Thank you for the response. This is all I was looking for. I appreciate you taking the time to look into this further.

It is a shame something is broken in WP8.1. That means all my projects are now on hold as every WP app I make will end up crashing after some time. I will pursue this with Microsoft further. Please feel free to close this bug, and please let me know if you know how I can get in touch with Microsoft about this (I have not dealt with them before).

I am re-opening this issue because your answer was completely inaccurate and incorrect.

You replied: "The snapshot you sent me on Twitter (note: it's more useful to keep it in the bug report itself)". That is wrong. I never sent you anything on twitter. I attached the screenshots of the memory leak directly to this bug report. I have no idea what you are referring to.

You replied: "screenshot only indicates 179mb usage". That is wrong. Again. The screenshots attached to my ticket clearly show the memory leak over time going up as high as 356mb after 90 minutes. Here it is again:

[attachment=0:t4ev4py4][/attachment:t4ev4py4]

You said: "If you want to send a more useful snapshot, export again with "minify script" off, since the minify option has made the majority of the snapshot unreadable". That is wrong (yet again). The screenshots (and the issue) was observed without minify script enabled.

I may be willing to accept that it is with Windows Phone 8.1 code, as it is after all still in beta, but the fact that you are talking about things I am completely unaware of (such as sending you minified code on Twitter) leads me to believe that you are either mistaken, or are responding to a completely wrong ticket. Furthermore, I have tested by running other WP8.1 apps (not from Construct 2) and they successfully run for over 8 hours without crashing.

I would truly appreciate if you could test with one of the tutorial files I mentioned, following the steps I mentioned, and send me your javascript memory test to prove that this is not a bug. Because as of right now, I have no confidence in what you are telling me. If you need me to send you a log, a test, a script, a file or anything else, just tell me exactly what to do, and I will do it.

I am a Windows Phone user, and the main reason I chose to create my app with your tool was for Windows Phone support. Prove to me that this is not a bug, or refund my money. Your choice.

-------

Problem Description

Over time, a project exported as a Universal 8.1 (Windows 8.1 and Windows Phone 8.1) app from Construct 2 r168 will leak memory only in Windows Phone 8.1. Despite sharing the same code, the Windows 8.1 build does not leak memory. Only the Windows Phone build does.

Attach a Capx

This issue was observed with my own capx as well as your own template files. No need to appach a capx file... simply use the Flapping Bird template with no modifications. I tested and confirmed the same issue occurs.

Description of Capx

Test with Flapping Bird template.

Steps to Reproduce Bug

• Export app from Construct 2 as Universal 8.1.
• Open the project in Visual Studio 2013 with Update 2 RC.
• Ensure you are are the Windows Phone app
• Debug -> Performance and Diagnostics -> select Javascript memory, and start
• Do nothing. The app immediately starts leaking memory, even just sitting on the main play screen. The memory number keeps going up. It never drops back down.
• Try with Windows 8 and notice there is no memory leak.

Observed Result

The app slowly increases total memory used, about 1-2 MBs per minute. Eventually, app crashes as it runs out of memory.

Expected Result

I expect the memory use to remain the same over time.

Affected Browsers

Windows Phone 8.1 Only.

Operating System and Service Pack

Windows Phone 8.1 (issue occurs on device with Lumia 920, and in emulator)

Construct 2 Version ID

Construct 2 release 168 64bit

Funny... I never sent you anything on twitter. I attached the screenshots of the memory leak directly to this bug report. I have no idea what you are referring to. Furthermore, the screenshots attached to this ticket clearly show the memory leak. Not to mention, the screenshots (and the issue) was observed without minify script enabled.

I may be willing to accept that it is with Windows Phone 8.1 code, as it is after all still in beta, but the fact that you are talking about things I am completely unaware of (such as sending you minified code on Twitter) leads me to believe that you are either mistaken, or are responding to a completely wrong ticket.

I would truly appreciate if you could test with one of the tutorial files I mentioned, and send me your javascript memory test to prove that this is not a bug. Because as of right now, I have no confidence in what you are telling me.

Andy

I'm going to bump this daily until I get a reply.

Knock knock... anybody home?

Ill run it again when I get home tonight.

I'd appreciate that. Thank you.

I did everything you mentioned and I was unable to reproduce this bug. Waited for over an hour.

Thanks for the test. I'd love to know more about your system... what versions of software you're running, etc. And I'd love to see a memory test to get more info (like the images I attached). You can get it by going to the menu: Debug -> Performance and Diagnostics -> and selecting Javascript memory, and start. Make sure you are on the Windows Phone version... it's not an issue with Windows.

I'd also love to get confirmation from Scirra directly if they tested and reproduced the issue.

We deal with a very large number of bug reports, the majority of which are user's own mistakes or unrelated issues, and not bugs in Construct 2. ... strongest possible evidence that the defect is in Construct 2 itself, or in as specific a component as you can possibly identify.

I understand that. But when every one of your own tutorials is causing the problem, clearly it is not an issue with my code. And, I gave very specific steps to reproduce the issue with a very small tutorial file that came packaged with the software. It would take you (as it takes me) under 5 minutes to load the tutorial, export as universal, load it in VS2013, and see the problem.

I'm starting to regret my decision to purchase. I like your software, but this is unacceptable. I found a severe bug, and provided you with plenty of information on how to reproduce it. What exactly did I pay for if I'm going to be told I have to solve my own problems with your own code? That's not what I consider to be good customer service.

Isn't that your job, since I paid for your application?

I'll see if I have time to do this tonight.

I have tested and I can confirm that the issue occurs with the Ghost Shooter (tutorial example) at 264kb, the flapping bird template at 196kb and even the Autorunner template at only 56kb. The Driving Game template at 13kb was not an issue (it did the rise and fall motion I described earlier).

Hope that helps speed things up for you. The sooner I know what's going on, the sooner I can make my future plans. My app is just sitting in the Google Play store waiting for proper promotion while I wait for this to get sorted out.

The Flapping Bird template is (considerably) smaller than my alarm app, so that's the best I can do. I'll test a few other templates and let you know.

Worth noting: Interestingly, if I try the examples (like, say, the 8 direction templates), I see a similar behaviour but with cleanup. The memory rises slowly, then returns to normal, rises, returns to normal, and so on. The same example on Windows 8, the memory just stays steady. It looks like as soon the project has events, the cleanup phase doesn't happen.

Andy

Any updates or an eta as to when I might hear back regarding this issue? I've released my app in Google Play and I'm waiting to hear back before I can release to Windows and Windows Phone.