Ajax Toolkit evaluation.

Herb Wideman herb at yorku.ca
Fri May 4 13:03:40 UTC 2007

Hi Jonathan,

You have probably considered these additional criteria but just in case :

- efficiency of coding process using toolkits - how quickly can standard 
tasks/builds get done? Are there usability issues with the toolkits 
themselves that could impede their functionality for certain specialized 
purposes that are important to our planned development work?

- efficiency of resultant code - I'm speculating here as I have no 
programming/computer science background, but perhaps less efficient code 
could cause delays on low-bandwidth networks? There was quite a 
discrepancy in the Dr Dobbs-reported test Colin referenced recently - 
the Dojo JavaScript file generated in their test development was about 
8X larger that that generated by YUI. Could this be an issue for the 
development of high-access UI components which might possibly be more 
code- and bandwidth-intensive than standard UI structures?


Jonathan Hung wrote:

>Hi everyone.
>We have come up with a shortlist of  Ajax toolkits we will consider as 
>the technology to develop FLUID components on.
>1. Dojo
>2. Mochikit
>3. YUI
>4. JQuery
>5. Prototype with Scriptaculous
>Going forward, we will write a simple component to evaluate each 
>component. The criteria used in evaluation include:
>- debugging support
>- cross browser support (IE 6 and 7, FF1.5 and FF2+, Safari, Opera)
>- accessibility support (i.e. ARIA)
>- portal (JSR-168) compatibility (i.e. plays nice in a portal)
>- skinnable
>- community support and momentum
>- security
>- event abstraction
>- extensible
>We'd love to have input from the community regarding the 5 toolkits we 
>have selected for consideration.
>Also, if there are other criteria we should consider, please let us know.
>- Jonathan.

More information about the fluid-work mailing list