Ajax Toolkit evaluation.

[Colin Clark]

Hi Josh and everyone,

Joshua Ryan wrote:
>> It may be more a matter that the reverse relationships (eg, JSF 'playing well' with the AJAX toolkit) in the end, but the AJAX toolkit should not be the 'bad guy'....
> 
> Jon, I think you nailed that one square on the head, I'd rather have a
> tooth pulled out then do any more ajax work in JSF.

Yep, I also agree.

Along with the AJAX toolkit evaluation, we're planning to look at what 
server-side presentation frameworks are most amenable to easily building 
user interfaces with Ajax, DHTML, and JavaScript in both Sakai and 
uPortal. Based on this ongoing assessment, we'll likely choose to bind 
with a small number of server-side technologies that will give us the 
best coverage.

When I was in Berkeley in March, Ray and Josh suggested that we should 
avoid trying to work with JSF if at all possible for FLUID. It sounds 
like your experience is similar, so I hear that advice loud and clear.

Given Antranig's involvement, and the very AJAX/JavaScript-friendly 
environment that RSF provides, we will be working with RSF right away to 
get FLUID started.

> I'm glad this evaluation is being done, there are many of us out there
> that eagerly await the results as doing a thorough evaluation on and
> then keeping up with more then two or three of these is more then many
> have time for in the midst of work and deadlines.

Good to hear that this will be useful to you. The uPortal community did 
some work about a year ago to compare AJAX toolkits:

http://www.ja-sig.org/wiki/display/UP3/JQuery+JS+Library
http://www.ja-sig.org/wiki/display/UP3/Mochikit+JS+Toolkit
http://www.ja-sig.org/wiki/display/UP3/Google+Web+Toolkit
http://www.ja-sig.org/wiki/display/UP3/Dojo+JS+Toolkit
http://www.ja-sig.org/wiki/display/UP3/Prototype+JS+Library

The challenge with these types of evaluations is the fact that the AJAX 
and JavaScript landscape is changing so quickly as people push the 
envelope of what is possible within the browser and eventually roll it 
into a framework or toolkit.

As I mentioned in context of the recent Dr. Dobb's comparison, my 
feeling is that it's also important to evaluate the road map, community 
supports, documentation, and long term vision of the communities who are 
producing these toolkits. There's no crystal ball, and we'll have to be 
careful selecting a technology, but assessing both the technology and 
the community will likely provide us with the broadest perspective.

Interested to hear others' thoughts,

Colin

-- 
Colin Clark
Technical Lead, FLUID Project
Adaptive Technology Resource Centre, University of Toronto
http://fluidproject.org