Ashley's Forum Posts

  • You do not have permission to view this post

  • Triggers fire outside of the event loop. So when 'On left button clicked' fires, it's not currently running the event sheet - the next tick it will run the event sheet from the top, so the next event that runs after the click event is event 1, not event 7.

  • Have you looked at the Space Blaster demo game? It uses invisible objects placed in the layout to spawn enemies. No XML/JSON/Dictionaries used at all.

  • I can't reproduce in r143 - it appears to be correct in both Chrome and Firefox. Can you verify it's fixed as well? r143 had a fix to how AdjustHSL handles alpha, so that seems likely to have fixed it (your sample images have lots of alpha).

  • Closing: this is in fact by design. Objects are only marked to be destroyed by the 'Destroy' action, and are in fact truly removed only at the end of the next top-level event. This is for two reasons:

    1) it's a useful feature: sequences like "Destroy sprite"; "Create explosion at Sprite.X, Sprite.Y" in fact work when it would not if the sprite was truly destroyed immediately

    2) the architecture of the engine requires that objects are not truly destroyed mid-event, because it may still be picked by conditions in parent events that still reference it.

  • Should be fixed in the next build.

  • Try Construct 3

    Develop games in your browser. Powerful, performant & highly capable.

    Try Now Construct 3 users don't see these ads

  • Your bug report does not really contain enough information to warrant investigation. I need to know things like the exact variables in question (did you mean "banlength"? The "banned variable" seems an ambiguous thing to say compared to just saying "banlength"); the specific event that you believe to be working incorrectly; precise steps to reproduce (which buttons to press how many times exactly in which sequence and then the expected event result vs. observed event result/variable values/etc), and so on. A large fraction of bug reports are simply mistakes in events, or users misunderstanding how the execution flow works when it is working correctly. The event engine is pretty mature and well tested at this point, so it would also surprise me if a simple tree of events like this did not work properly. So on that basis the report was closed as not sufficient evidence to prove there is a Construct 2 bug. Please feel free to repost with significantly more information if you think there really is a bug.

  • I think it might be a mistake in the AJAX plugin. Try deleting that line (which calls setRequestHeader for Content-Length) and see if it works. I've removed that line for the next build; I've read the specification does not actually allow that header to be set.

  • Closing, please post in English and follow the bug report guidelines.

  • You can download text files to disk by AJAX requesting them then using the node-webkit object to write a file on disk with the data returned from AJAX.

  • If there isn't a way to get input data in to a web page via a browser, we can't support it.

  • It appears to be pointing at line 199 in index.html. The official template that ships with C2 doesn't have that many lines. Did you modify it yourself? If so, perhaps your own modifications are breaking it. It also mentions "MicrosoftAdvertising.Shared.WinRT.Advertisement" - perhaps you're using a third party advertising addon which is causing the error?

  • I think Tetris is an ideal candidate for web delivery anyway. Then you get a nice and small download - mainly just the images and sound - offline support, mobile support, much easier accessibility (just a URL away), and greater chance of viral marketing/sharing. Why would you want to weigh that down with a desktop app? Modern versions of Windows even stop you running a downloaded EXE if it's not run often, for security reasons.

    You cannot rely on WebGL being present in node-webkit either. Node-webkit basically is Chrome, and just like Chrome, WebGL support is subject to driver and hardware blacklists.

  • thehen - I read somewhere it makes the Lumia 920 go too wide... hmm...

    Is that the only change necessary to index.html to fix it? Are there any other changes that would be useful to make?

  • Closing, please follow the bug report guidelines. It would also be useful to narrow this down to the specific beta that caused the problem. Note we haven't changed the Anchor behavior for some time (not since way before r139).