Rob Morris

Members
  • Content Count

    6
  • Joined

  • Last visited

About Rob Morris

  • Rank
    Newbie

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. Hey trsh, I took a peek. Verified that the dirty counter is updating properly, tried messing around with starting and stopping the underlying video to see if that would help. No dice. The problem seems to be some kind of race condition, where seeking the underlying video results in the data copy failing if performed too quickly. Change your sample's change handler to: fs.onchange = function(e){ videoEl.fastSeek(e.target.value); setTimeout(function() { baseTexture.update(); }, 100); } And it will work as expected. I'm not sure what the solution would be should you be wanting real-time response times, but at least this should help narrow down the problem.
  2. Thanks, I'll give one of those other font loading libraries a try... Frustrating heisenbugs are the best!
  3. Hey all! I'm using Pixi.js in my edutainment platform, and it rocks. We're about to go live, and in our testing have found an intermittent issue with how text in Pixi.Text elements is rendered. Context: we're using a webfont loaded using Google's webfont.js like so: <script src="https://ajax.googleapis.com/ajax/libs/webfont/1.6.26/webfont.js"></script> <script> WebFont.load({ custom: { families: ['Short Stack'] } }); </script> This produces a synchronous load, and is run before we bootstrap up Pixi and our framework/game. This seems to be working fine; I include it purely for context. The problem is that very seldomly, our Pixi.Text elements render with the text offset vertically by ~5 pixels. When this happens, ALL our text is offset. Doing a hard refresh (ctrl+shift+R on Firefox) resolves the issue, usually. See the attached screenshots for what it looks like when it borks. So! Any ideas what could be causing this issue? I know folks have had issues with load timing in the past, but the webfont is clearly loaded by the time the game renders - the font is correct in both versions. There's zero code changes between the working and offset versions - just a hard refresh. Help?
  4. Thanks Ivan, that's very helpful. I'll go refactor my main app and see if I can't improve things. Much appreciated!
  5. Any pointers on how to skip rendering cleanly? I like the idea, and given my architecture (where everything is wrapped by accessors) it would be easy enough to set a global "dirty" flag, but I'm not clear on how to explicitly skip a render using the PIXI.Application model. Where can I trap the "I'm about to render" event/call?
  6. Yeah, I'm really struggling with interaction right now. I implemented drag'n'drop, and it works great for some objects, and breaks for others. It seems to break based on the parent chain of the object being dragged - I stop getting pointermove events over non-interactive objects and other oddness. I guess I'll go read through the source, but it sure would be great if there were some better documentation on how interaction is supposed to work. The event descriptions leave a lot to be desired!