for.

-ben
Re: Verbose warning splurged to the log with Secure Coding (dearchiving) -- can they be serious?

Graham Cox
 

Well, I figured out a solution, though I’m not sure — it seems slightly smelly, even though it looks neat from the outside and is easy to use. Would appreciate any feedback.

Basically, I walk up the class hierarchy until I hit NSObject, then return the class once removed from that. Then I added a category on NSCoder that uses this to encode that class name in the archive. Using these methods in NSCoding is easy and hides the minor pain involved.




Class GCLowestSpecificAncestorClass( Class aClass )
{
// walks back up the class hierarchy from <aClass> until it reaches NSObject (or nil), and returns the
// class that subclasses NSObject.


Class sc = [aClass superclass];


while( sc && sc != [NSObject class])
{
aClass = sc;
sc = [aClass superclass];
}


return aClass;
}



@implementation NSCoder (GCSecureAnonymousClassSupport)

- (void) encodeAnonymousObject:(id) theObject forKey:(NSString*) key
{
Class umbrella = GCLowestSpecificAncestorClass([theObject class]);


NSAssert( umbrella != [NSObject class], @"cannot archive NSObject as anonymous class");


NSString* className = NSStringFromClass( umbrella );
[self encodeObject:className forKey:[NSString stringWithFormat:@"class_for_%@", key]];
[self encodeObject:theObject forKey:key];
}



- (id) decodeAnonymousObjectForKey:(NSString*) key
{
NSString* className = [self decodeObjectOfClass:[NSString class] forKey:[NSString stringWithFormat:@"class_for_%@", key]];
Class umbrella = NSClassFromString( className );


return [self decodeObjectOfClass:umbrella forKey:key];
}



@end








On 24 Feb 2022, at 9:07 pm, Graham Cox <graham@...> wrote:

Hi Glenn,

This seemed like a perfectly fine idea in this case, so I tried it.

Now I get a new message spewed to the log:

2022-02-24 20:50:13.158117+1100 GCSimpleContainers[26354:31119213] [general] *** -[NSKeyedUnarchiver validateAllowedClass:forKey:] allowed unarchiving safe plist type ''NSString' (0x7ff85c670920) [/System/Library/Frameworks/Foundation.framework]' for key 'data', even though it was not explicitly included in the client allowed classes set: '{(
    "'__NSCFConstantString' (0x7ff85c645908) [/System/Library/Frameworks/CoreFoundation.framework]"
)}'. This will be disallowed in the future.


This happens for the line that decodes the payload object, which happens to be a string. My code is:

Class dataClass = NSClassFromString([coder decodeObjectOfClass:[NSString class] forKey:@"data_class"]); //<—— this is OK
self.data = [coder decodeObjectOfClass:dataClass forKey:@"data”]; //<— this is not

The actual class is __NSCFConstantString, rather than just a generic NSString, that seems to be what it is complaining about — if I change it to use [NSString class] it’s fine.

Since the archive originally wrote the result of NSStringFromClass(), it will of course write the explicit class rather than the more general umbrella class of a cluster. Can you think of a simple solution for this, in general?


—G.




On 24 Feb 2022, at 6:44 pm, Glenn L. Austin <glenn@...> wrote:

On Feb 23, 2022, at 11:30 PM, Graham Cox <graham@...> wrote:

Hi all,

I was working on some demo code investigating the use of secure archiving, since I guess we all have to adopt that if we can. I have a simple container class (a linked list) that I decided to make secure archivable. It works, but when dearchiving, I need to dearchive the ‘payload’ of the list nodes, which can be anything at all - it’s defined as an id<NSObject>. So when I ask for this item securely, I use

-decodeObjectOfClass:[NSObject class] forKey:<blah blah>

A suggestion? Archive the class name as a part of your archiving process, then get the class instance by NSClassFromString. You can then use that as the parameter for decodeObjectOfClass:forKey:.

It's not entirely secure, it's not the best style to avoid the warning, but it should prevent the spew into the system log. Also, if you can't get the class from the class name, you know you'll not be able to decode the value anyway.

-- 
Glenn L. Austin, Computer Wizard and Race Car Driver         <><
<http://www.austinsoft.com>
Re: Verbose warning splurged to the log with Secure Coding (dearchiving) -- can they be serious?

Graham Cox
 

Hi Glenn,

This seemed like a perfectly fine idea in this case, so I tried it.

Now I get a new message spewed to the log:

2022-02-24 20:50:13.158117+1100 GCSimpleContainers[26354:31119213] [general] *** -[NSKeyedUnarchiver validateAllowedClass:forKey:] allowed unarchiving safe plist type ''NSString' (0x7ff85c670920) [/System/Library/Frameworks/Foundation.framework]' for key 'data', even though it was not explicitly included in the client allowed classes set: '{(
    "'__NSCFConstantString' (0x7ff85c645908) [/System/Library/Frameworks/CoreFoundation.framework]"
)}'. This will be disallowed in the future.


This happens for the line that decodes the payload object, which happens to be a string. My code is:

Class dataClass = NSClassFromString([coder decodeObjectOfClass:[NSString class] forKey:@"data_class"]); //<—— this is OK
self.data = [coder decodeObjectOfClass:dataClass forKey:@"data”]; //<— this is not

The actual class is __NSCFConstantString, rather than just a generic NSString, that seems to be what it is complaining about — if I change it to use [NSString class] it’s fine.

Since the archive originally wrote the result of NSStringFromClass(), it will of course write the explicit class rather than the more general umbrella class of a cluster. Can you think of a simple solution for this, in general?


—G.




On 24 Feb 2022, at 6:44 pm, Glenn L. Austin <glenn@...> wrote:

On Feb 23, 2022, at 11:30 PM, Graham Cox <graham@...> wrote:

Hi all,

I was working on some demo code investigating the use of secure archiving, since I guess we all have to adopt that if we can. I have a simple container class (a linked list) that I decided to make secure archivable. It works, but when dearchiving, I need to dearchive the ‘payload’ of the list nodes, which can be anything at all - it’s defined as an id<NSObject>. So when I ask for this item securely, I use

-decodeObjectOfClass:[NSObject class] forKey:<blah blah>

A suggestion? Archive the class name as a part of your archiving process, then get the class instance by NSClassFromString. You can then use that as the parameter for decodeObjectOfClass:forKey:.

It's not entirely secure, it's not the best style to avoid the warning, but it should prevent the spew into the system log. Also, if you can't get the class from the class name, you know you'll not be able to decode the value anyway.

-- 
Glenn L. Austin, Computer Wizard and Race Car Driver         <><
<http://www.austinsoft.com>
Re: Verbose warning splurged to the log with Secure Coding (dearchiving) -- can they be serious?

Glenn L. Austin
 

On Feb 23, 2022, at 11:30 PM, Graham Cox <graham@...> wrote:

Hi all,

I was working on some demo code investigating the use of secure archiving, since I guess we all have to adopt that if we can. I have a simple container class (a linked list) that I decided to make secure archivable. It works, but when dearchiving, I need to dearchive the ‘payload’ of the list nodes, which can be anything at all - it’s defined as an id<NSObject>. So when I ask for this item securely, I use

-decodeObjectOfClass:[NSObject class] forKey:<blah blah>

A suggestion? Archive the class name as a part of your archiving process, then get the class instance by NSClassFromString. You can then use that as the parameter for decodeObjectOfClass:forKey:.

It's not entirely secure, it's not the best style to avoid the warning, but it should prevent the spew into the system log. Also, if you can't get the class from the class name, you know you'll not be able to decode the value anyway.

-- 
Glenn L. Austin, Computer Wizard and Race Car Driver         <><
<http://www.austinsoft.com>
Verbose warning splurged to the log with Secure Coding (dearchiving) -- can they be serious?

Graham Cox
 

Hi all,

I was working on some demo code investigating the use of secure archiving, since I guess we all have to adopt that if we can. I have a simple container class (a linked list) that I decided to make secure archivable. It works, but when dearchiving, I need to dearchive the ‘payload’ of the list nodes, which can be anything at all - it’s defined as an id<NSObject>. So when I ask for this item securely, I use

-decodeObjectOfClass:[NSObject class] forKey:<blah blah>

But the dearchiver spits this to the log:

2022-02-24 17:59:55.701579+1100 GCSimpleContainers[25591:31009796] [general] *** -[NSKeyedUnarchiver validateAllowedClass:forKey:]: NSSecureCoding allowed classes list contains [NSObject class], which bypasses security by allowing any Objective-C class to be implicitly decoded. Consider reducing the scope of allowed classes during decoding by listing only the classes you expect to decode, or a more specific base class than NSObject. This will become an error in the future. Allowed class list: {(
    "'NSObject' (0x7ff85c7c2b88) [/usr/lib]"
)}


I get what it’s saying — that asking for NSObject is too open-ended, and it renders the secure coding moot. But what else can I do? I don’t know the payload class, I want it to be as flexible as possible. Forcing this to be a subclass of NSObject of my own means I lose the ability to store all standard Cocoa objects ‘as is’ in my container. Adding a wrapper class doesn’t help because it would still need to archive what it wrapped. It strikes me that Apple themselves must run into this when archiving NSDictionary, since keys can be anything, but that works, so perhaps they use an internal bit of trickery.

It concerns me that they say “this will become an error in the future”. I can go back to traditional archiving with no security, no errors there, but then what’s the point of secure coding?

A way to simply suppress the log of the message would do for now. It’s extra bad because it logs this every time it happens, not just once. Is this a case of Apple being too zealous in overlooking a common use-case, or have I missed something?

—G
UICollectionView with DiffableDataSource

Rick Aurbach
 

Xcode 13.1, iOS 25 SDK
 
I have a UICollectionView (with a custom UICollectionViewLayout) which I am updating with a DiffableDataSource. This CollectionView displays a particular data object (with a given number of sections and items).
 
The user can choose to display a different object (with a different number of sections and items), reusing this CollectionView. When this occurs, I do the following:
 
    (pass the new data object to the custom layout)
    let layout = collectionView.collectionViewLayout
    layout.invalidateLayout()
    collectionView.contentSize = layout.collectionViewContentSize
    (calculate a new snapshot and apply it)
 
When the new data object has FEWER sections and/or items as the first, I get a series of warning messages (one for each of the cells that are no longer in the collection view) as follows:
 
    2021-12-07 14:29:02.239301-0600 WineCorner[82916:1921357] [CollectionView] Layout attributes <UICollectionViewLayoutAttributes: 0x7f77b3047e00> index path: (<NSIndexPath: 0xa0f0b1eb018e754a> {length = 2, path = 0 - 4}); frame = (578.118 6; 160 160); transform = [0.70710678118654757, -0.70710678118654746, 0.70710678118654746, 0.70710678118654757, 0, 0]; zIndex = 1;  were received from the layout <WineCorner.WineRackLayout: 0x7f77b1f1a0c0> but are not valid for the data source counts. Attributes will be ignored.
 
Obviously, I am not handling the situation correctly. What should I do so that these warning messages are not issued?
NSTextView: displaying "virtual" symbols?

David Young
 

[I'd sent this to cocoa-dev@lists.apple.com, but ISTR that list is
deprecated. Resending here.]

I am working on a document storage format that consists of logical zones
that nest. You can think of them sort of like <div> elements from
(X)HTML:

<div id="foo"><div id="bar">All work and no play.</div></div>

Ordinarily, I would like for the zone boundaries marked by <div> and
</div> to be invisible to the user. Ordinary cursor motion commands
should step from character to character, ignoring the <div> boundaries.

However, I also want to offer a display & editing mode to the user where
they can see the boundaries and use special cursor motion commands to
"creep" the cursor across the boundaries. In this mode, there are
additional locations addressable by the cursor. I have marked those
locations with a pipe (|):

|<div id="foo">|<div id="bar">All work and no play.</div>|</div>|

I took the attached screenshot in Numbers. It shows roughly the visual
effect I'm looking for. Incidentally, it's possible to move the cursor
back and forth over the crescent that "closes" the AVERAGE function
cell. That's good. It is also possible to delete the crescent, which
is neither necessary nor desirable for my purposes, but I guess that I
could accept it.

It would be nice if I could accomplish this effect without introducing
characters to my NSTextStorage.

It looks like TextKit 2 lets one do something like this rather easily,
but I'm leary of adopting it while it's so green.

Is anyone aware of a "worked example" I can read for TextKit 1?

Dave

--
David Young
dyoung@pobox.com Urbana, IL (217) 721-9981
Re: Scanner oddity

Sandor Szatmari
 

Laurent,

On Sep 29, 2021, at 15:11, Laurent Daudelin <laurent@...> wrote:

Anybody uses the Scanner class?

I’m using an instance to extract some data from a string received from a web server.

If I have a string that is “45\nBlah”, read the “45”, the

Have you tried -setCharactersToBeSkipped:

Set that to skip new line, carriage return… 

Maybe that will work for you?

Sandor

scanner location is just before the carriage return. If I then try a someString = scanner.scanString(“\nBlah”), someString is nil. But if I do scanner.scanLocation += 1, then someString = scanner.scanString(“Blah”), then someString contains “Blah”. Not sure if I’m using it correctly or if there is some kind of a bug.

By the way, speaking of “scanLocation”, the editor gives me a warning that it’s been deprecated as of iOS 13. But when I open the documentation about the Scanner, there is no mention that it’s been deprecated. Further, the editor doesn’t suggest a replacement.

What am I missing?

-Laurent.
-- 
Laurent Daudelin   laurent@...
Skype: LaurentDaudelin
Logiciels Némésys Software   http://www.nemesys-soft.com/
Scanner oddity

Laurent Daudelin
 

Anybody uses the Scanner class?

I’m using an instance to extract some data from a string received from a web server.

If I have a string that is “45\nBlah”, read the “45”, the scanner location is just before the carriage return. If I then try a someString = scanner.scanString(“\nBlah”), someString is nil. But if I do scanner.scanLocation += 1, then someString = scanner.scanString(“Blah”), then someString contains “Blah”. Not sure if I’m using it correctly or if there is some kind of a bug.

By the way, speaking of “scanLocation”, the editor gives me a warning that it’s been deprecated as of iOS 13. But when I open the documentation about the Scanner, there is no mention that it’s been deprecated. Further, the editor doesn’t suggest a replacement.

What am I missing?

-Laurent.
-- 
Laurent Daudelin   laurent@...
Skype: LaurentDaudelin
Logiciels Némésys Software   http://www.nemesys-soft.com/
How to embed Swift Package executable into macOS app

Jim
 

I have a macOS app I’d like to embed an executable from a swift package. However, while I can add the swift package under the app target’s “Frameworks, Libraries, and Embedded Content” section, and add it to Dependencies in Build Phases, it doesn’t show up as a project resource in the “Copy Bundle Resources” section. Is there some step I’m missing?

Also, I had accidentally dragged the executable to my project and added that in the “Copy Bundle Resources” section, but when I tried to notarize it refused because the executable wasn’t built with hardened runtime. Is it possible to build a command-line executable from a Swift package with hardened runtime?

This is all pointing me back to how I had first added this, with a separate target and scheme in the main project for the command line tool. However, that also added the command line tool as a product, and even though I turned off “Archive” in the scheme for the command line tool it was still archived as a separate product. This archived as a folder with both products and I wasn’t able to use Xcode’s built-in notarization process. I can notarize outside of Xcode if I have to but it doesn’t seem like this should be so difficult. Also I’d prefer to have the command line tool in its own project if possible.

Jim Crate
Re: Async/Await and Model Dialogs

David M. Cotter
 

also you can enlargen your screen resolution:

apple->system prefs->displays->display
resolution: scaled, click "more space" on the far right

-dave

On Sep 21, 2021, at 2:27 PM, Rick Aurbach via groups.io <rlaurb@...> wrote:

In a UIKit context, has anyone had experience/success in using async/await to synchronize a modal dialog with other logic? I've tried it a bit without success.

I.e, given a presented dialog, I want to capture data in the dialog, then use the results in a simple, linear fashion. (Something that looks like "Present the dialog, wait for results, use results" -- all inline.) It seems to me that async/await with @MainActor ought to make that possible, but  I haven't yet figured out how. I'd really like to see a real-world example.

Can you help?

Async/Await and Model Dialogs

Rick Aurbach
 

In a UIKit context, has anyone had experience/success in using async/await to synchronize a modal dialog with other logic? I've tried it a bit without success.

I.e, given a presented dialog, I want to capture data in the dialog, then use the results in a simple, linear fashion. (Something that looks like "Present the dialog, wait for results, use results" -- all inline.) It seems to me that async/await with @MainActor ought to make that possible, but  I haven't yet figured out how. I'd really like to see a real-world example.

Can you help?
Re: Just a little sharing - it's nice when your product finally hits the market.

Alex Zavatone
 

I don’t know. I was at MorphoTrust in 2014 when we worked on what would become the Apple Drivers’ Licence app. I did the first iOS working prototype with Dan, Ted and Clayton. Our PM was Steve Miu. Since then, MorphoTrust has gone through a few corporate hands. It was even owned by a French company for a while. Not sure who owns it now. It took forever to get the product picked up from what I know. Would have loved to be still working on it!

Cheers!
Alex Zavatone

On Sep 1, 2021, at 3:14 PM, Jack Brindle via groups.io <jackbrindle=me.com@groups.io> wrote:

Louisiana has had a driver’s license app for several years. Is that your work, Alex?

Jack

On Sep 1, 2021, at 2:32 PM, Alex Zavatone via groups.io <zav=mac.com@groups.io> wrote:

My olden app has finally hit the market.

https://www.macrumors.com/2021/09/01/apple-shares-us-states-adopting-wallet-ids/

It’s only taken 7 years since I worked on it. : )

Cheers to everyone else who was able to bring it to market.

Best,
Alex Zavatone
Re: Just a little sharing - it's nice when your product finally hits the market.

Jack Brindle
 

Louisiana has had a driver’s license app for several years. Is that your work, Alex?

Jack

On Sep 1, 2021, at 2:32 PM, Alex Zavatone via groups.io <zav=mac.com@groups.io> wrote:

My olden app has finally hit the market.

https://www.macrumors.com/2021/09/01/apple-shares-us-states-adopting-wallet-ids/

It’s only taken 7 years since I worked on it. : )

Cheers to everyone else who was able to bring it to market.

Best,
Alex Zavatone
Just a little sharing - it's nice when your product finally hits the market.

Alex Zavatone
 

My olden app has finally hit the market.

https://www.macrumors.com/2021/09/01/apple-shares-us-states-adopting-wallet-ids/

It’s only taken 7 years since I worked on it. : )

Cheers to everyone else who was able to bring it to market.

Best,
Alex Zavatone
WKWebKit and iOS 13 in Xcode 12

Alex Zavatone
 

Has anyone found a workaround for the crash of WKWebKit under the above configuration?  Xcode reposts “Error acquiring assertion” trying to display any webpage.

A workaround to prevent the screen from blanking out is this, but it doesn’t stop WKWebKit from crashing.

    override func willMove(toParent: UIViewController? ) {
        print("Something")
    }

Basically, loading any webpage crashes out in WebKit, iOS 13 and Xcode 12.  It’s fine in iOS 14.

Thanks in advance.
Alex Zavatone
Re: New syntax in Xcode 12.5.

Alex Zavatone
 

I found out what the issue was.  The programmer didn’t add a self in front of variables that had the same name as the local that he was creating.   Xcode 12.4 reports an error until you add self in front of the properties.  Xcode 12.5 can handle 

The actual line looked more like this.

let myVar = myVar != someValue ? myVar : myOtherAssignment

It actually should be    

let myVar = self.myVar != someValue ? self.myVar : myOtherAssignment

Thanks.
Alex Zavatone

On Jul 26, 2021, at 11:19 AM, Jeremy Hughes via groups.io <moon.rabbit@...> wrote:

1. myVar != someValue is the condition.

2. It determines whether the ternary expression evaluates to resultOne or resultTwo

3. resultOne or resultTwo is what is assigned in let myVar =

This seems like a standard use of the ternary conditional operator. I don’t have any problems compiling it in Xcode 12.4.

Jeremy

—

On 26 Jul 2021, at 16:57, Alex Zavatone via groups.io <zav@...> wrote:

The first half.  Ternary has been there since time began.  I’ve never seen this part be legitimate before.

let myVar = myVar != someValue



On Jul 26, 2021, at 10:09 AM, Jeremy Hughes via groups.io <moon.rabbit@...> wrote:

Are you referring to the ternary conditional operator, which has existed in Swift since the beginning, or to the fact that it creates a local variable with the same name (myVar) as a parameter or member (myVar)?

Jeremy

—

On 26 Jul 2021, at 15:37, Alex Zavatone via groups.io <zav@...> wrote:

I just saw someone use this syntax in a project and haven’t see it before.  It compiles in Xcode 12.5, but not in Xcode 12.4.  Does anyone have any more details on it?

let myVar = myVar != someValue ? resultOne : resultTwo

Swift sure is getting muddier and muddier with regards to visual comprehension.

Thanks in advance.
Alex Zavatone
Re: New syntax in Xcode 12.5.

Jeremy Hughes
 

1. myVar != someValue is the condition.

2. It determines whether the ternary expression evaluates to resultOne or resultTwo

3. resultOne or resultTwo is what is assigned in let myVar =

This seems like a standard use of the ternary conditional operator. I don’t have any problems compiling it in Xcode 12.4.

Jeremy

—

On 26 Jul 2021, at 16:57, Alex Zavatone via groups.io <zav=mac.com@groups.io> wrote:

The first half. Ternary has been there since time began. I’ve never seen this part be legitimate before.

let myVar = myVar != someValue


On Jul 26, 2021, at 10:09 AM, Jeremy Hughes via groups.io <moon.rabbit=virginmedia.com@groups.io> wrote:

Are you referring to the ternary conditional operator, which has existed in Swift since the beginning, or to the fact that it creates a local variable with the same name (myVar) as a parameter or member (myVar)?

Jeremy

—

On 26 Jul 2021, at 15:37, Alex Zavatone via groups.io <zav=mac.com@groups.io> wrote:

I just saw someone use this syntax in a project and haven’t see it before. It compiles in Xcode 12.5, but not in Xcode 12.4. Does anyone have any more details on it?

let myVar = myVar != someValue ? resultOne : resultTwo

Swift sure is getting muddier and muddier with regards to visual comprehension.

Thanks in advance.
Alex Zavatone
21 - 40 of 1454