Ashley's Forum Posts

  • So it sounds like if you disable cookies in Firefox, accessing storage throws exceptions. (I don't know why they'd do that - returning empty results would be easier to support!) That line is the debugger trying to load from storage which items you added to the watch, which throws an exception. Post it to the bugs forum and I can try and make storage exception-proof.

  • It works in NW.js (desktop) and Crosswalk (mobile) / Chrome for Android, since all are based on Chrome.

  • I think you should report this issue to Microsoft: the touch feature of C2 is very simple and just forwards what the browser tells it. I can't reproduce on the Lumia 520 so I think it must be some Windows Phone issue specific to your circumstances.

  • OK, maybe it was someone different reporting the same problem, but I couldn't reproduce it - iOS 9 still opened links just fine, but it first pops up a dialog saying "Do you want to allow the page to open a popup window?" with "Allow" and "Block" buttons. If you press "Block" it may be permanent which might make you think it's not working. I guess it's also possible you could have added a new content blocking extension that blocks popups as well?

  • You can add effects to layouts as well as layers.

  • According to the whitelist plugin documentation, the "allow-navigation" tag only whitelists origins for navigating the webview itself, which should not affect AJAX requests. It sounds like the <access origin="*"> tag would do that instead. A new beta is going out today so I'll add in this and we'll see if it makes a difference:

    <gap:plugin name="cordova-plugin-whitelist" source="npm" />

    <access origin="*" />

  • Maybe in this beta cycle. We don't have a 6S device in the office to test on yet, but I'll try get one.

  • It's so if your tilemap has a large object, e.g. a tree, made from a patch of 3x3 tiles, you can quickly pick it up and stamp it out again. Also if your tilemap is arranged such that 9-patch sections are intact on the spritesheet too, you can pick up any 9-patch area with the same feature. I think you're asking for a select/copy/paste feature, which is a separate feature request.

  • Fillrate isn't affected by what you're drawing: 10 of the same or 10 different objects drawn at the same size use the same fillrate. Using the same object over and over helps save the overall memory use, which is a separate concern to fillrate.

  • Everything should work if you save valid JSON data. The problem stems from the fact you are using invalid JSON data, probably saved with an old build that had a bug.

  • I can't reproduce this in the latest Chrome version. Even switching between all the layouts very quickly for some time, memory usage in task manager is stable.

    If it reproduces only in NW.js, it's probably because the version most people use is well over a year old now, so it might be a bug in the Chromium engine that was fixed some time ago. We have an NW.js update to 0.13.0-alpha5 in the works so hopefully that will make a difference.

  • find() appears to be deliberately case insensitive, probably because most features in C2 are case insensitive too (e.g. matching animation names is case insensitive). I won't change it, since it could break compatibility with existing projects. For the next build I've added a new findCase() system expression which works identically but is case sensitive. I'll update the documentation to reflect this as well (which normally we do just before the next stable release).

  • Thanks, should be fixed in the next build.

  • I don't think this is a bug. Non-static local variables are reset to their initial value when they are "run" (i.e. imagine an "every tick" event in the same position resetting the variable). So first of all a local variable is not suitable for a countdown that needs to remember its old value. Secondly that message is just a debug note. I've removed it for the next build since it doesn't really matter. It's a bit unusual to access local variables from a trigger, because local variables only really "exist" for the normal run of evaluated events within their group - and triggers fire outside of that. So it's taking a "this variable doesn't exist" code path (which previously logged a debug message), and falling back to returning the initial value.

    Making the variable static also works, since that stops it resetting every tick and makes its lifetime permanent.

  • Try Construct 3

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

    Try Now Construct 3 users don't see these ads

  • I don't really understand the .capx. If I just run it, nothing happens, and I'm not sure if that's the problem or if the problem only happens when I create some extra objects as the instructions say. Turret and LOS are designed to do different things: LOS takes in to account obstacles between the two objects, but Turret does not, so it might just be down to that misunderstanding. Can you make a simpler demo?