[dojo-contributors] what is the raison detre for dojo 2.0.
dylan at dojotoolkit.org
Fri Oct 5 11:30:28 EDT 2012
-----BEGIN PGP SIGNED MESSAGE-----
P.S. to show this off all, Dojo need to be built with 3 incredibly TBD
sexy demo apps, that are fully documented with tutorials that show how
to build those apps. This is how we address Peter's comment, which is
that people don't even know what these features are. e.g. see this nice
demo with live real time charts, how was it built... real-time aware
stores, gfx, charts, events, etc. People will care to learn what the
features are only when they see how nicely they come together.
on 10/5/12 8:28 AM (GMT-07:00) Dylan Schiemann said the following:
> 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
> 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.
> 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.
>> dojo-contributors mailing list
>> dojo-contributors at mail.dojotoolkit.org
> dojo-contributors mailing list
> dojo-contributors at mail.dojotoolkit.org
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (Darwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
-----END PGP SIGNATURE-----
More information about the dojo-contributors