Rhindon's Forum Posts

  • In both C2 and C3, you open the instance variable list window and it, obviously, brings up the list of variables for that object. Adding a new variable opens yet another window. No big deal.

    But in the C3 list window, it's displayed in a kind of Excel-style grid. What if, instead of clicking to open a new window and filling out each field, you simply enter the new variable info directly into that list window - just make the bottom-most entry blank, ready to be filled in. Upon creating a new variable, a new blank entry field in that list is automatically made available in the "grid" list.

    One less window to pop up and, hopefully, more efficient.

    ALSO, it means you can clearly see - without the other window to obstruct it, at first (and yes, it's easy enough to move aside) - the whole list of existing variables. (Trust me, when you get a lot of variables as many of us do, it's easy to forget if we've created a needed variable for "this" or "that". Seeing the list readily at a glance makes it all that much easier...

    Additionally...

    • make the notes info that describes what each variable is for also show up in the variable list and not just in the drop-down list in the event tab.
    • right-click in Properties window on an instance variable to bring up mini-window to be able to edit name or delete variable.

    What do you think, Scirra Bros?

  • Try Construct 3

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

    Try Now Construct 3 users don't see these ads

  • I'm having a bit of an issue with this, too. Upon the original unveiling of the light/shadow caster feature (which I loved!), I seem to recall being able to create penumbras as well. I've reviewed both manual entries for the light object and the shadow caster behavior, but I can't get penumbras to appear (it might be the limitations of my computer...which is why I can't load the new Construct 3...

    A test site reveals that I can run WebGL 1, but WebGL 2 is possible to run on my computer but is either disabled or my processor (I don't have a dedicated graphics card on this laptop) isn't able to run it.

    Simply looking for more info that might help solve this issue. Anything I'm overlooking?

  • Thank you, both. I'll look up those links.

    The main issue with disabling collisions is that I still need a collision detection of sorts, regardless of which layer the player is on.

  • I realize it's not easy. That's why I'm asking for help.

    I know there is a solution - I can almost see it in my head. I just need help sorting out the details.

    I already figured out part of it but I can see there's still more to account for. It can be done.

    Already someone on one of the FB C2 groups suggested a workaround that doesn't involve using the Platform behavior for the enemies. So while more work, it potentially can solve my problem.

    So, while I appreciate your advice, as I said, it's not the answer I'm looking for.

  • Not really the answer I'm looking for.

    I don't have time to learn a new game editor...

  • I'm working on a game where you can switch between the foreground and background to avoid detection by guards. Doing this also allows the player to go behind or in front of solid objects (ie: a stack of boxes in the background can be passed in the front).

    Regarding just the player, this is not an issue.

    In order to avoid stopping the player from walking passed the boxes (or other objects that would normally have the Solid behavior), I can simply test if the player is on the same layer as the boxes. If not, I turn off the Solid behavior and the boxes are nothing more than sprites.

    The problem is when I include the enemies who also will use the Platform behavior and will need to be able to interact with objects that are supposed to be solid (regardless of the layer the player is on). Because if the payer is on a different layer than the solid object, then I'd have to turn off the Solid behavior. But if an enemy is on top of those solid objects, they'll fall through the sprites due to their Platform behavior...no Solid to hold them up.

    So I need to work out a different system of collision detection AND, more importantly, how to make "solid" objects act like solid objects without the solid behavior.

    So far, I have figured out, for example, how to use Overlap or Overlap At Offset to place the player on top of the boxes when he jumps up and falls to land on top of them. Problem is, if I walk up to the boxes from the side, the conditions still puts the player on top of the box. Part of that, I'm sure, just requires a couple extra conditions to test for the X position as well as the Y position of the player.

    Still, even with that understanding, I'm struggling to find the best way to test for these conditions in the same vein that the Solid + Platform behaviors do. May I get some input on this, please?

  • I've long-since deleted it. This was nearly four years ago.

  • No, that was the to ignore the GPU blacklist. For WebGL2 I had a separate flag.

    Well, in either case, both are enabled.

  • Try setting the WebGL2 flag to Enabled (it's Automatic on default).

    Already done. That was one of the first things I did - by your suggestion, if I recall.

  • Please check the steps in this bug report : https://www.scirra.com/forum/oops-webgl-is-not-supported_t189525.

    Consider posting your own results in the topic.

    Reporting issues for C3 should be done in the bug reports forums, not the general discussion forums, thanks.

    Sorry about that, KYATRIC. From the construct.net page, when it provided a link for reporting bugs, it took me to the general C3 discussion sub-forum. I didn't see the bug report forum for C3. I'll go ahead and repost this matter correctly.

    And thanks everyone!

  • Also, I tried loading Construct 3 in my Chrome mobile browser directly through construct.net. Despite triple-verifying that I do, indeed, have the (currently) latest version (56), the page still acts like I'm on a desktop (I guess that makes sense...) and asks me to update to version 57.

    I'm using the Google Pixel (Android version 7.1.1 with security patch level: March 5, 2017).

    EVERYTHING is as up-to-date as I can figure.

    I'm clueless about any Chrome settings regarding WebGL or my (non-existent) GPU.

    Granted, this is all beta, but I don't know if this is because it's beta or if it's my computer.

  • I had the same problem on my notebook. Chrome said only software rendering is enabled under chrome://gpu/ . I managed to solve it by turning on the #ignore-gpu-blacklist flag under chrome://flags/ .

    Thank you for the tip!

    Sadly, it didn't work. I engaged the #ignore-gpu-blacklist flag AND cleared the browsing history, cached items, cookies, etc. It still gives me the same error page I reported above.

  • While attempting to launch C3, I ultimately get a message that states,

    [quote:2y2km6x1]Oops! WebGL is not supported.

    WebGL provides high-performance graphics for web apps. Construct 3 needs WebGL support to run, but this browser does not appear to support WebGL.

    Try installing any available system updates, and check for an update for your graphics driver. Alternatively if you have access to a newer device try using that.

    I have ensured that I updated my Chrome browser to the latest available version at this time (Version 57.0.2987.110 (64-bit)) but still I either get the above message, or, on repeated attempts, a page that says the page doesn't even exist on the server according to the link that took me there (but this is seemingly random as other attempts will get me to the next page...but that's a separate issue with the Construct.net website). On the System Requirements Check page, however, it says, "Congratulations! As far as we can tell, you're able to run Construct 3!" So at this point, I'm not sure if it's my computer or a bug in C3 or the website, itself.

    If it at all helps, here are my basic computer specs...

    Acer Aspire 5349

    Windows 10 Home OS

    Intel Celeron CPU B800 1.5GHz

    8.00 GB RAM

    64-bit OS, x64-based processor

    (No dedicated graphics processor.)

  • Everything is assigned to a layer and the only object jumping layers is the MrSpy object. So nothing else should be changing.

  • My apologies if the topic heading is confusing. The issue I'm having is, indeed, confusing. I don't think I've encountered this EVER before.

    I'm working on a 2D side-scroller.

    I have three groups of layers - A, B, and C - that make up three layers of rooms - back, middle, and front.

    Each group is made up of four layers, themselves...

    • Background - force own texture (true), transparent (no)
    • Main - where the main character exists
    • Lighting - for all light-based objects (blend mode Destination out)
    • Foreground

    Up to this point, everything works and appears as it should.

    I've successfully made it so my main character, Mr. Spy, moves between the layer groups just fine.

    When switching between layers, if Mr. Spy goes back a layer group (such as from C to B ), then the layer closest to the player's vision is turned invisible (if moving from B to C, layer groups A and B are invisible), thus making the non-transparency and texture forcing a non-issue.

    So far, for testing purposes, I have a HangingLight object with a light cone in the B layer group only (on the Layer B Lighting layer). Going through either pair of doors - that lead one to Layer Group A or one to Layer Group C - is fine. But when moving BACK TO Layer Group B, where the lights are, the lights are in effect, and yet they cannot be VISIBLY seen like when you first run the game. Meaning, you can see the EFFECT of the light cone making anything overlapping the light object more visible, just like normal light, but you can't actually see the white transparency of the light object, itself.

    I am at a complete loss as to why this is.

    CONTROLS: Arrow keys to move, A on a door to open

    https://dl.dropboxusercontent.com/u/36080515/MrSpy%20beta%2002.capx