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

Dylan Schiemann dylan at dojotoolkit.org
Fri Oct 5 11:28:31 EDT 2012

Hash: SHA1

I've waited a few days to answer while thinking about this, and have
enjoyed reading everyone's ideas.

To me, I think of it as follows:

Dojo 1.x is the path for apps that need to support legacy (ES < 5, HTML
<5, CSS <3, VML/Silverlight/Canvas, no WebSockets, no WebGL, etc.). I
could easily imagine this living on as 1.10, 1.11, etc., to support
users as long as this is important to them, much like Apache 1.x vs. 2.x.

Dojo 2.x to me is like Rick and others have said, assuming a modern
platform and building a toolkit that makes it an order of magnitude
easier than it is today to build robust web apps with JavaScript treated
as a first-class citizen. It is also a toolkit to support emerging
features like WebRTC, WebGL, etc., that provide excellent new
capabilities, but in many cases challenging and difficult to work with
APIs. Colin's brief note about how to build UIs with Dojo 2 is how I
would start thinking about what Dijit, widgets, gfx, mobile, DnD,
charts, dgrid, etc. need to migrate towards.

James notes about how a toolkit works make a lot of sense with how we
start building and structuring things. I still think we would want Dojo
to be able to release aggregated builds twice a year, which would be the
stable track.

In general, I want to see independently developed modules and packages
and systems that work well together. For example, I don't want to see us
having 10 systems for implementing themes or plugins or documentation or
tests, and while I don't want to lock us down, I think building on the
right common sets of tools and approaches to building things will give
us the opportunity to build and rev things independently, but aggregate
them together to make releases very easy for our users to manage and
work with.

After we're happy with the approach to tools and utilities and pulling
things together, we would start thinking about features and APIs for
those features. Let's take an example of gfx. What does gfx need to
become to be not only the best drawing API out there, but also what
would it need to do to support 3D, WebGL, and more importantly, what
does it need to do to become the best possible foundation for charts,
gauges, other visualizations, and even projects like Bonsai. How can it
work well with other modules and packages like animations to note
duplicate effort. How can it share a common theming system with normal
widgets (does that even make sense).

Throughout this process, we need to document what we're thinking, and
start marketing it. That way people know what's coming and start getting
excited about the vision, and know where help is needed and how to best
contribute towards our efforts.

There's so much more to say, but those are my top of the head thoughts
after a couple of days.

- -Dylan

on 10/3/12 4:54 PM (GMT-07:00) Rawld Gill said the following:
> As we being moving forward with 2.0, it seems this foundational question
> is important to answer: what is the raison d?etre for dojo 2.0?
> Everybody should try to answer. This includes current contribs as well
> as  past contribs that have moved on.
> Answers can?t be feature lists. Pretend you are justifying this project
> to somebody who is going to pay for it. After all, that?s exactly what
> is happening: we are all donating the most valuable thing we have?our
> time. As such, we all ought to ask ourselves, ?is this something I want
> to do??
> Answers should be a least a sentence, but not more than a paragraph. We
> can have long debates if necessary after we throw the gas on the fire.
> Warning: this is not a simple question. If we get it wrong, at best Dojo
> will be also-ran.
> Best,
> Rawld
> _______________________________________________
> dojo-contributors mailing list
> dojo-contributors at mail.dojotoolkit.org
> http://mail.dojotoolkit.org/mailman/listinfo/dojo-contributors
Version: GnuPG v1.4.9 (Darwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/


More information about the dojo-contributors mailing list