Hacker News Reader

Comment by arkitaip | Hacker News Reader

I am incredibly thankful that jQuery exists. The other day I had to code some basic javascript in a environment where embedding jQuery wasn't an option (too much overhead) - boy what a PITA. Basic stuff like writing selectors took forever and I had to resolve to "low level" coding. What would have taken a couple of seconds in jQuery took several hours of research, and trial and error. I planned on adding a basic fading/sliding effect - forget about it. Takes to much time in ordinary javascript.

So thanks John Resig and the rest of the jQuery dev team for making us web dev guys more productive and actually making JavaScript fun.

View on HN

On the other hand... congrats! Now you know more about the inner-workings of jQuery and are a better javascript programmer.

Question is for how long. I've written several thousand lines of native, cross-browser javascript some years ago and from time to time again. But i don't remember a lot of the workarounds, from the top of my head.

Fiddling with cross-browser support is something that my brain don't like to keep today, be it js or css. Things like jQuery for js and compass for css come in handy to brush over that. :)

Remember that jQuery's selectors can be used separately: they're a wrapper around the Sizzle library. It's 5kb after minification and gzip.

http://sizzlejs.com/

Can't speak to the performance relative to JQuery or to raw javascript however.

jQuery too much overhead? What kind of environment was that?

Pretty much any environment where you care about performance of your code, actually. A lot of the simple-looking jquery stuff is implemented with tons of indirection, branching, etc.

Just compare the work a browser has to do for something like:

  document.getElementById("foo").setAttribute("bar", "baz");

to the work needed for the jquery equivalent...

Now the good thing is that most consumers of jquery don't care about performance of their code (in that they're doing small enough amounts of stuff it doesn't matter much).

Out of curiousity I tested that particular statement, and you're right. Compared to

    $('#foo').attr('bar', 'baz');

your code is

30 times faster in Chrome 9

81 times faster in Firefox 3.6

Could you share your test code?

In most cases, though, using jquery to do something like adding a class, or setting a value, is still plenty fast enough. We're talking about code that runs in the browser, where things move on human timescales. Don't optimize prematurely.

The problems start when you add a class to 10,000 elements on the page. And yes, I've seen pages doing this with jquery where the jquery overhead was what made the difference between a 20ms operation (not user-perceptible) and a 1s operation (very much user-perceptible).

47 times faster in Firefox 3.6.1 7 times faster in IE 8

  > Now the good thing is that most consumers of jquery don't
  > care about performance of their code

Many do, but more do care about cross-browser compatibility and when you need to support IE6,7 home-rolled code is not likely to cut all the research and effort that's embedded in jQuery.

Bookmarklets

Reply on news.ycombinator.com