Strange graphical errors on Android devices.

Solar2D Android
#1

We’ve finished our latest game and have it in the review process for the App Store. So we’ve been in the process of making sure the Android version is ready to go.

When I installed a build to my Evo 4G it seemed to run fine except when I went to our ingame store.

The top and bottom black areas are supposed to have navigation buttons. The center area is a scrollView, if you scroll around with it sometimes the buttons will flash on screen. I can also press the buttons whether they are there or not. Also when I hit the back button (one of the hidden buttons) it goes back to the main game which just looks like this:

The green bars are drawn using the display.newRect() function while all other items on screen were images loaded from a file.

Any ideas why these things are acting this way? [import]uid: 69531 topic_id: 24725 reply_id: 324725[/import]

#2

Hey what version of OS does the 4g have on it?

Just wondering because I have seen a few other apps have some issues like that on my galaxy S running ICS Passion ( Icream Sandwich 4.x )

Larry [import]uid: 11860 topic_id: 24725 reply_id: 100202[/import]

#3

Hi, I’m not sure, but I think we may have the same problem (http://developer.anscamobile.com/forum/2011/12/31/android-graphics-problem-build-2011703?page=1#comment-99646), graphic images disappearing when there’s a masked object in the scene.

As you can see from the reply from Joshua Quick, it’s something they’re working to fix in the next few weeks.

Currently, the last version of Corona SDK not having this problem is the beta built 758, so you should try building with that release [import]uid: 9158 topic_id: 24725 reply_id: 100204[/import]

#4

hmm that’s definitely a problem
We are using IAP for our game so we’d need to use build 760 or later. I don’t even see a download for 758 anyways, it only goes back to 760.

I had a feeling it was due to the mask of the scrollView. Any idea on when we can expect a fix? Or a possible work around? [import]uid: 69531 topic_id: 24725 reply_id: 100209[/import]

#5

I’ve recently bought a Galaxy Nexus which is running 4.0.4 and this issue does not exist on it.

My Evo 4G running 2.3.7 does have the issue and my brother’s Evo 3D running 2.3.4 also has the issue. [import]uid: 69531 topic_id: 24725 reply_id: 102922[/import]

#6

This is very important to fix. Currently a game with masks can not be submitted with IAP. So big bottleneck in launching next version.

[import]uid: 22829 topic_id: 24725 reply_id: 102929[/import]

#7

This definitely should be a SUPER-TOP HIGH PRIORITY bug fix.

It’s a huge problem, we also have an app that we can’t submit because of this.

Any news Ansca? [import]uid: 9158 topic_id: 24725 reply_id: 102986[/import]

#8

Any news on this issue?

I’m stuck with daily build 758 but need some new functions… and building with a newer build show this error.

It’s definitely a problem on scenes with masks on tableviews [import]uid: 9158 topic_id: 24725 reply_id: 103441[/import]

#9

Everyone,

This is a known issue that was created since our rendering performance change in build 759. It happens on devices with low-end graphics hardware (typically Android devices) when using masks or displaying polygons (rectangles, circles, and lines). The only work-around at the moment is to revert to build 758 as noted earlier in this thread.

We plan on addressing this issue soon. Hopefully within the next 2 weeks. Sorry about the inconvenience. [import]uid: 32256 topic_id: 24725 reply_id: 103588[/import]

#10

I really hope that it gets fixed soon. It’s the only thing holding us back at this point and I definitely feel its a market we’re missing out on lot’s of potential money made.

I was contemplating just releasing for newer devices that we know it works on but I feel like that’ll be fairly small number of devices and not really worth it. [import]uid: 69531 topic_id: 24725 reply_id: 104918[/import]

#11

Great just built with daily build 807 and it seems to be fixed.
Thanks a lot! [import]uid: 69531 topic_id: 24725 reply_id: 105476[/import]

#12

Oh good! Yeah, this issue “should” be fixed now. We still need to do some additional testing on our end just to be sure, but so far it looks like we’ve nailed it. [import]uid: 32256 topic_id: 24725 reply_id: 105528[/import]