Comments on: Performance is in the Details http://briancrescimanno.com/2008/05/29/performance-is-in-the-details/ Thoughts on Ajax, Flex, and Rich Internet Applications Thu, 04 Dec 2008 03:05:37 +0000 http://wordpress.org/?v=2.6.3 By: diyism http://briancrescimanno.com/2008/05/29/performance-is-in-the-details/#comment-244 diyism Thu, 21 Aug 2008 07:52:07 +0000 http://briancrescimanno.com/?p=10#comment-244 Why not use "jQuery Live Bind": $.fn.is_match=function(selector) {selector=selector.split(/\s /).reverse().join(' ') .split(' > ').join('").parent("') .split(' ').join('").prev("') .split(' ~ ').join('").prevAll("') .split(/\s /).join('").parents("'); return eval('this.filter("' selector '").length'); } $.live_bind=function(selector, etype, fn) {$(document).bind(etype, function(event) {var event=event || window.event, src_ele=event.srcElement || event.target; if ($(src_ele).is_match(selector)) {fn(); } } ); } hello2 hello3 hello1 hello hello2 hello3 hello1 function alr(){alert('helloooo');} $.live_bind(".kkkk .ppp1 .ppp[title='333']", 'click', alr); $('.kkkk').append('hello'); Why not use “jQuery Live Bind”:

$.fn.is_match=function(selector)
{selector=selector.split(/\s /).reverse().join(’ ‘)
.split(’ > ‘).join(’”).parent(”‘)
.split(’ ‘).join(’”).prev(”‘)
.split(’ ~ ‘).join(’”).prevAll(”‘)
.split(/\s /).join(’”).parents(”‘);
return eval(’this.filter(”‘ selector ‘”).length’);
}
$.live_bind=function(selector, etype, fn)
{$(document).bind(etype,
function(event)
{var event=event || window.event,
src_ele=event.srcElement || event.target;
if ($(src_ele).is_match(selector))
{fn();
}
}
);
}

hello2
hello3
hello1
hello

hello2
hello3
hello1

function alr(){alert(’helloooo’);}
$.live_bind(”.kkkk .ppp1 .ppp[title='333']“, ‘click’, alr);
$(’.kkkk’).append(’hello’);

]]> By: Brian http://briancrescimanno.com/2008/05/29/performance-is-in-the-details/#comment-45 Brian Sat, 31 May 2008 13:20:59 +0000 http://briancrescimanno.com/?p=10#comment-45 Here's what's interesting (in response to to Antonelli's post above): My (admittedly non-scientific) testing has shown that IE6 will still lose its lunch over this loop structure: for(var i=0, len = $$(’div.contents’).length; i<len; i ){ /** .. */ } Yes, it *should* work as the value of the length should be cached--but everything that I've seen from what IE6's JScript parser is that is NOT the case. Here’s what’s interesting (in response to to Antonelli’s post above):

My (admittedly non-scientific) testing has shown that IE6 will still lose its lunch over this loop structure:

for(var i=0, len = $$(’div.contents’).length; i

Yes, it *should* work as the value of the length should be cached–but everything that I’ve seen from what IE6’s JScript parser is that is NOT the case.

]]> By: Neil http://briancrescimanno.com/2008/05/29/performance-is-in-the-details/#comment-44 Neil Sat, 31 May 2008 12:20:34 +0000 http://briancrescimanno.com/?p=10#comment-44 Ah, Little Johnny Toggler and his wonderful for loop. The overaching problem (IMHO) isn't just DOM traversal, but a gross lack of understanding of the way IE6 (and to a lesser degree IE7) looks up pointer references in JavaScript. Man, we need team blog/discussion forum. Ah, Little Johnny Toggler and his wonderful for loop.

The overaching problem (IMHO) isn’t just DOM traversal, but a gross lack of understanding of the way IE6 (and to a lesser degree IE7) looks up pointer references in JavaScript.

Man, we need team blog/discussion forum.

]]> By: Brian http://briancrescimanno.com/2008/05/29/performance-is-in-the-details/#comment-42 Brian Fri, 30 May 2008 23:34:52 +0000 http://briancrescimanno.com/?p=10#comment-42 $$ has been heavily optimized since it first appeared in the Prototype framework so I wouldn't be too afraid of using it; however, it does not make up for this code. Tell me who did this so I can smack them. Also, to improve performance, remove the life of a variable (letting garbage collection remove it sooner too) do this: for(var i=0, len = $$('div.contents').length; i<len; i ){ /** .. */ } Sorry for trolling your blog, I got bored and saw the link from your LinkedIn profile. :) $$ has been heavily optimized since it first appeared in the Prototype framework so I wouldn’t be too afraid of using it; however, it does not make up for this code. Tell me who did this so I can smack them. Also, to improve performance, remove the life of a variable (letting garbage collection remove it sooner too) do this:

for(var i=0, len = $$(’div.contents’).length; i<len; i ){ /** .. */ }

Sorry for trolling your blog, I got bored and saw the link from your LinkedIn profile. :)

]]> By: Brian http://briancrescimanno.com/2008/05/29/performance-is-in-the-details/#comment-39 Brian Thu, 29 May 2008 15:38:17 +0000 http://briancrescimanno.com/?p=10#comment-39 Very interesting post on your site, Eric. You are right in your assessment that the short-hand nature of the $() method can encourage poor practices. Thanks for the feedback! Very interesting post on your site, Eric. You are right in your assessment that the short-hand nature of the $() method can encourage poor practices. Thanks for the feedback!

]]> By: Eric DeLabar http://briancrescimanno.com/2008/05/29/performance-is-in-the-details/#comment-38 Eric DeLabar Thu, 29 May 2008 15:09:12 +0000 http://briancrescimanno.com/?p=10#comment-38 At my employer we're dealing with some similar performance problems, as much as I love prototype, in the wrong hands its a very dangerous tool. I suppose it's a case of learning by scraping you fingers with a hand tool before you cut them off entirely with a power tool. If you're interested, my link goes to a post I wrote on my website talking about a similar performance problem with the '$' function. It also includes some performance metrics on a few different browsers. At my employer we’re dealing with some similar performance problems, as much as I love prototype, in the wrong hands its a very dangerous tool. I suppose it’s a case of learning by scraping you fingers with a hand tool before you cut them off entirely with a power tool.

If you’re interested, my link goes to a post I wrote on my website talking about a similar performance problem with the ‘$’ function. It also includes some performance metrics on a few different browsers.

]]>