Comments on: WebWait: Time your Ajax apps

By: Matthias

Matthias — Tue, 06 Mar 2007 16:16:24 +0000

before firebug was around, I always used Charles. It's a local proxy that lets you look into the data you download, and takes the time as well. Of course, the time the browser itself takes to calculate and display is missing here, but that's, in my case, pretty much negligible.

By: Michael Mahemoff

Michael Mahemoff — Wed, 14 Feb 2007 10:47:01 +0000

Guys, thanks for the feedback.

Tom, correct about perceived load time, ultimately that’s all what matters. In some respects, the IFrame is more accurate than other techniques, but not totally if the page is image-heavy as. WebWait works best for Ajax-heavy sites where JS processing is the bottleneck. The idea actually came to me while writing Quizr.com, where a large quiz can take a while to load due to all the event registration. I was using the FasterFox plugin for profiling, but wanted a consistent tool across browsers.

Firebug is awesome. Seconded, Thirded, Fourthed!!! With its feature set, there’s not much reason to use webwait over firebug if you have it installed. One benefit of WebWait right now is the ability to cut-and-paste a URL and try it in IE, or send it to a friend.

By: markus941

markus941 — Wed, 14 Feb 2007 00:15:05 +0000

Yeah, I’d recommend using Firebug – one less thing to install. Jonathan is right, the “net” tab is REALLY good for testing load speeds of website elements in detail.

By: Tom

Tom — Tue, 13 Feb 2007 20:28:42 +0000

This is a great idea, although after using it, it dawned on me that even this time isn’t the “actual” time a user has to wait.

For example, I pulled up TechCrunch in the tool and it told me 12 seconds. It didn’t actually take me, the end user, 12 seconds before I could start using an interacting with the site. A lot of that extra time was pulling down images that weren’t really interfering with my ability to use the site. In fact, some of the images were scrolled out of view, so they didn’t even matter. As a user, I wouldn’t even have noticed.

It’s a cool idea though, and I totally agree that tools like curl and httperf don’t tell the whole story.

By: Jonathan Aquino

Jonathan Aquino — Tue, 13 Feb 2007 20:16:19 +0000

Recommended for graphing times of individual page elements (images, ajax calls, css files, etc.):

– IBM Page detailer
– Firebug “Net” tab

By: Michael Mahemoff

Michael Mahemoff — Tue, 13 Feb 2007 19:40:00 +0000

(1) It will ignore any new asynchronous arrival of info (after all, that could go on indefinitely). All it does is time how long until the IFrame’s onload event fires, which means any XHRs fired off on page startup or later will be ignored.
(2) Sorry, I don’t follow what you mean by that?

By: Edward Miller

Edward Miller — Tue, 13 Feb 2007 19:10:44 +0000

This is interesting, but two questions:

(1) what happens in an AJAX app when there
is asynchronous arrival of information?

(2) does this work (we suspect not) when you want
to measure the total arrival times of individual page
components?