• Content Count

  • Joined

  • Last visited

  1. Works on desktop, crashes on Mobile. There's a small chance it's due to using Scala JS, but had no issues prior to changing alpha values. Update: I forced the renderer in the newly uploaded version to be canvas. This fixes the issue. Thus, the only difference between the version that hard crashes the mobile android browsers and the one that works fine is the choice of autodetect versus canvas. I'm going to take the demo down for a bit and will put up a simplified a/b later. Thanks.
  2. I get straight up browser crashes when I do transparency on graphics objects with current WebGL on Samsung Galaxy Note 3. Happens in both firefox and chrome, and can be replicated on samsung galaxy s3.
  3. Hi All, the latest build of Pixi (both dev and regular versions) seems to break Chrome on my Android Phone (Galaxy Note 3). As well, the Samsung stock android browser remains broken. I've tried using both 'getAnimFrame' and 'setTimeout' but doesn't seem to make a difference. Any idea what's causing this behavior? Works fine in: Mobile Firefox Safari Chrome on PC Firefox on PC Thanks, T.
  4. Thanks Rich, I put in your suggested fix and it did indeed solve the issue.
  5. Hello, I created a little routine that shows the bounding boxes of a circle and a bit of text I've rendered inside it. Whenever I turn "wordWrap" on, I get an extra line on the text. Is this behavior by design? If so, could we get a way of detecting how large the blank line is so we can recenter vertically, as needed? I can kind of fake it by remembering the original line height, but that doesn't account for spacing. Cheers, Travis
  6. Ok. Updated version here: Runs buttery smooth on everything except non-Chrome Samsung Internet browser, in which it exhibits a series of 'trails' and strange rendering artifacts. Not sure why the Samsung browser isn't clearing the screen properly now, but it looks like removing GSAP solved the main (lag related) issue. Thanks! Travis
  7. Interesting. I'll switch to a hand coded tween and see if it improves.
  8. Hi there, I was tremendously impressed with the performance of the PixiJS example applications on my various mobile phones and tablet devices. I have not had nearly that level of performance when using KineticJS. Accordingly, I decided to make my own stress test using the starter tutorial on Pixi and the GSAP tweening framework. The stress test is here: Performance is mostly buttery smooth in Desktop Chrome and Firefox, with minor (OS related?) herky-jerkiness on occasion. Firefox on Android is pretty slow (not a huge surprise). Chrome on Android is very inconsistent (surprising - Using a Samsung Note 3). Safari on the Apple 5c is buttery smooth. Is this a fair test? Or did I do something really suboptimal from a coding standpoint? It's true I'm not using the minified versions of anything, and the Pixi version is Dev. If it is a fair test, I'm curious as to what I'm doing that puts the performance so far outside of the example apps on Android. Any help is greatly appreciated. Cheers, Travis