When I go to type a message into my JSQMessagesViewController, the view is not being pushed up when the keyboard appears.
I can't send a message and I can't see the message I'm typing.
I don't even see any code in my VC that would be causing this / that would have changed this so I don't know where to start looking or what code to supply. I've downloaded examples using JSQMessages and I feel like this is just built in functionality behind the scenes.
Any ideas?
If you have viewDidAppear/viewDidDissapear being used you have to use
super.viewDidAppear(true)
and
super.viewDidDissapear(true)
also called respectively.
Adding those in fixed the issue.
Related
I've got a TableViewController with static table; one of it's cells houses an UIView named graphArea. The view renders a chart, it's background and an axis line - all inside it's drawRect(). There are also two another views (sunView & markerView), that are made with Interface Builder and used for chart dynamics (moving marker line and point on touch events).
All worked buttery smooth until I've implemented and tried to test a share button, that employs the ordinary activityViewController mechanism.
The magic begins, when one from a couple of share activities, whose share plugin window takes the full screen, is finished (no matter whether sharing succeeded or cancelled). The app crashes.
Discovery using debugger made apparent to me, that the crash happens, because some views, including graphArea, sunView, markerView are nil after return from sharing screen.
Only some of fullscreen share plugins (like preinstalled Mail and Messages, or, in my case, "Download to DropBox" action) lead app to crash. Other fullscreen share plugins do not (tested Telegram, WhatsApp, Skype). No one of those non-fullscreen plugins has ever caused crash (Evernote, Twitter, 2Do etc.).
It looks like graphArea, sunView, markerView are deallocated from memory when "malicious" share plugins take full screen. I haven't figured out, why.
Here's some debug info:
The traceback and assembly of fatalErrorMessage.
The next screenshot shows a part of controller code and properties, that are nil on return from share plugin (gray selection). And yes, they were all non-nil before.
Please, help me! Thank you in advance!
Thank you, Palpatim. My friend also pointed me at the same thing: I've put graphArea.removeFromSuperview() in viewDidDisappear(), and this caused the exception after share plugins, that have .presentationStyle = fullScreen. So at the point, when the app is to show again, there is no more graphArea in on the tableView.
I have been working on this app for months now and from as far back as I can remember I have never had an issue with segues. The code is unchanged in terms of calling performSegueWithIdentifier but since my recent update to Xcode 7 and iOS 9 I have not been able to tack this issue.
I have tried:
Deleting button and creating new button w/ segue link
Using a direct segue from button to view, without the use of performSegueWithIdentifier
Connecting button to new blank viewController
When I press the button, no initial load functions are called on the destination VC (Ex: ViewDidLoad, ViewWillAppear, etc). When I connect it to a blank view, the segue works fine with the same code in place.
Since the code never stops, or breaks, and just seems to "freeze" in place while still running on Xcode I can't seem to even narrow this down to whats causing the issue. I have a similar segue that is also called from another button on the same ViewController that has no issues whatsoever.
Any thoughts on the matter are greatly appreciated!
EDIT: I have narrowed the issue down to the UITextView's causing the problem. Once the Text Views were removed the page loads fine via segue. I wonder what changed between iOS 8 and iOS 9 in terms of UITextView as I will have to remove the text views and completely re add new text views.
So basically the segue was freezing because of the UITextView's I was using in the destinationViewController. The following fixed the issue:
Delete all UITextView's
Add new UITextView's
you must leave the default lorem imposed text and change this programmatically in the viewDidLoad()
This was the fix for me, and from the research I have done on the issue it seems this is a bug in iOS 9 and Xcode 7.
Cheers!
NOTE: Removing the text in the UITextView (or making it longer then ~12 characters) is sufficient to work around it, no need to delete and recreate them. This is fixed in Xcode 7.1.1 and later.
I ran into the same issue and the fixes in this post (Xcode 7 crash: [NSLocalizableString length] 30000) solved the issue for me.
The first is to enable a localisation other than the base for the storyboard (see https://stackoverflow.com/a/32688815/3718974)
The second is to turn off the base localisation (see https://stackoverflow.com/a/32719247/3718974)
I think I have the same problem: I have a UITabelView with cells created from a nib file, when a user tap a cell this method is called:
and when I have the following method prepareForSegue:: the application crashes:
if I delete the line 129 Everything is ok , the method prepareForSegue:: open the right view and the label contactName is shown with its default text.
If I modify the method as follows prepareForSegue:: get exactly what you expect, without having any type of error:
let me know if you also get the same result
Any one who is facing this issue, i solved it by turning off the "Optimize rendering for windows scale" option in Debug of simulator window. I already had tried all of the above answers but could not solve the issue.
In the method in the first viewController where you activate the segue, do you have beginIgnoringInteractionEvents anywhere? If so the screen you segue to will be frozen and will ignore interaction events like you describe. If this is the case you can fix this by adding an endIgnoringInteractionEvents method before your segue method:
UIApplication.sharedApplication().endIgnoringInteractionEvents()
self.performSegueWithIdentifier("editItemToMyGearSegue", sender: self)
I realize this is an old topic, but appears to be still relevant. I was facing the same problem in Xcode 9, iOS11. My UITextViews are embedded inside UITableViewCells. Same symptoms as described here. The tricks with default text and placeholders did nothing for me, but I solved it by turning off the scrolling indicators for the text view in the xib. They were on by default, I guess, though unused.
Edit: this is probably an important detail... the views that were hanging all had an image NSTextAttachment in the attributed string of the text view. I think the image was wider than the available table cell content. With scrolling turned off, they appear to downscale.
I have a basic watchkit app that loads a page based navigation of 3 interface controllers. This works well, but I'd then like to trigger an action to remove the page-control and essentially revert back to the original InterfaceController that was present when the app first loads.
// load page based control, with 3 views. this works ok
[WKInterfaceController reloadRootControllersWithNames:#[#"pageController1",#"pageController2",#"pageController3"]
contexts:#[#"data1",#"data2",#"data3"]];
// attempt to reload original interface controller, identified by storyboard id
[WKInterfaceController reloadRootControllersWithNames:#[#"myInterfaceController"] contexts:#[#{}]];
The page based navigation remove, the original navigation loads after a short spinner. However it fails to function correctly and original Actions result in this error.
Extension[6766:123665] *********** ERROR
-[SPRemoteInterface _interfaceControllerClientIDForControllerID:] clientIdentifier for interfaceControllerID:(null) not found
Is there a better way to cleanly reload the original InterfaceController?
EDIT, 2/19
It seems there are some other actions that are causing this error too. For instance, if segue to a second InterfaceController and then popController to get back, the error often appears. It is always related to a secondary call to this function.
[WKInterfaceController reloadRootControllersWithNames: contexts:]
EDIT2, 3/18
As previously mentioned, this is reproducible 100% of the time by doing the seguePush, the popController, then attempting to reloadRootControllersWithNames.
If the seguePush/popController is not done beforehand, then the reloadRootControllersWithNames will work fine.
This situation seems to be in addition to the multi->single-multi instance of this bug.
This is actually not a bug because according to Apple:
You cannot combine hierarchical and page-based interface styles. At design time, you must choose the style that best suits your app’s content and design for that style.
So unfortunately, we can't mix Hierarchical and Page-based navigation patterns within the same Watch app.
Just one of many limitations we have to deal with when developing apps for Watch
This is a bug in WatchKit in Xcode 6.2 Beta 5. Please dupe the following radar on Apple's Bug Reporting System to help raise the priority to get this fixed.
In the meantime, a workaround that I've found can be found on the dev forums. What you can do is add a dummy interface controller to any single interface controller page set so you always have two. This will fix the error until Apple get's the bug fixed (hopefully in Beta 6). Please dupe!
I was able to solve my instance of this problem by not using popController on a pushed view controller. Instead I use a reloadRootControllersWithNames in place of the popController.
How this allows both push and paging, via an example:
Push a view controller
reloadRootControllersWithNames to return to the original controller. (The transition is not quite as animated, but is sufficient)
Create page based view controller.
reloadRootControllersWithNames to return to the original controller
Repeat 1 or 3 as needed.
This eliminates the error at the cost of non-animated popControllers, and allows partial pushing and paging. It would not allow more complex push navigation though.
There may be a better method of navigating to a sub interface controller without a push call, but I'm not aware of it on the watch yet.
None or the answers above worked for me. This problem began when I changed the icon names for the app and the watch app name. I solved it like this:
1) Click on your Watch app Target > Capabilities > make sure app Group
is in ON.
2) Make sure the App Group is selected.
3) Clic on the circled arrow Refresh icon (this will apparently just
refresh this thing if you already had it)
4-Repeat steps 1-3, but for your Watch App EXTENSION target too.
5-Click on the Scheme button (on the right side of the STOP button),
and clic on Edit Schemes.
6-Click Run > Info 7-In executable select your target (Actually it
should already be selecting but opening this window seems to
refresh the option, and wipe the error)
Apparently all these things above are not updated automatically when you change the icon name (Target names) and you have to go to those menus and open them to refresh them manually. Shame on Apple perhaps?
I have an iPad app that has been running fine until iOS7. This issue seems to be only on ipad 2nd gen models and earlier when iOS7 is installed. Anyway, I've been tearing my hair out trying to figure out where this error is coming from, but have had no luck. The console in xcode (5) reports the following error after I perform a logged in segue:
2013-11-18 11:17:31.768 MyApp[400:60b] *** -[UIToolbar backdropView:willChangeToGraphicsQuality:]: message sent to deallocated instance 0x18ec23e0
I can't lookup the address for more info (image lookup -a 0x18ec23e0) it just returns nothing.
In instruments running zombies, it reports that a message was sent to a UIToolbar like so:
When I inspect the instance, I get the following:
How do I debug this? I have no idea where this call is being made and it seems dependent upon a physical deivce (doesn't happen on the iPad mini or ipad 3/4)
I was struggling with a very similar error, also with a UIToolbar, that I couldn’t figure out until a couple hours ago. I also had to use and try to understand the zombies’ instrument but without any luck.
What I did was to pay a close attention to the call stack that was presented when the Exception Breakpoint was activated as described in the following tutorial:
http://www.raywenderlich.com/10209/my-app-crashed-now-what-part-1
Even though the call stack didn’t point me to the exact code line, I noticed that the app was trying to add a UIToolbar to a ViewController. Turns out that what I was doing was creating a local UIToolbar inside of a method and adding it to the presented UIView. After have modified this behavior, I stopped having the annoying sudden crash. It was difficult for me to find the issue because looking at the code of the ViewController that caused the crash, there was no code that created or used a UIToolbar; however this VC included a custom view that did exactly that, as I explained before.
Have said all of this I recommend you to closely inspect the VC that generates the crash. If you need to create a UIToolbar programmatically I recommend you to declare it as a strong property to maintain the memory reference as long as needed.
I hope this helps you.
I struggled with this for a while today. I had two storyboards, one for login/signup (set as the main storyboard for the project) and another with the rest of the application. The app delegate would detect if a user was logged in and instantiate the root view controller of the other storyboard. The root view controller of the login storyboard is a navigation controller and after after some investigation with instruments I realized there was a UIToolbar being instantiated from the nib. Opening up the storyboard file revealed an off-screen UIToolbar object in the root view controller. I deleted it and I'm not crashing any more.
I should also mention this crash was only occurring when I was using MKMapView.
I'm presenting an MFMessageComposeViewController which works fine with iOS4 and iOS 5 but has problems with iOS6.
The view is presented ok but if its dismissed and then represented it doesn't display correctly - only the To: line is displayed, the body and keyboard are missing. (Sorry I can't post a screen shot at the moment as XCode crashes when I take a shot, I'm downloading an older version of XCode as I type).
Stepping through the code in the debugger I noticed that the problem may originate earlier than the re-presenting - I noticed than when dismissViewControllerAnimated: gets called the entire view does not get dismissed immediately, rather it is the message body and keyboard that gets dismissed leaving behind the To:, the same thing that is displayed when the view is re-presented.
It might be easier to describe with screenshot, I'll post some shortly.
I'm using presentViewController: and dismissViewControllerAnimated: to present/dismiss the MFMessageComposeViewController.
+++ UPDATE +++
I've found the problem can be solved if instead of using the same MFMessageComposeViewController object to re-present the view I first delete it and then create a new one.
That seems a little inefficient though, and it should not be necessary I'd have though, like i mentioned it worked on iOS5.
In iOS 6 apple introduced a new feature "remote view controller". Some external view controllers are no longer part of your app, and the messaging controller is one of them.
I guess that is the problem in your case.
you can read more about it there: http://oleb.net/blog/2012/10/remote-view-controllers-in-ios-6/