I know JQuery is free but there is no official standard of widgets, and it seems for ‘hacker-level’ developers – not me. I just want to use them as building blocks.

Last weekend, I found another framework ZK Light that looks great. Switching framework is painful. Does anyone have experiences of using it?

But what features or performance related implementations are you eluding to? I think you’ll find many reliable tests here http://yura.thinkweb2.com/cft/. And just look at what jQuery did with version 1.3 – the features they were able to detect. And it isn’t just detecting features but working around them such as with jQuery’s ready function.

It isn’t an elegant solution, but it’s a solution none the less. The fact of the matter is ExtJS doesn’t even try, electing instead for the quick shortcut. In my book that is of far greater concern than performance, all it took was an upgrade to one of their “supported browsers” to learn that lesson.

While I mostly agree with your reasoning (I have just read your well written article) you seem to forget that not everything can be feature detected reliably and you entirely miss the point that ua detection is not only about testing for features – differences in performance between browsers cannot be feature detected.

There are therefore areas where it is unfortunately still neccessary to detect the browser.

I’ve been using ExtJS for a few months now to develop large web apps and I am very impressed by it’s possibilities. However when a customer wants a simple webpage or just a little animation i use jQuery (UI), because it’s more lightweight and less complex. So to Nosredna I would say that comparing the usage percentage of jQuery to that of ExtJS is just not right.

And about the open source discussion: the GPL is very much an open source license, so you can use ExtJS for free, just as long as you open source your project (which shouldn’t be a problem for developers that seem to want every good project to be open source and use it for free…). And if you don’t want to open source your code, you can always but a license, and I must say that they are not very expensive, regarding the quality of thet ExtJS library.

They may be doing “alright,” but what percentage of web pages use ExtJS and what percentage use jQuery?

I think they’ve baffled, if not irritated, many JS programmers along the way.

At this point JQuery continues to trail blaze, and while ExtJS enjoyed that opportunity for while thanks to a liberal license, they slapped the community in the face for their help, and they’ve lost that mindshare.

$(‘#extjs’).remove();

SmartClient is available under LGPL (free!). Also make sure you check out v7.0 of SmartClient. It is much more polished that the previous version. Demo : http://www.smartclient.com/smartgwt/showcase/

Cheers.

I’ve been using ExtJS for a few weeks now. Sometimes the docs are lacking but I’m impressed with how much I can do with little energy. I haven’t used SmartClient, but my initial impression from their website is that it is very enterprisey…that is to say…ugly and slow (hrm, maybe they have a Java background). Compare the demos on ExtJS’s website and SmartClient’s website. ExtJS seems faster, smoother and scales (text size) better.

SmartClient seems to have more features. I guess it’s just a matter of “are these extra features the ones we need”.

Also ExtJS is dramatically cheaper!

How does ExtJS fares against SmartClient ?

/Happy Ext camper

I fear how library authors might respond with year numbers, angles or roman numerals; jQuery2009? dojo360? MooToolsX? ;P

(Just for the record – I’m a mootooler – that’s my bias)

However, because of the GPL restrictions, it’s very easy to tell an enterprise that if they really don’t want GPL they can buy out of it. For most people, GPL isn’t a problem, and for enterprises (who most likely have the cash to pony up) it’s a lot cheaper than many closed source solutions to buy out.

I think it was realized very late for ExtJS and the change left a bad taste in a lot of vocal community members mouths.
But I think the only people it truly “hurts” those who have MIT frameworks/products and wish to use the component library.

But I think this is an incredibly savvy move on their part. Their big value adds are in the component suite arena, and not in the area that every other JS lib is already working (namely events, dom utilities, ajax, etc) and I can see why they want that portion to be free of cost, if not free of the GPL.

They’ve also changed their Class based mechanism to be truly different from YUI’s (whose classical mechanism is pretty darn painful).

Overall, I think this is smart of them, and I think if they can ditch the bad smell that has been left by the GPL change snafu, it will help them.
My suggestion would be for them to take a few of their components and MIT those as well as a sort of community give back.
I would namely do the DataGrid, Menus, and maybe 1 other one that is already commonly done in other libraries.
They have some real unique stuff (for instance, the tabs, the sortable tree, the forms, etc) that they can keep GPL on, but I think they should consider reaching out to a community that feels a little burned and maybe giving back some components as a bit of an olive branch.

Just my opinion though. There will always be people who are never quite satisfied.

They just sold out their first conference. I’m sure they are still doing alright.

http://extjs.com/conference/
http://www.extjs.com/forum/showthread.php?p=311029#post311029

http://ajaxian.com/archives/thinking-about-the-difference-between-frameworks#comment-267880

My guess is that they would have had this out the door sooner, but they’ve been extremely busy with the upcoming 3.0 release.