Juriy Zaytsev

kangax Well that was quick. Here's a more or less perfect explanation of Javascript quiz posted earlier — http://bit.ly/afis0D (thanks @cmsalvado )

kangax Today's browser stats on a blog show that FF 3.6 already surpassed previous versions. Nice update speed, but not quite as fast as Chrome.

kangax Javascript quiz (http://perfectionkills.com/javascript-quiz/). What's your score?

kangax ES5 is slowly spreading around. FF 3.7 now has `Object.create` (also in WebKit) and `Array.isArray` (also in WebKit & Opera).

kangax Validome appears to be a more intelligent (X)HTML validator, comparing to W3C one or others — http://j.mp/b1ObBU

kangax And @jdalton comes up with a version 41 chars less than mine :) (safer version — 25 less) http://gist.github.com/296128 Thanks, everyone!

kangax Shortened @rem 's html5 enabling script down by 29 chars (http://gist.github.com/296114). Can we do better?

kangax Garrett came up with a brilliant workaround for Safari 2 inability to fire "load" on SCRIPT elements— use image instead (http://j.mp/9slTco)

kangax It's clear now that JSLint is not fully conforming ES parser. Fails: `x=/x/g//f` `x=1,2` `x=1/ /x/` `x=void 0` `throw x=1` `var \u005F=1`

kangax Completely forgot about Function Statements (thanks @abozhilov ), which explain Firebug's hoisting behavior (http://bit.ly/8OMnY).

kangax Firebug console doesn't hoist function declarations. `alert(typeof f); function f(){}` alerts "undefined". Lovely.

kangax Speaking of `arguments` performance, here's a simple example of how bad it is: http://j.mp/c50aqb (in at least Safari, Opera and FF <3.6)

kangax It really seems like JSLint can't parse some valid ECMAScript programs. E.g.: `1/ /x/` terminates scanning, instead of returning `NaN`.

kangax Nice overview of Javascript speedups in FF 3.6 (http://bit.ly/bU8tSD). Btw, `arguments` are still very slow in many browsers.

kangax There's a lot of ECMAScript parsers, but none written in ECMAScript itself? Narcissus doesn't count, as it relies on SpiderMonkey extensions

kangax People seem to _love_ unrelated inference. Two popular client-storage libraries branch to IE's userData when ActiveXObject is present. Ugh.

kangax We need smth. like PPK's compat. tables, but open sourced, to keep content up to date. 1/2-yr-old info is not too useful: http://j.mp/dyxKhR

kangax ES committee is not planning to release HTML version of specs (http://bit.ly/98XbPK). I guess we'll have to do it ourselves :)

kangax Sad to see people leave comp.lang.javascript. If you don't participate, at least read the archives (that's how I truly learned most of JS).

kangax .@philogb If Im not mistaken, Object.keys returns _enumerable_ properties only, whereas Object.getOwnPropertyNames — _all_ own properties.

kangax Too bad ES5 `Object.keys` (already in Chrome and WebKit) differs from Prototype's counterpart (ES5 one returns only _own_ properties)

kangax And doesn't respect overwrites, transforming `Array = customArray; alert(new Array());` to `Array = customArray; alert([]);` :/

kangax Closure Compiler is still full of oversights. Transforms `new Array()` to `[]` but not `Array()` to `[]` or `new Array(1,2)` to `[1,2]`

kangax Btw, shortest valid HTML 4.01 document is "