103 bug (103.2 too) disapearing sprites.

Since 103 some of the sprites in my game started to disappear at random times.

Does anyone else have a problem similar to this?

I've checked a lot to see if it wasn't a problem in my event programming, I even placed debugs to check for visibility and opacity on the disappearing sprites and for layers too.

And 100% of the times the sprites disappear everything remain "visible" and with "100 opacity" although you can't actually see the sprites, =(

This doesn't affect the game in any way other than visually, since all the collisions remains there.

The problem occurred in chrome and on internet explorer, both when ran from construct2 as preview and when exported. it also happens on all the pc I've tested (although all of them are windows 7 pro X 64)

Nonetheless it only seems to be happening with 3 object's sprites, sometimes 4. and it happens 100% of the times before the end of the layout. It fixes on restart.

I have no idea how to recreate the problem other than playing my game, and since it's a 700 event game D: I don't know if I should send ashley the capx.

however, this is a previous version of the game when the problem didn't happen yet: scirra.com/arcade/shooter/1764/notevenalphayet-stage-1

Please see how to report bugs. It is extremely difficult to do anything about a problem without a .capx file, and steps to reproduce the problem. It will help me fix the problem even quicker if you can provide a minimal .capx designed from scratch which reproduces the problem.

I'm terrible sorry ashley, I did look at it. It said there you guys don't wanna look into big .capx files, and I have absolutely no idea how to reproduce the problem via events, but only by playing the game I made. and as I said above, the sprites are getting invisible while the program still considers then visible, valid and with 100 opacity, so I know it's not a problem on my side.

Wowever, if you want to, I could email you to the mail you've provided if you wanna take a look at the 700 events .capx (which I know is anything but ideal T_T and I'm terribly sorry about it)

edit: ... and i guess my best option here is just send you what i can =/

Sargas - if you can (back up then) delete as much of your project as possible while still having something that reliably shows the issue, that should do. It's also helpful if you test as many browsers as possible, and on different computers, and let us know if it's the same everywhere or just one computer. The problem is even if it's the same everywhere, it could just be a mistake in your events, and it changed recently due to a breaking change or something else altered in the engine rather than a bug.

ok, will do. I'll post a reduced capx as soon as I can.

and just for the sake of wondering, could something like this be triggered by having too much stuff on a game? maybe is a render issue?

I started on may i think, and construct 2 is my first ever approach to making games, and I wonder how big can a game be and remain stable or maybe it isn't even an issue?

I have exactly the same issue. It got so bad that I reverted to 95. I have tried on multiple computers in multiple browsers with exactly the same effect. This is what is happening:

dl.dropbox.com/u/22577765/CelestiBug/index.html

It might be hard to see if you don't know what to look for, but if you watch the stars carefully, one every few seconds will vanish. After a little while the game will throw and error and simply freeze. After reverting to 95 and recovering my save file from dropbox, I no longer have the problem. I'll post my capx soon. Unfortunately I will need to reinstall the new update to fiddle with the broken file.

ryguydavis - we cannot investigate problems unless we have a .capx files for the project. Exported projects cannot be reverse engineered in to .capx files.

ryguydavis well, they do banish, though my game doesn't freeze then the sprites start to disappear :S

things related to this:

1)not all sprites are affected

2)at the time the disappearing starts bullets-behaving objects start to move much faster.

And for some reason my game that runs a 60fps on chrome, runs at 3-12 on internet explorer :S

I'm still working on reducing the size of the capx =(

Can anyone provide a .capx not using third party plugins that reproduces this issue? Otherwise I will close the report.

I just came onto the forum to see if anyone else had this issue as well, and I see that I am not the only person.

The problem I have is that just completely random instances of monsters are invisible. There is no consistent way to reproduce this that I can see, and it seems to just randomly effect sprites.

I uploaded my capx all the same, but I have no idea if it will shed any light on the problem. I only have noticed this issue after upgrading to 103.

Link to capx: dl.dropbox.com/u/13425402/ZombieTest.capx

I'm almost never having this problem anymore. I use to have nearly 2000 objects in my layout, and I reduced to around 800, which seems to have solved it. I'm pretty sure it is a workload issue on the renderer.

However, this problem, as we've all stated, wasn't present before 103.

Im not so sure the number of objects has much to do with it. I see it happen when there are only a few guys spawned, and I see it more often when there are many, but I think that might just have to do with the fact that the odds are higher with a higher number of objects.

I know i'm just guessing here, but I noticed it triggered a lot often when there was a lot of collision calculations. Some times you could make an object disappear just by standing on it.

It's just a tough bug to pinpoint. Just as an additional note, when my monsters disappear, they are not physics objects or anything.

Hi Ashley,

Please don't close this thread. I also notice an odd flickering of a sprite since build 103. A lot of us can only use Construct2 on our free time so sometimes it can take a while for a user to be able to create a simplified project file which demonstrates a bug. Keeping this post alive lets us discuss the bug and hopefully we can figure out the exact conditions to reproduce it.

I suggest instead of closing threads like this, just remind us until the bug is reproducible you can't do much about it and leave it to us to report to you when we finally figure out how to reproduce it or at least have a simple project file which sows the bug.

-Mike