Even on my exact computer with my exact project file, repeating what I did in the video, I can't replicate the crash on command, so I see no point in posting my file. I had hoped that by sharing some more footage I would help someone more knowledgeable about Construct 2 than myself narrow down a cause, but apparently not.

Again, my project is light. A half-dozen sprites with a few animations and a couple tilemaps. If you want to experience the crash for yourself, I suggest you work on anything in Construct 2 instead of Construct 3 for a day, and you probably will.

As for myself I will try downgrading to an earlier version, editing the <saved-with-version> line in my project file in a text editor, as another user already suggested.

Is anyone who experienced issues in this thread able to confirm if r276 fixed the problem?

Still crashing on r276, but not as often as in r275. You wanted video? streamable.com/v8wgj

Working on tilesets in this case, though I had been working with animations earlier. Tabbing back from a preview and the program just closes without a crash dialog box.

Mine is also a pretty light project and my specs are more than adequate; i5-6600K 3.5 GHz, 16 GB memory.

Came back to Construct for a project after a long time away. Never used to get these crashes.

Thanks again! That did it.

As for both loops running for each instance, I'm not confident enough to say I didn't do it wrong, or that I didn't do it in a terribly roundabout way, but I think it's working as intended (until I break it again).

Basically the first loop creates the tiles and transfers variables from the "map" to them -- ideally, later, this map could be some kind of text file or spreadsheet that I can import data from instead of being drawn by hand in the layout... if I were to ever figure out how to do something that complicated.

Then, the second loop from the nested function. It's no longer dealing with the map tiles. It runs through each of the created IsoTiles, this time according to the view angle: the northernmost tiles will be at the top by default, but could be at the bottom if the view has been adjusted with the (][) bracket keys. Lastly the third loop sorts the z-order accordingly for each tile, which depends on the view angle and so has to be done last. Any time the view is rotated it redraws all these tiles again starting over at the first step, creating them from the map again. Sensible enough? Maybe? Who am I to say.

Thank you for the quick response! It was very helpful.

I had to do some tweaks to get things running in the right order, but it seems to be working now. I just have one problem with the last part that sorts the Z-order of the tiles now.

If I put it in the function, I believe the logic is that it can't really "sort" them because it's only seeing one at a time. After it "sorts" that one tile, only then is it handed another. But if I nest it anywhere else it doesn't want to work. The same problem is as before, where it's moving onto another event "without any specific IsoTile picked" as you said. (I just want it to pick all of them, once, each time the other stuff is activated).

And if I just have it at the end, it works, but as you can see from the "test1" global variable in the debugger, it's working a little too hard. What would someone normally do in this case?

capx: drive.google.com/open

Hello. I may be misunderstanding something fundamental about how events work, but I've been giving myself a headache searching through tutorials, and trying various adjustments without success, so I hope someone can help me here.

Specifically, after one event creates several of an isometric tile object, transferring a bunch of variables over to them, I'm unsure how to come back to the lot of them and have them move according to some global variable or otherwise.

It's hard to explain, but as you can see from the capx, there are two sections which don't apply to my tiles as I had intended, so the objects do not get separated, and the values are left empty for some of their variables.

capx: drive.google.com/open