Kendo UI - a framework for modern HTML UI

If you take away how this looks, and start digging into the project from a beginner's perspective, this project is awful. I find this with most of the supposed "UI frameworks" out there for HTML. With a few exceptions, they mostly lack:<p>1. Good documentation that doesn't just <i>define</i> the framework, but teaches you how to use it and get stuff done with it. Code already defines what it is, your docs should tell me why it's this way and how to use it. In Kendo UI they've got a list of dependencies for javascript projects they need, then a few code snippets with no explanation as to why or how they work.<p>2. Good sample code, in a full complete project you can download, with documentation on getting it up and running. Your first sample code is how <i>everyone</i> will write code using your project. If you've got bad samples, poor formatting, and weird file layouts (or none), then that's what everyone will write and that's what you'll be known for.<p>3. Examples that gradually increase in complexity. Start off with a simple hello world, graduate to a chat app or something simple, and get them to a full blown large application. In this Kendo example they've got a demo picture viewer, with no explanation for how it was built, and viewing the source it looks like a huge mess.<p>4. Humor. These kinds of documentation are boring as hell, especially if you're just defining everything. It doesn't have to be insanely hilarious, but at least throw a few little funny tidbits in the code. Even the great tech books of our time have tiny little jokes for the people who pay attention.<p>5. Finally, these frameworks rarely have a "theme". MVC is a theme. Convention over configuration is a theme. There's only one way to do it. There's more than one way to do it. Themes work to help people keep the script for why everything works the way it does in their head.<p>It's too bad because this looks really good, and it could be the most awesome thing on the planet. But if I can't figure it out even if I want to, then I'm never going to try.<p>Finally, none of what I wrote above applies if your project is for fun and not meant to be a "product".

Note: <i>You cannnot use this library on your web site</i>. The licensing agreement forbids you from redistributing the library, whether or not it's minified. It further states, "You are not allowed to integrate the Software into end products or use it for any commercial or productive purpose." So private deployment is out too. It's strictly for your own evaluation and amusement.<p><a href="http://www.kendoui.com/download/licenseagreement.aspx?skuId=436" rel="nofollow">http://www.kendoui.com/download/licenseagreement.aspx?skuId=...</a><p>Have fun with that.<p>[Edit: Their web site has conflicting information in the FAQ. See below.]<p>[Edit: Updated link. Thanks pakitan.]

The question I am asking myself right now is how did this project make it to the top slot of HN. It seems to be a re-hash of some standard frameworks, some bad documentation, some buggy widgets and be backed by a largish vendor. I hate to say it, but it reeks of a voting ring.

Hate to jump on the negative bandwagon, but... I had a momentary hope for something truly novel, but found the usual aggregation of data-binding framework, templating language, and widget kit, where the widgets have various bugs/quirks that make them undesirable to use as is.

Using Chrome13 on Ubuntu, when I try to use the Drag&#38;Drop Demo, the draggable element jumps to the lower right corner of the mouse cursor.<p>In the slider demo, rapidly clicking multiple times on the left or right arrow to increase/decrease the value fires a doubleclick event, highlighting most of the text on the site.<p>In the window demo, the mouse cursor does not change when I hover over the title bar, although the window is draggable.<p>It's these little details that scare me off. When I use a framework, I want it to take care of everything. If I have to add css classes for the mouse cursor or fix element positioning, I'd just build what I need myself.

I'd seriously consider using this, just because of the stagnancy of jQuery UI. It's a massive project with hundreds of long-open tickets (despite thousands of dollars spent on incentivizing developers over the summer through <a href="http://rewardjs.com/" rel="nofollow">http://rewardjs.com/</a>). 1.8 was released in March of 2010, and the last milestone release for 1.9 was back in May.<p>To be fair, a lot of jQuery UI's development headaches come from supporting IE6, while Kendo only touts its support for IE7+...

Demo pages are completely broken for keyboard navigation (try tabbing or activating the accordion widget). Wake me up when the hard stuff is actually working (accessibility!)

I'm having a hard time nailing down what the killer features are here. For example, I saw that they advertise drag-and-drop with support for touch devices, but I couldn't even find the drag and drop demo on the site.

I've learned from experience to run away from these kinds of UI kits. I always end up hacking around their shortcomings (e.g., tokenized inputs with autocomplete, etc.)<p>This one does seem to have a nice, compact, intelligible stylesheet, though – big improvement over jQuery UI there.

I couldn't say if jQuery UI is better or this is better, honestly seems they are just different. Even if Kendo UI is faster I have never had a speed issue with jQuery UI (not speaking for everyone, just, I personally have never had a speed issue).

The problem I have with this and JQueryUI is that both are still too stylized such that they don't lend themselves well to being integrated into an existing design. And the JqueryUI themeroller doesn't help much. YUI probably does the best job of being generic enough to utilize broadly.

Let's hope the code produced by the controls is better than the code Telerik's CMS generates...

For some reason, in Chrome 13 on OS X, some of the UI animations cause the browser view to go black for a second and redraw. Not sure if this is just a Chrome bug but it's really obnoxious and means I won't use it until it's smoother.

We arent far away from being able to copy and paste beautiful front-end designs.

Aeroviewr button borders look ugly in Chrome and on Samsung Galaxy S also arrow on play button is of center.<p>Dragging on SGS shows circular dragged object as if dragged by left top corner of bounding box.

This is a bit off the topic of your actual framework, but as far as branding, I love your logo. Do you mind sharing the person/company that designed it for you?

It looks pretty good. Good UX and well designed, and I see that it's based on jQuery. That's useful for many reasons, e.g. you can least pull jQuery from a CDN.

Has anybody downloaded this yet? I see three css files and one minified js. What is the total size of all three of these?

I'm always put off when a framework doesn't get straight down to the code/usage examples. Down with meta bloat.

Nice but decent upload widgets these days include support for drop-zone uploading.

I apologize in advance if this opinion offends anyone, but every HTML UI kit that I've seen simply can't hold a candle to native kits such as WinForms, WPF or even Cocoa. I would really love a write-once browser based solution, but I just can't see that type of solution ever catching up to native tech.<p>How long do we have to wait before the browser can catch up? Do you think it will ever happen?<p>The closest thing that I've ever seen is Silverlight. With it I've been able to make some very excellent front-ends, with EASE, that look and behave identically between Mac and Windows. EDIT: The only challenge with Silverlight has been wheel-scrolling, which works fine on Windows but only works in Out-Of-Browser on the Mac.

Seems like an obvious FAQ would be, "How does this compare to jQuery UI?".

It actually looks pretty good and would save you a ton of time rather than trying to build some of these widgets yourself.<p>But are people willing to buy a framework like this? Or is everyone just using JQuery UI and leaving it at that?

Needs a lot more baking.<p>For web apps built now, I use ExtJS. The newest release has been a little too buggy, but they are working hard to make it better.