you need simple stuff, you need some effect ? hit moo or 10s of likes or more quicker, make your own with some set/clear interval routines. go for it.

finally, for some nice little tiny piece of dojo code, please check the js of 0.4x’s gravity function( dont remember the exact js filename) .

wtg dojo

With that said, I have given up on dojo and have gone to mootools. Why you ask… Well you guess it …. documentation and simplicity.

I would always run into a wall when using dojo and end up reverting my code back to plain HTML/form or end up coding the js myself.

My life is much easier now that I use something that has DOCUMENTATION and is simple to use.

DOCUMENTATION ROCKS!!!

just my 2cents

And to Laurent, the days of desktop apps are coming to an end. Adapt or lose.

use jQuery.

That’s not true. What is true about Dojo is that adopters will loose quite some time figuring out by themselves how it works.

It is a shame, I picked up Dojo at version 3.something, was impressed and used it on a few projects. But I left it again when I realised that instead of making better documentation available , the Dojo developers anounced plans for a version 0.9. During this period, the dojotoolkit.org website was degraded to some sort of glorified blog and that made things even less interesting. I went in search for other frameworks that simply got me going faster and KEEP me going faster
Mootools and Prototype are prime examples. Allthough neither of the two come even remotely close to the functionality that Dojo has to offer.

I’m afraid that Dojo has lost the battle, even before it entered the fight. A real shame and hard to imagine with sponsors like IBM and Sun.

You’re right.

Beeing a troll sucks too ;-)

@Matt

Yes, Dojo makes me bitter.

Or rather, the software development process it implies makes me angry at people who promote it.

I mean, Dojo is like Spring, Maven and all those monstrous Java frameworks that have the same purpose: to replace creative programmers by certified enablers with the promise (but just the promise) of lower risks and cost cuts. This policy is not just taking the fun out of the job, it’s also making my job a lot more difficult because I have to struggle with a ton of unstable and broken API.

Now, talking specifically about Dojo, why would you bloat a browser with yet another fat client?

To develop Client/Server applications like we did in the 80’s?

Well, if you really want that, use GWT: it translates C/S fat clients into lean web applications that leverage web protocols. But trying hard to work around those protocols as Dojo does, that’s just stupid.

There are no more constructive comment to make about Dojo than there are to make about COBOL, SNA, OS/2 or WS-*.

It sucks by design.

The only useful comment to make about Dojo is to warn potential adopters that they are about to make a big mistake and loose quite some time figuring out by themselves how much it sucks.

Along side Dojo, I was also taking a look at MooTools/Fx and I certainly feel the technology is moving along at a nice pace however documentation is king.

Dojo is lacking quite severely in this department, I can appreciated the developers have not had time to document the features – it’s understandable, when your being creative, you want to stay being creative!

They should take a leaf out of MooTools, the documentation for this is comprehensive to say the least (as a note, I am pretty sure the dev’s had contributors working on the documentation).

One of the first problems I hit with Dojo is the sparce documentation, the API doc is there however it is completely meaningless if you cannot put it into context with examples and background reading. If the developers are reading the comments then I would advise the developers to stop coding and start documenting the framework from the ground up. If you need to pull contributors in then do so – right now the perception of Dojo is wrong.

To many it appears like the developers personal pet project that only they, and a select few have knowledge of the inner workings of the framework (allowing them to fully utilise it’s power).

Personally I struggled so much with the framework (because I had to make guesses at how to approach a particular problem, due to not having the documentation in place) I got fed up and left it alone… not a good start.

You sound bitter – as if Dojo personally took away your first child or something.

Ask yourself this: Are my comments constructive? Am I citing specific examples?

I agree the email sample is horribly slow loading, but I don’t think that is at all indicative of the experience you get with a production deployment of Dojo. I am speaking from experience on that. And by all appearance, 0.9 appears to be a huge advance over 0.4. Hopefully now that 0.9 is out the door, the team can concentrate on documentation, new samples and demos.

I’m using 0.4 in production and even with many widgets the library is nowhere near “multi-megabyte”.

Perhaps, if you can collect yourself, could you describe what it is about the Dojo APIs that you find overly complicated, and explain how a different framework does something similar (or the same) with a better API? That might be useful feedback.

Cheers,

Fact: it took FF2 about 20 seconds to load and display Dojo’s demo e-mail UI, which is not exactly fast to run.

Fact: that demo page is ranked by YSlow as F(59) and yet it provides a demo with only a tiny fraction of Gmail UI features (which is rated A(99) and loading under 5 seconds).

Fact: compared to the other AJAX libraries, Dojo’s API are overly complicated, obviously coming from a j-spelled “entreprisey” land (beeing 40 years old I went through enough of that shit to know when I see some).

Fact: building a multi-megabyte framework in JavaScript is not required to emulate a GUI in a browser (GWT does it in much less code) and in a browser JavaScript is intended to be applied as directly as possible preferably leveraging the protocols implemented by the browsers supported.

Fact: Dojo sucks.

Not trying to sound like a fanboy here … But your comments were and still are almost entirely untrue. Short of a personal anecdote about “a few freshmen java developers”, you stated nothing but opinions. poorly formed [unsubstantiated] opinions in some cases. (fact: you didn’t even download it before you spouted the fud? ) “how could you?!?” think of the children. ;)

Also worth noting: you don’t actually have do download any of it to use Dojo because of the AOL CDN (including the DojoX snapshop as of 8-20). Just call in the minified dojo.js from the CDN (remember, only 26k from AOL) and give it a run, preferably _before_ spreading more FUD. I am led to believe you didn’t download and look. The “minified” comment gave you away.

@all, @mike: with regard to documentation: Yes, it is an issue. This 0.9release took tons of work to get together, 0.9beta just a few weeks ago, major house-cleaning on the part of the developers. The roadmap provided for dojo _promises_ 100% documentation / api stability for 1.0 … it’s “The goal” … Currently, there are tons of useful comments inline the code, and The API browser is being worked on in parallel, parsing/feeding off those comments. The volunteers that exist will be working feverishly to make the online book and API references all tied in together over the next few weeks. In their defense: the API wasn’t locked until yesterday just before the release, making documenting an ever changing source tree a living nightmare, even with the automation available. Hopefully, with the lock in place, this will all be ironed out, and some very shiny/useful demos and examples will be available.

Karl is asking: “What’s the problem with new version?”

All its 8+MB of minified and compressed sources suck.

Dojo suffer from the same defect as the Java framework that inspired it: it is bloated with useless complications developed on simplistic ideas or plain stupidly gross mistake. Along the lines of “Why make it small, simple and fast when you can make it big, complicated and slow?”.

Lack of demos is very much an outspoken thing on the website, and a very brief explanation for it (whether anyone cares or not) is that the demoEngine that was developed for the 0.3.x+ branch of Dojo was found to be extremely klunky and painful to maintain, so it was dropped in favor of a new demo system which from what I’ve gather is turning into a demo/theme tester. We do however still maintain a test case that covers all the features of a module just like before, however now they are kept in seperate folders (not 1 huge test directoy for the entire Dojo project).

As for documentation, 0.9’s big push was to include inline commenting so our custom API utility can extract the information and make a nice commentable API similiar to the PHP.net communities. People have been doing extremely well on this, and in the background we have a much larger “book team” that has been pushing pages and pages of documentation/examples into the online Dojo Book for the 0.9 release. It is leaps and bounds better than the original 0.4 one, if you overlook the mistakes that were recently introduced to it by major API changes days before the 0.9 release… (these are being fixed in the book as quickly as possible).

For the person up top that asked about what is Dojo, there are several very informative pages found in the first few pages of the Dojo 0.9 Book (found here: http://dojotoolkit.org/book/dojo-book-0-9/introduction )

-Karl Tiedt