sqiddster's Forum Posts

  • You forgot to add your example ;)

  • boolean I'm almost 100% sure every browser with webgl support has per-layout memory management.

  • R0J0hound

    I see what they did there :P

  • Whiteclaws, different levels should have different effects, and I can't prerender a whole bunch of levels.

  • Oh... that's not good news.

    Ashley , is there anything at all we can do about this? My game pretty much hinges on this being added. How complex is this architecture change?

  • 'Bump' This thread disappeared a little too quickly. My fault for posting in How Do I.

  • Apparently, Google's name is a misspelling of Googol, which means 1 followed by 100 zeros.

    So yeah, Google is actually a typo :D

    There's a random fact of the day!

  • That definitely sounds better, if a bit generic. Something unique yet easily spellable/googleable is ideal.

  • XManBG

    Someone already finished it just for you!

    scirra.com/forum/plugin-polygon-update-20130218_topic62075.html

  • C2's library of shader effects is excellent, however it presents one major limitation - We cannot make an effect apply across multiple layers.

    Let's assume we are making a game with a few parallaxed background layers and a few layers for the foreground - perhaps a lighting layer, layer for objects in front of player, etc.

    This very common layout has a few very big problems when webGL effects come into the picture.

    Let's consider the problem with layer-wide effects. Perhaps we want to adjust the HSL of the foreground and the background to simulate a night atmosphere. As Construct 2 is at the moment, we have to apply redundant effects to every single layer of the game in order to achieve this. This may not seem like such a problem, however it has huge performance ramifications that only really show their faces at high resolutions. I've done testing in a game with about 8-10 layers and having 10 layer effects is crippling in fullscreen, where the same game with a couple of layout effects runs very smoothly at the same resolution.

    A nice way to fix this problem would be to let us have 'multi-layer' effects. Layout effects as-is are pretty much useless for any game with a UI as-is anyway, so that's not really a valid option.

    Now, I would be very happy to find that there is a way to write my own shader that does this however I feel like this specific problem is hardcoded into the editor.

    Thanks for listening! Let me know if you think this is an important subtle change.

  • e1two , unfortunately I doubt Scirra can spend time ATM making C2 work better with something it's not designed to do.

  • This may not actually be possible in firefox using HTML5, period. I see the workaround that you tried to use, but I have a feeling that it would give you problems.

  • boolean, unless you understand integral calculus it's very hard to explain.

    The best explanation I have seen is

    Subscribe to Construct videos now
  • blackhornet, trying to fix, no idea why it's happening.

    blackhornet, fixed it partially.

    Heh, apparently you can't link to a file with an apostrophe.

  • Try Construct 3

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

    Try Now Construct 3 users don't see these ads
  • I recently heard about an interesting mathematical experiment called Buffon's Needle.

    Basically, it suggests that if you drop enough needles (or matchsticks) of length l on a wooden floor with planks of width 2l, you can take the ratio of total number of matchsticks dropped to the number of matchsticks that cross the crack. This ratio should eventually equal pi (3.14159265).

    I thought I'd write a C2 program to simulate the proposition, and it worked pretty well.

    EDIT: dropbox hyperlinks aren't working for some reason, so try copying and pasting these links.

    Here's a demo.

    Here's the capx.

    If you let the calculation run for a little while, you should see the ratio value approach pi. For a variety of reasons it won't be perfect but it gets pretty darn close.

    It's a very interesting proposition, and its solution goes into the depths of probability calculus.

    So, there's your uncomfortable math truth for the day!