I have a question or two about using AVFoundationFramework in XCode.
I downloaded AVCamDemo source code. I am trying to leverage the code in my app to replace UIImagePickerController (because it is slow) with the AVFF functionality. However, I cannot figure out how to exit the AVFF code when "Done" button is tapped in the UI. My app will provide the Done button.
Specifically, I need to know:
What do i need to stop the capture session? Just [session stopRunning] ??
My app's ViewController will invoke the AVFF, and will provide the "Done" button in the view controller that I will create for the AVFF functionality. When the user taps the "Done" button, I need to stop the capture process and exit back to my app. I am not sure how this can be done.
Any help will be highly appreciated.
Regards, Sam.
Yes. Although you may also want to ensure you turn the video torch off if you had turned it on, and if you called lockForConfiguration on the device you should be sure to unlock it.
There is no "exiting". If you presented your view controller using presentModalViewController:animated:, use dismissModalViewControllerAnimated:. If you pushed it onto a UINavigationController, pop it back off. If you did it some other way, dismiss it in the way appropriate for that method.
regarding your question on AVFoundationFramework:
I don't exactly understand what you are not succeeding to do. Some general options:
[session stop] will stop capturing.
You can play with the AVCaptureVideoPreviewLayer layer - that's what actually shows the camera preview.
All the ui code is in AVCamViewController - so you can just pop it from the stack, or remove it from superview - which ever is correct in your code.
If these don't help, please try to elaborate further on the problem, and perhaps supply some code.
Cheers,
Oded.
Related
My app uses multiple features for Apple's demo project AVCam. Everything works just fine except when I exit the app and go back in it, it doesn't show the camera preview anymore. Does anyone know what piece of code I am supposed to use and where it belongs? I tried searching but a lot of questions relating to android popped up.
You need to reinitialize your camera once the App becomes active again. In your app delegate methods, override, applicationDidBecomeActive and send a notification so your view controller knows that your app became active again.
Pending the notification received, you can reload the viewDidLoad, or move the contents of viewDidLoad to viewDidAppear. There's multiple ways to do this. You can also reload the contents of viewDidLoad in viewWillAppear. There's many many many ways to do this, like I said.
I've read a lot here about misusage of UIAlertView and the fact that Apple may reject an app where UIAlertView is overused.
In my app I'm writing I have to update data rarely via online request and recalculating internal data structures. While this is in effect and running in a second thread I don't want that the user touches the GUI and the app's current settings etc. So I decided to show an UIAlertView with UIActivityIndicator and a UIProgessView for some seconds which is dismissed automatically when background work is done.
Do you think that it is a reason against HIG rules? If so do you have suggestions for good practice on this kind of work flow?
Cheers,
Konran
Instead of a UIAlertView, I'd use MBProgressHUD. It looks nicer and has a built in activity indicator. Apple uses a similar component in some of their apps.
https://github.com/matej/MBProgressHUD
I don't think an alertView should be used to stop a user from interacting with the screen. What you can easily do is add a "mask" on top of your content. Create a view; give a background color of .5 alpha black, and add it as a subview of your main view. This will dim everything else, and stop the user from interacting with it.
I have used an alert view with integrated activity indicator in a couple of apps and none of them was rejected from entering the AppStore. If you don't display any button inside the alert view the user won't even consider it as an alert view, so I don't see any problem with that.
I've made a custom activity indicator (actually just an imageView)
When user clicks something and I expect it will take a bit long to process(alloc a UIViewController and push on to navigation stack),
I alloc the indicator and add it as subview of current view just before the lengthy process starts.
Strange thing is, indicator doesn't show up until the push (left-right) animation starts.
Is it because the lengthy job takes the system, and ui drawing for activity indicator is delayed?
Am I doing something wrong here?
Thank you
Edit
Looks like I can do the "push" in background.. i'm trying it now
IPhone SDK - Leaking Memory with performSelectorInBackground
Is your job synchrone or asynchrone ?
If it's the first case, then it can be the problem.
Check all the method like :
[ self performSelector:<#(SEL)aSelector#> ];
You can thread this to avoid your [potential] problem.
Good luck.
You should process your lengthy tasks in the background. The UI won't update if you block the main thread.
So you have to refactor your app, the alloc and push of the viewController should happen within the blink of an eye, because you can't do this in the background.
But you can do the processing (downloading data I guess) in the background.
There is plenty information available about background processing. The way to go varies heavily on what you want to do exactly.
I am using the UIWebView to load both streaming audio and video. I have properly set up the UIWebView delegate and I am receiving webViewDidStartLoading and webViewFinishedLoading events perfectly. The webview launches a full screen window (likely a MPMoviePlayerController)
Apple's MoviePlayer example gets the array of Windows to determine which window the moviePlayerWindow is for adding custom drawing/getting at the GUI components. I believe this to be a bad practice/hack.
My expectation is that I should be able to figure out when that button was clicked by either a delegate method or an NSNotification. It may also be the case that I have to poke around subviews or controllers with isKindOf calls, but I don't think those are correct approaches.
Are my expectations incorrect, and if so, why?
What is the correct way to bind an action to that "Done" button?
There isn't an MPMoviePlayer instance method that covers this. You can use - (void) moviePlayBackDidFinish:(NSNotification*)notification to find out when the movie has finished. Or you could overlay the existing Done button with your own and have complete control that way.
You can also use MPMoviePlayerWillExitFullscreenNotification in order to conrol the action provided that youe MoviePlayer is in fulscreen mode.
I can't find any documentation to confirm this, but it appears that you can only call the method stringByEvaluatingJavaScriptFromString in overridden methods from a UIWebView delegate. Can anyone confirm this?
Here's what I've tried. I setup a button on a view, link it to a method on my viewcontroller, and make sure it works fine. My view has a UIWebView control on it as well. If I run the project on the simulator or on the iPhone, there are no issues. Then I add this code to the button's method.
[theWebView stringByEvaluatingJavaScriptFromString:#"alert('Hi there!');"];
When I run the project, I can click the button and see the 'Hi there' prompt and I can click OK to dismiss it. Usually 4-5 seconds later the simulator crashes. I occasionally see the "__TERMINATING_DUE_TO_UNCAUGHT_EXCEPTION__" error, but not consistently; sometimes there's no error. It also doesn't always crash the first time. Sometimes I go to another page, and then try it again, and it crashes.
If I put the same code in the webPageDidFinishLoad event it works fine. But I'd like the code to be called when the user demands it so that event doesn't suit my needs.
I'm open to a workaround if you have any ideas? Thanks in advance!
I still don't know the exact reason this didn't work, but I found I could rewrite my code to get called during the UIWebView delegate methods instead.