51 ms·

Mozilla are leading the way in other areas. Firefox OS is entirely based on HTML5 - even the UI. Mozilla invented asm.js, an open, backwards-compatible and fasc

Mozilla are leading the way in other areas. Firefox OS is entirely based on HTML5 - even the UI. Mozilla invented asm.js, an open, backwards-compatible and fascinating alternative to Google's proprietary Chrome-only Native Client technology. They've done a lot with emscripten to compile large native engines to the web, complete with WebGL and Web Audio support. Chrome looks to follow Mozilla's lead by adding asm.js support to V8. I agree Firefox are overall more behind, but they are still an innovative company coming up with some fantastic tech. The main reason they're behind is probably because their codebase is pretty old, back from the days when they were busy saving us from IE6.

dragonwriter 13y ago
> Mozilla invented asm.js, an open, backwards-compatible and fascinating alternative to Google's proprietary Chrome-only Native Client technology. Native Client isn't proprietary, its open -- no other browser vendor has adopted it, but that doesn't change the fact that its open. Backward compatible is a real issue with asm.js vs. NaCl, there is no reason to obscure the real issue by throwing a fake one in with it.
- AshleysBrain 13y ago
  Oh, fair dos. But I think asm.js is so much of a better idea that there's no reason to support NaCl - asm.js does everything it's supposed to do but better. I think it was a pretty nice move from Mozilla.
  - dragonwriter 13y ago
    > But I think asm.js is so much of a better idea that there's no reason to support NaCl - asm.js does everything it's supposed to do but better. A key focus of NaCl and asm.js is performance, and IIRC most benchmarks are showing NaCl with somewhere between a 25%-33% additional overhead over native C, and asm.js with somewhere in the neighborhood of 100% additional overhead, so I don't think I can quite agree that asm.js does everything NaCl is supposed to do but better. That said, the other advantages are there, and the performance gap may close.
    - pcwalton 13y ago
      Those statistics are misleading. The asm.js numbers count compilation time, while the PNaCl numbers don't.
      - dragonwriter 13y ago
        If you've got statistics that show asm.js is equal or better including compilation time, I'm sure lots of people would like to see them.
- josephlord 13y ago
  Can one company change the next version of the spec without the collaboration of other users? If they can then it is proprietary even if it is open. Was it developed by a collaborative effort across the market or was it dropped as a complete solution that they then tried to get approved by a standards body? Then it is probably proprietary even if open to everyone to use.

dragonwriter 13y ago

> Mozilla invented asm.js, an open, backwards-compatible and fascinating alternative to Google's proprietary Chrome-only Native Client technology. Native Client isn't proprietary, its open -- no other browser vendor has adopted it, but that doesn't change the fact that its open. Backward compatible is a real issue with asm.js vs. NaCl, there is no reason to obscure the real issue by throwing a fake one in with it.

AshleysBrain 13y ago
Oh, fair dos. But I think asm.js is so much of a better idea that there's no reason to support NaCl - asm.js does everything it's supposed to do but better. I think it was a pretty nice move from Mozilla.
- dragonwriter 13y ago
  > But I think asm.js is so much of a better idea that there's no reason to support NaCl - asm.js does everything it's supposed to do but better. A key focus of NaCl and asm.js is performance, and IIRC most benchmarks are showing NaCl with somewhere between a 25%-33% additional overhead over native C, and asm.js with somewhere in the neighborhood of 100% additional overhead, so I don't think I can quite agree that asm.js does everything NaCl is supposed to do but better. That said, the other advantages are there, and the performance gap may close.
  - pcwalton 13y ago
    Those statistics are misleading. The asm.js numbers count compilation time, while the PNaCl numbers don't.
    - dragonwriter 13y ago
      If you've got statistics that show asm.js is equal or better including compilation time, I'm sure lots of people would like to see them.
josephlord 13y ago
Can one company change the next version of the spec without the collaboration of other users? If they can then it is proprietary even if it is open. Was it developed by a collaborative effort across the market or was it dropped as a complete solution that they then tried to get approved by a standards body? Then it is probably proprietary even if open to everyone to use.