John Resig

"Once you're in edit mode just hit the 'Run' button to re-run it (right now the only way to "go back" to the formatted code mode is to reload the page - of course then you lose all your changes). Eventually I'd like to have each slide persist a little revision history (so you can revert back to the original slide - and be able to reload the page without losing your changes)." - John Resig

"Awesome, glad you found it to be useful. I suspect that this talk (in its current state) is most useful for people who've used JavaScript before. When I gave this talk at the Web 2.0 Expo it was pretty obvious that there were some in the audience who had never used it before, so I had to spend some time explaining why the syntax for object literals was different from other syntax used in the language - stuff like that." - John Resig

"Unfortunately there wasn't any video recorded of the talk (which was about 3 hours long). I could record something but I'm not sure how useful it would be (since it pretty much requires interaction with the people watching - having no feedback would make the presentation kind of dull, imo). Maybe I could give the talk as a ustream livecast and have people join in." - John Resig

"*Explained* by a 12 year old - an important distinction." - John Resig

"They are, double-click the examples. More details were provided in my blog post announcing this: http://ejohn.org/blog/adv-java..." - John Resig

"I think you may have mis-read me - obviously I can't really point someone to the candidate's web sites (since those are huge sources of bias) but rather unbiased third parties (such as domain experts) who are able to provide critiqued analysis of the positions that the candidates take. For example, the NPR interview with Jonathan Oberlander was excellent (based upon his article for the New England Journal of Medicine on the candidate health plans). I learned a lot about the positions and felt much more informed." - John Resig

"Yeahhh... Reddit is great and all - but unbiased is one thing that it's not, especially when it comes to politics." - John Resig

"Some links that I plan on passing along, so far: * [The Washington Post on the consumer tax plans](http://chartjunk.karmanaut.com...) * [Political scientist Jonathan Oberlander analyzing Health Care policies on NPR](http://www.npr.org/templates/r...) I want to have reasonable conversations with my family (who are still undecided) and so I'm looking for more information (beyond what you can find on the candidate's web sites). Unbiased third-parties would be preferred!" - John Resig

factcheck.org - Claude Betancourt

Also, have them watch/listen to the candidates' speeches from the conventions. My previously undecided mother-in-law did and she made up her mind. - Claude Betancourt

"Animations. By having animations actually be able to run faster the probability for smoother animations increases. Combine this with the "process per site" architecture (aka. less chance for a garbage collection cycle to occur) and I'm willing to lay down money that Chrome will have the smoothest JavaScript animations of any browser on the market." - John Resig

"Internet Explorer 8 is still in very, very, rough shape. This suite is actually quite good at testing the memory capabilities of browsers. I've been able to crash WebKit, Gecko, and Internet Explorer - some are just more responsive at fixing the crashes than others." - John Resig

"I'm confused - the Dromaeo v2 tests do include library independent DOM benchmakrs: DOM attributes, DOM modification, and DOM querying - none of which rely upon JavaScript libraries (they're just pure tests of the individual DOM methods/properties). I guess I could look at GQuery, but I'm not in a huge rush (wanted to get the major libraries tackled first)." - John Resig

"I didn't leave out the JavaScript tests, at all. I'm referring to the second version of the Dromaeo suite, which includes JavaScript, DOM, and JavaScript library performance tests: http://v2.dromaeo.com/ As it stands, v1 of Dromaeo is pretty much just a subset of the tests that are in SunSpider. It's better to just tests SunSpider and v2 Dromaeo, at this point." - John Resig

"Sorry everyone - server is on fire, working to put it out." - John Resig

"The V8 engine is really fast at function calls - it's not clear if they designed the test to emphasize that, or if they built the tests then built the engine to do good at it. Regardless of this fact - it's still a really fast engine - it's just not OMG crazy-go-nuts." - John Resig

"As I mentioned in the comments of the blog post: Correct, a process per-tab is the same as IE. But that is not the point that I was making: IE doesn't provide a Tab Process Manager - allowing you to monitor how much memory, CPU, and bandwidth your tabs are utilizing. *That* is the killer feature - not the "process per-tab"." - John Resig

"First an illustration, next, dancing!" - John Resig

"That was intentional - the game doesn't work in IE (IE doesn't have Canvas support) - no need to enable to code for people who won't be able to enjoy it!" - John Resig

"The only thing stopping me is the one gotcha: If jQuery is loaded dynamically (say via a dynamic script tag) then the last script on the page will be executed twice - which is hardly desirable. I don't have a good solution for this yet, but if I find one, I'll definitely lean strongly towards this." - John Resig

"Firebug uses the internal JSD service in Firefox - and it has... to put it gently... not aged well. It's been a long time since it's seen a serious update. We're [proposing a bunch of changes](https://bugzilla.mozilla.org/s...) that will help to modernize the service and give us better feedback (so that we, in turn, can actually help web developers)." - John Resig

"Firebug uses the internal JSD service in Firefox - and it has... to put it gently... not aged well. It's been a long time since it's seen a serious update. We're [proposing a bunch of changes](https://bugzilla.mozilla.org/s...) that will help to modernize the service and give us better feedback (so that we, in turn, can actually help web developers)." - John Resig

"I will be explaining a number of aspects of this selector engine in [my upcoming book](http://jsninja.com/) (I just finished the chapter on CSS Selector Engines). I'm not sure if I'll cover every technique that I use but I'll certainly be covering a majority (with a focus on writing cross-browser code). Thanks for the complements, I appreciate it!" - John Resig

"I will be explaining a number of aspects of this selector engine in [my upcoming book](http://jsninja.com/) (I just finished the chapter on CSS Selector Engines). I'm not sure if I'll cover every technique that I use but I'll certainly be covering a majority (with a focus on writing cross-browser code). Thanks for the complements, I appreciate it!" - John Resig

"a) Comparatively little. I consider this re-write to be the last one of any selector engines. The new Selector API (querySelectorAll) is going to make this piece of technology obsolete in a few years. That being said, there is still a noticeable percentage spent doing selectors. It varies from browser-to-browser and site-to-site but let's say a general ballpark of 25%. b) Yes. Although, it'll probably affect hardcore users more (the ones who have dramatically large sites with complex structures). That being said, I'm not just singlely-optimizing jQuery here. We've done some deep analysis of how jQuery is used on sites "in the wild" and have found other factors to be a much larger concern, namely: DOM manipulation (append, prepend, before, after), hide/show, and attribute methods to be some of the largest culprits. You can expect all of these to see some major love in the upcoming 1.3 release (in addition to the selector engine - if I can get it to a state that I like)." - John Resig

"a) Comparatively little. I consider this re-write to be the last one of any selector engines. The new Selector API (querySelectorAll) is going to make this piece of technology obsolete in a few years. That being said, there is still a noticeable percentage spent doing selectors. It varies from browser-to-browser and site-to-site but let's say a general ballpark of 25%. b) Yes. Although, it'll probably affect hardcore users more (the ones who have dramatically large sites with complex structures). That being said, I'm not just singlely-optimizing jQuery here. We've done some deep analysis of how jQuery is used on sites "in the wild" and have found other factors to be a much larger concern, namely: DOM manipulation (append, prepend, before, after), hide/show, and attribute methods to be some of the largest culprits. You can expect all of these to see some major love in the upcoming 1.3 release (in addition to the selector engine - if I can get it to a state that I like)." - John Resig

"I go in this order: Try to use querySelectorAll (IE 8, FF 3.1, Safari 3, Opera 10) then regular DOM with caching (FF 2-3, Opera 9), then regular DOM (IE 6-7). So yeah, no magic pixie caching dust for IE yet. The DOM modification events are fantastic here and IE doesn't have them. Right now the speed of the engine, in IE 6-7, is comparable to current jQuery - which is quite fast - just not 2-4x as fast. I won't release this engine until I can have an IE number that I'm proud of. I should note that the speed of the engine, without caching, is still faster than the major libraries - caching just makes it go "lightspeed" on these comparison tests." - John Resig

"I go in this order: Try to use querySelectorAll (IE 8, FF 3.1, Safari 3, Opera 10) then regular DOM with caching (FF 2-3, Opera 9), then regular DOM (IE 6-7). So yeah, no magic pixie caching dust for IE yet. The DOM modification events are fantastic here and IE doesn't have them. Right now the speed of the engine, in IE 6-7, is comparable to current jQuery - which is quite fast - just not 2-4x as fast. I won't release this engine until I can have an IE number that I'm proud of. I should note that the speed of the engine, without caching, is still faster than the major libraries - caching just makes it go "lightspeed" on these comparison tests." - John Resig

"D'oh! That's what I get for using publicly-viewable source control. It's definitely not ready yet (got some minor outlier bugs in the standards-compliant browsers - and a bunch of major bugs in IE still left to tackle) but the results are already promising. 4x faster in Firefox 3, 3x faster in Opera 9, 1.5x faster in Safari 3 than the other major JavaScript libraries. It's completely standalone (no library dependencies) and clocks in at under 4kb. I was just committing my code before moving on to work on IE - so beware. And yes, I expect this engine to become the new default selector engine in jQuery." - John Resig

"D'oh! That's what I get for using publicly-viewable source control. It's definitely not ready yet (got some minor outlier bugs in the standards-compliant browsers - and a bunch of major bugs in IE still left to tackle) but the results are already promising. 4x faster in Firefox 3, 3x faster in Opera 9, 1.5x faster in Safari 3 than the other major JavaScript libraries. It's completely standalone (no library dependencies) and clocks in at under 4kb. I was just committing my code before moving on to work on IE - so beware. And yes, I expect this engine to become the new default selector engine in jQuery." - John Resig

"[These comparisons](http://www.masonchang.com/2008...) were posted today (Squirrelfish compared to TraceMonkey)." - John Resig

"[These comparisons](http://www.masonchang.com/2008...) were posted today (Squirrelfish compared to TraceMonkey)." - John Resig