I don't know if NIH is applicable here, ActionScript 3 adopted the optional type syntax of what was supposed to be ES4 , not the other way around (I think).
If I remember correctly, early proposals of the ES4 standard had a type system. It was later removed, but ActionScript continued down that path anyway.
And now MS is trying to replace JS with Typescript ... the irony. So much time wasted because of petty fighting and stupid egos if anybody remembers what happened around 2006/2007 with ES4. So much stupid people that slowed down Javascript's evolution, and the web in general. MS bears a huge responsibility in this.
As a long time Flex developer who has a lot of affection for the technology I'm was not hugely impressed by how a dev would implement an app that would render to HTML5 in FlexJS.
As someone who actually liked AS3 (and to a lesser extent MXML), but loathes the Flash runtime, this is kinda exciting. Can't wait to try this when it's more than just a PoC.
FlexJS is actually beyond PoC at this point, but the docs are proving to be slow to catch up. The contributors have been discussing doc improvements on the mailing list recently, and I think once the content is better fleshed out, the perception of the project's maturity will be way different. You should check it out, ask questions on the mailing list, and point out what seems especially missing in the docs.
Have you tried Haxe[0]? Influenced by ActionScript, compiles to several platforms (including Flash itself, as well as HTML5 and other languages). Has a Flash-API like library as well[1].
I have! In fact, I wrote an early PoC implementation of a hAxe <-> PHP "compiler". When I learned someone else was doing it, I scrapped mine and helped with theirs.
I take it this does not bring any of the flash.* actionscript APIs nor any support for Flash SWF sprites, nor the flex 1/2/3/4 UI components, since they are talking about a <js:xyz/> set of libraries. So this is slightly interesting in its own, but leaves a lot of work on the table for anyone looking to porting existing flex apps.
If you're interested in the Flash runtime APIs, someone's working on an open source project that implements them for ActionScript compiled with FlexJS.
I wish they'd focus on the Flash/AS3 part and drop MXML. AS3 (aka DOA ECMAscript v4) was a pretty solid language offering the flexibility of JS with native classes and types that were optional... The declarative aspects of MXML were beneficial but half-baked IMO.
I disagree... looking at the rise of JSX and HTMLBars, MXML was, if anything, ahead of it's time. Being able to compose a page of components allowed for a clear understanding of the composition hierarchy, while allowing for structural components (like tabs) that have their children defined by the parent.
MXML is a key part of why Flex as so popular... AS3 helped, but Flex let teams of developers and designers work on projects... something that couldn't be managed w/ JS frameworks of-the-time.
If all you want is AS3's features, look into TypeScript.
Seeing XAML/WPF/Silverlight in the microsoft c# world after seeing MXML was like going one step backwards, in my opinion. Very similar but XAML is more awkward and verbose.
Apache FleJS is getting more and more interesting each month. Now with 0.7.0 release it got many things up like maven build, debugging with AS3, NodeJS integration, Visual Studio Code extension, and many others...
Unlike other JS frameworks, FlexJS is a complete technology composed of languages (AS3/MXML), compilers, frameworks, and is development cost is greater than others, but as things are coming into place the wait is worth it :).
I recommend you to join and play a little with technology! :)
OpenLazlo was a flex-alike that targeted flash and dhtml (yup, dhtml) but is no longer amongst the living. I shipped a project with it a long time ago.
This looks great, although roughly 5 years too late.
Adobe's late 2011 communications around the Flash/Flex platform and the ensuing alienation and loss of a large developer community must go down as one of the low-points in corporate planning and communication.
The rationale of this project is unclear to me. For existing Flex projects one should just continue to use regular Flex on Flash and try to slowly migrate to modern JavaScript.
Irregardless of the target platform new projects should never be started on Flex nowadays.
28 comments
[ 2.3 ms ] story [ 66.9 ms ] threadFor a more complete idea of ES4's type system, see "Evolutionary Programming and Gradual Typing in ES4": http://www.ecmascript.org/es4/spec/evolutionary-programming-...
Also, I see React with JSX repeating what MXML (AS3) / FXML (JavaFX) did 10 years back.
[0]: https://haxe.org/
[1]: http://www.openfl.org/
https://github.com/matrix3d/spriteflexjs
MXML is a key part of why Flex as so popular... AS3 helped, but Flex let teams of developers and designers work on projects... something that couldn't be managed w/ JS frameworks of-the-time.
If all you want is AS3's features, look into TypeScript.
Unlike other JS frameworks, FlexJS is a complete technology composed of languages (AS3/MXML), compilers, frameworks, and is development cost is greater than others, but as things are coming into place the wait is worth it :).
I recommend you to join and play a little with technology! :)
https://en.wikipedia.org/wiki/OpenLaszlo
Adobe's late 2011 communications around the Flash/Flex platform and the ensuing alienation and loss of a large developer community must go down as one of the low-points in corporate planning and communication.
Irregardless of the target platform new projects should never be started on Flex nowadays.