[dojo-contributors] what is the raison detre for dojo 2.0.

Rawld Gill rgill at altoviso.com
Thu Oct 11 14:22:35 EDT 2012


Thank you very much for finally saying this. I mostly disagree, see below. But this is certainly one of the key points of contention that must be decided.

> -----Original Message-----
> From: Colin Snover

> 
> We should not attempt to compete in the brochure site space.
> 
> We should not attempt to compete in the "helper library for a PHP/RoR/Java
> server-side-rendered site" space.

I'm not sure I totally understand the difference between the two. So my comments below are limited to just the "brochure site space".

> 
> Those areas are both well-served by jQuery and other more specialised
> libraries, and frankly, I doubt anybody here is even interested in working in
> those areas, so why they are the topic of discussion is beyond me. Because
> we are weak there? Of course we are. It isn't our focus, others do it better
> already, and making Web *app* maintenance easy precludes the sorts of API
> shortcuts that these other libraries take. I'm not saying we should not look to
> other places for inspiration, but it's absurd to say Dojo must be great at
> everything, just as absurd as the notion that jQuery or MooTools or YUI need
> to be great at everything.

I am not trying to be a jerk, but from an engineering pov, I don't understand why a simple, optimal, easy [e.g.] dom api that's is an end point is different than a dom api that is used to construct higher-level abstractions.

I'm *NOT* arguing we should copy JQuery or anybody else. I am arguing that the goal should be to have every one of our APIs be every bit as simple, optimal, easy as any competing API.

Can you give me an example (as in a specific function or group of functions) of where it is impossible to build a low-level API targeted as an end point in and of itself that does not imply a cost when it is used as a block in a higher level abstraction?

> 
> We should continue to be the best in the rich client-side application space,
> since 1. that has been our focus for a long time, 2. it is where we excel, 3. it is
> where the majority of contributors are invested in their day-to-day work, and
> 4. it is a space that is still in need of our experience and innovation.

Agree.

> 
> What does this mean for Dojo more generally? It means it will *never* be
> #1 in the JavaScript library space. So stop playing these stupid popularity
> contests already, or go work on jQuery and win the popularity contest but
> spend the rest of your life wedging square pegs into round holes as a Web
> app dev because it's not designed for that and never will be, just as Dojo is
> not designed for brochure sites and never will be.
> 
> We will ingratiate ourselves with decision-makers and developers in the Web
> app space by being the first well-known library to provide a fully integrated
> OSS toolkit for delivering Web apps across platforms using new Web
> technology. We have our own core technologies in place (modules and
> feature detection), so stop wasting time talking in circles about purpose and
> start figuring out how to build what *we* want to use. If these discussions
> keep going on, I'm going to start suspecting people don't actually want to do
> the work because it's a lot easier to complain than it is to write code and
> offer solutions.
> 

You are very sure of yourself. Yet you are missing the point that this discussion is not about JQuery. You are free to stop participating whenever you want. This kind of attack is exactly why you upset people.

--Rawld



More information about the dojo-contributors mailing list