3 ms·
I find it very interesting that he says asm.js was "discovered" rather than "invented". Clues to a person's philosophy.
by Torgo 11y ago
I find it very interesting that he says asm.js was "discovered" rather than "invented". Clues to a person's philosophy.
- ravishi 11y agoI read that as a joke. Wasn't it?
- azakai 11y agoI think that's similar to how people say JSON was discovered by Crockford. Both only use things already present in the language, but they do form something new out of them. It is sort of a middle ground between discovery and invention.
- TazeTSchnitzel 11y agoMoreover, asm.js effectively already existed, right? It's not quite the same, but it's essentially a formalisation of what emscripten and the like already produced.
- Torgo 11y agoIs there anything written that explains this? I sort of got that impression when he talked about how JS was untyped but the js bytecode vm had types.
- azakai 11y agoModern JS VMs have type-specializing JITs. That's how they make code run fast, basically - they detect types at runtime, and emit efficient code for that. asm.js just gives them the types, which makes things easier, but it all builds upon the existing power of JS VMs. See e.g. http://kripken.github.io/mloc_emscripten_talk/sotu.html#/16 http://kripken.github.io/mloc_emscripten_talk/sotu.html#/16
- azakai 11y agoYes, correct. Emscripten and mandreel already produced something quite like asm.js, before asm.js was defined.
- tambourine_man 11y agoI thought it was a joke
- snarfy 11y agoIt's a minor quip. Are numbers real? Would you say we discovered them, or invented them? Do ideas exist even if you haven't thought of them yet? If so then they are discoveries.
- thousande 11y agoEverything is a discovery. No one sat down and said now I am going to invent or think up something called numbers. A thing happens because of what happened before it, and before that thing another thing happened etc.