[TIMOB-23220] iOS crash on databinding when debugging in studio with SDK 5.2.X
GitHub Issue | n/a |
---|---|
Type | Bug |
Priority | Critical |
Status | Closed |
Resolution | Cannot Reproduce |
Resolution Date | 2016-08-08T22:35:18.000+0000 |
Affected Version/s | Release 5.2.2 |
Fix Version/s | Release 5.4.0 |
Components | iOS |
Labels | Debugging, Studio4.5.0.201602170821, ios |
Reporter | Alexander |
Assignee | Angel Petkov |
Created | 2016-04-12T10:16:33.000+0000 |
Updated | 2016-08-08T22:35:18.000+0000 |
Description
Attachments
File | Date | Size |
---|---|---|
HarnessTemplate.zip | 2016-04-14T07:31:07.000+0000 | 8754465 |
Hello, Can you please explain a little more. I am not understanding your guidance in steps to reproduce "create a test app with jake using models/binding_tableview". Please provide a sample test case or a sample project for us to test in our environment. We will let you know the result ones we test the issue in our environment. Regards, Sharif.
Hello Sharif, I made an effort in avoiding as much custom coding as possible. So I reverted to the alloy repository to try and recreate the issue. This is what I did: I created a test app using appcelerators repo: https://github.com/appcelerator/alloy Following these steps provided in the readme of the repo: [Readme section B](https://github.com/appcelerator/alloy#b-by-cloning-the-repo-and-using-the-jake-test-runner) {quote}B) By cloning the repo and using the Jake test runner
first, clone the repo
git clone https://github.com/appcelerator/alloy.git cd alloyinstall jake globally
[sudo] npm install -g jakeinstall alloy globally from the cloned repo
[sudo] npm install -g .install alloy's local testing dependencies
npm installrun a test app
jake app:run dir=models/binding_tableview {quote} NOTE: I changed *_jake app:run dir=basics/simple_* from the read me to *_jake app:run dir=models/binding_tableview_* I run this app from CLI and after that imported it in studio as stated in section C: [Readme Section C](https://github.com/appcelerator/alloy#c-importing-the-harness-to-titanium-studio) {quote} C) Importing the "Harness" to Titanium Studio You can use these apps through Titanium Studio too. The easiest way to do that would be to import the test/project/Harness into Titanium Studio. After that, everytime you run jake, your project in Studio will be updated. Once in Studio, you can run for any platform, Titanium SDK version, or change any settings you want. This will give you a lot more options and power than running solely from the command line.{quote} After that run the app from Studio in debug mode. As soon as you click addItem (rightnav button) the issue appears and the app crashes. Best regards, AlexanderAdded an attachement with the example project that causes the issue.
Hello Angel, Testing the prescribed sdk (5.4.0.v20160411201753) did not work. With or without the run-on-main-thread element the app crashes as soon as an item is added to the list using the model. However, the newer sdk did have success. I tested using sdk [5.4.0.v20160419234222](http://builds.appcelerator.com/mobile/master/mobilesdk-5.4.0.v20160419234222-osx.zip) Now the app doesn't crash and I am able to debug. The debugger stops on every breakpoint (single or consecutive ones). Best regards, Alexander
Hi, Alex thats great to here. Ill be resolving the issue now, thank you for testing it with the newest SDK!
One note though, It works in the simulator only. I was not able to install an app on device in a debugging session, using this sdk. I suppose this is a different issue.
Ah thank you for letting me know! We'll investigate the issue and create a new ticket if its re-producable of which I'm sure it will be. Are you able to build to devices using the other SDK's ?
Hello Angel, I tried installing on an iPad 2 with iOS 9.3.1, and it installed both in run and debugging mode for sdk: 5.4.0.v20160419234222, 5.4.0.v20160411201753, 5.2.3.v20160404160237. So success, however the debug-connection was never established. I tried installing on an iPhone 6 with iOS 9.3.2(beta) and I succeeded installing only once in debug mode using sdk: 5.4.0.v20160419234222. After that I kept getting this error, on all three sdks upon install, both from debug and run mode on all attempts, restarting or reconnecting didn't help: [ERROR] : An error occurred during build after 19s 740ms [ERROR] : Failed to start "com.apple.syslog_relay" service (0xe800007f) [ERROR] : Try reconnecting your device and try again. Best regards, Alexander
[~Alexander Baars] Hello, I'm sorry to hear you've ran in to another bug , your patience is appreciated. Does this error message appear only with iOS 9.3.2 ? Does it work with iOS 9.2 and 9.3 GA ? Edit: I think the crash might be happening because i misled you, instead of doing run-on-main-thread the way i suggested above. Use
Add this above the iOS tag *not* after. We changed the way we declared mainThread for parity reasons between platforms.
Hello Angel, Sorry for my late response. Using the sdk 5.4.0.v20160419234222 I was able to install a build from a debug session on my iPhone running iOS 9.3.1 GA, with and without the run-on-main-thread property. However the debugger-session hung in Waiting for debugger to connect... Cancelling this message makes the app run. Also I noticed the app starts with a black screen (regression?). I also tried sdk 5.4.0.v20160429102159 with the same results. Best regards, Alexander
I am unable to reproduce this issue, using: MacOS 10.11.5 (15F31a) Studio 4.7.0.201607250649 Ti SDK 5.4.0.v20160802165655 Appc NPM 4.2.7 Appc CLI 5.4.0-37 Alloy 1.9.1 Xcode 7.3.1 (7D1014) iPhone 6S, iOS 9.3.3 Tested using the data binding test apps in the Alloy repo as well as my own test apps using data binding. I had to turn off
<run-on-main-thread>
due to TIMOB-23726, but I was able to debug on devices and simulators. All apps were able to build, install, and debug properly.Hi [~Alexander Baars], We are having trouble reproducing the original issue you reported, however I have recently confirmed your issue with the Debugger hanging on "waiting to connect" as a valid bug and have created a ticket in regards to it: (See TIMOB-23726). For the time being, I will close this issue as "unable to reproduce", and the progress towards the fix for your current issue may be tracked in the aforementioned ticket. If you are still experiencing the original issue here, we will reopen this ticket and investigate further. Best Regards, Harry Bryant Quality Engineering