It's entirely possible. I'd say it's likely, even.

Easy thing to test, though: run your project in an offline folder for an evening and see if the same thing happens.

Changing to a project folder will probably help as well. A capx is a compressed version of a project folder, so it has to sync the whole thing every time, including all sprites and event sheets. If you work in an "expanded" version of this - a project folder - it'll only have to sync the individual files that are changed.

Sounds like something an anti-virus would do. I suggest that you try disabling that or at least tell it to ignore your working folder.

I've also had issues like this in the past where I've been working directly in a sychronised folder (cloud storage). If I save too frequently before the file has finished synching, I'd get errors trying to save. This happened specifically with Syncplicity, though - Dropbox doesn't have the same problem.

Lastly, I'd suggest working with a project folder, instead of a capx. You'll probably notice improved save times, as well.

But it can't be a driver bug - it's happening on both AMD and NVIDIA cards, as has been demonstrated.

I strongly urge you to test on more devices. Testing on three is hardly conclusive, given how erratic this issue can be.

On start triggers are processed in that order. So if you have On start events in your included event sheet and your parent sheet, it'll first run in the included sheet, then the parent sheet. If an action in your parent sheet is dependent on something happening first in your included sheet, it should work fine in this case.

I use this principle when creating a game that has multiple levels that share one common event sheet but each have their own sheet for special circumstances (in a puzzle game, for example): I can be sure that anything that's set both in the common sheet and in the unique sheet will be overwritten by the unique sheet if necessary.

Ooh, I see now. It's a z-order thing. If you set the light emitter to be at the top of the z-order, then the blue sprite is shrouded in shadow as expected.

Is it not just an opacity issue? The light sources are completely blacked out because they're in 100 opacity shadows, but even then they still cast shadows. If you drop the opacity then it seems to work how I'd expect it to.

Unless I'm misunderstanding..?

You had a few logic issues and inconsistencies, so I've uploaded a fixed up version with comments so you can see some general improvements on your code.

The actual issue was that you weren't converting your WebStorage.LocalValue("HighScore") from a string to an integer, so it couldn't be compared with the global variable TotalDistance.

Included event sheets are always read first, regardless of where the Include event sheet code is used in the event sheet. They are, however, read in the order that they're called in the "parent" event sheet, so there is certainty that can be relied upon.

Chromium will also blacklist the accelerated canvas2d on some device, e.g. the older Qualcomm Adreno device, see the link crbug.com/161575. To confirm the gpu feature status on your device, I suggest to use Chrome for Android for help, you can type about://gpu to check the output for Canvas category.

Hi hmin, I'm currently in the process of moving over from CJS to Crosswalk, and I came across this issue. What is the resolution process for this bug? Is this something that you guys deal with on your side, or is WebGL just never going to work with Crosswalk on a Galaxy S3 due to a driver bug that's beyond your control?

I understand that blacklist ignoring is something that's coming up soon in Intel XDK. Will it be safe to simply ignore blacklistsed GPUs to ensure WebGL works? I've been testing on my GS3 using Cocoon for some time now (with WebGL enabled), and it's always behaved itself fine.

It's not a bug. You need to set the layout's "unbounded scrolling" property to "yes".

This issue has come up before (and was not resolved) but I've been able to consistently reproduce it on my side, so I figured I'd post an official bug report.

Problem Description

What happens is any SpriteFont object that has the Tint effect applied will exhibit strange stretching glitches in the layout view (runtime is fine). This only applies to multiple instance of the object - the first instance works as it should.

As the user pans around the layout view, the stretching warps and twists. It makes it quite difficult to work with text in this way.

Attach a Capx

Attached

Description of Capx

Just open the file and it should look wonky.

Steps to Reproduce Bug

• Create blank project with WebGL enabled.
• Create SpriteFont object, drag out multiple instances of it.
• Apply the Tint WebGL effect.

Observed Result

See attached image "text stretched"

Expected Result

See attached image "text normal"

Affected Browsers

N/A

Operating System and Service Pack

Windows 8.1

For reference, I'm using an AMD HD 7950 with driver version 13.251-131206a-166151E-ATI

Construct 2 Version ID

R169

In my experience, functions work 100% in CocoonJS, but there are a number of things that don't work. Check the tutorial (possibly outdated - a few things have changed since it was written) for more details.