Skip to content

Scott Lembcke slembcke

slembcke commented on pull request cocos2d/cocos2d-iphone#880
Scott Lembcke

Looks good to me. No more matrix nitpicks. ;)

slembcke commented on pull request cocos2d/cocos2d-iphone#880
Scott Lembcke

:p Ok. I was kind of hesitant to bring it up since it's sort of a dumb nitpick.

slembcke commented on pull request cocos2d/cocos2d-iphone#880
Scott Lembcke

This is a little pedantic, but why not use GLKMatrix4MakeAndTranspose() instead? Then you don't have to reverse the multiplication order nor mix tr…

slembcke commented on pull request cocos2d/cocos2d-iphone#879
Scott Lembcke

Otherwise looks fine. The color matrix can be added later.

Scott Lembcke
slembcke merged pull request cocos2d/cocos2d-iphone#879
Scott Lembcke
Fix effects in OSX project. Add CCEffectHue.
7 commits with 235 additions and 11 deletions
slembcke commented on pull request cocos2d/cocos2d-iphone#879
Scott Lembcke

Yes, RGB(1,1,1). It works okay if you don't care about preserving luminance. It's very simple if luminance is not important anyway.

slembcke commented on pull request cocos2d/cocos2d-iphone#879
Scott Lembcke

Looks like yes. I was wondering how that would avoid oversaturation, but you can only avoid it if you use a non-linear color space like HSV (which …

slembcke commented on pull request cocos2d/cocos2d-iphone#879
Scott Lembcke

You could apply a 2x2 rotation matrix calculated on the CPU here instead and save a lot of expensive trig calls. Actually... since the RGB <--> YIQ…

slembcke commented on pull request cocos2d/cocos2d-iphone#876
Scott Lembcke

I'm always a little leery of using conditionals in fragment shaders. On pre-A7 devices I've done something like this: vec2 compare = 0.5 - abs(refr…

Scott Lembcke
slembcke merged pull request cocos2d/cocos2d-iphone#873
Scott Lembcke
CCEffects - Designated initializer cleanup, fix nil environment in refraction
3 commits with 37 additions and 51 deletions
slembcke commented on pull request cocos2d/cocos2d-iphone#866
Scott Lembcke

You don't have to explicitly reset the render state. If the CCNode.shaderUniforms handles that for you if creating a custom uniform dictionary inva…

slembcke commented on pull request cocos2d/cocos2d-iphone#866
Scott Lembcke

Ah good. I've been thinking about making this change for a while, but was never needed since the second texture coordinate set hasn't been used for…

slembcke commented on pull request cocos2d/cocos2d-iphone#871
Scott Lembcke

If you added a min(src, dst) blending mode you could have a "gloom" filter too.

slembcke commented on pull request cocos2d/cocos2d-iphone#871
Scott Lembcke

Copy paste error from the comment below I assume?

Scott Lembcke
Scott Lembcke
  • Scott Lembcke 2ddcc43
    Make uniform dictionary copying behavior of CCRenderState explicit fo…
  • Scott Lembcke 7d242d3
    Merge branch 'v3.1' of github.com:cocos2d/cocos2d-iphone into v3.1
  • 1 more commit »
Scott Lembcke
Body's mass and moment must be positive to simulate.
Scott Lembcke

It's a bug with upgrading 1.0 projects. The workaround is to open up your physics object, make a throwaway change (like changing the corner radius,…

Scott Lembcke
Scott Lembcke

Unfortunately because it's inherited from CCNode, we can't remove or deprecate it. We could change it's behavior in a non-backwards compatible way …

Scott Lembcke

!! Somebody reported this as a bug with Apportable although I was never able to reproduce it after a fair amount of effort. I wonder if it's relate…

Scott Lembcke

Ooh. That is a really sbutle change. *facepalm*

Scott Lembcke

Ugh. So it turns out the 3.1 behavior is actually "correct" for what it's worth. If they were allowed, a negative threshold would draw everything. …

Scott Lembcke
  • Scott Lembcke 080fc6b
    Using a non-symmetrical image for the progress node tests.
Scott Lembcke
Scott Lembcke

Some of this conversation took place off of GitHub. So it's not entirely a one sided conversation. :p I ended up fixing this in a minimal way inst…

Scott Lembcke
CCRenderTexture cleanup.
Scott Lembcke
Something went wrong with that request. Please try again.