General interest in an AS3 -> JavaScript cross-compilation feature for Flex?

[Raju Bitter]

Since the first announcement of Falcon's cross compilation feature
Falcon JS, I've been following the blog posts and discussions around
cross-compilation of ActionScript to JavaScript. How much interest is
there in the Flex community to have JavaScript generation as a future
feature in Flex?

Since Adobe has been working on Falcon JS, it would make sense to
improve whatever code the company contributes to Apache, and continue
from there to build a solid JavaScript cross-compiler.

Bernd Paradies has posted some information in his blog about the
cross-compilation effort http://blogs.adobe.com/bparadie/, in case you
haven't seen that.

Does anyone know more about the status of Falcon JS, and how far Adobe
is planning to implement the cross-compilation feature? Based on the
videos from Flex Summit it's just an experimental feature, and there
are no plans to release Falcon JS as a finished product. So maybe
that's a good opportunity for the community to help Adobe improve the
product.

I'd be interested in supporting such an effort, although I won't have
more than 5-6 hours per week available for it.

- Raju

[Jonathan Campos]

On Tue, Jan 10, 2012 at 9:57 AM, Raju Bitter <[hidden email]>wrote:

> Does anyone know more about the status of Falcon JS, and how far Adobe
> is planning to implement the cross-compilation feature? Based on the
> videos from Flex Summit it's just an experimental feature, and there
> are no plans to release Falcon JS as a finished product. So maybe
> that's a good opportunity for the community to help Adobe improve the
> product.
>

Long story short. I wouldn't hold my breath for what Adobe releases as
Falcon JS. Not to say that the devs that have worked on it are doing a bad
job, but it is incomplete and won't be intended as a final product.

Based on discussions with many parties there is some real work needed to be
done on this to make it work. 1st, some pretty significant updates to Flex
to support the JS cross-compile. 2nd, the cross-compiler itself.

This is 100% my guess, but I would suspect that a rouge group would need to
work on a new Flex standard... let's call it Flex 5... that is remade with
this plan in mind. Then as the cross-compiler comes the Flex code will be
created in a way that it will be friendlier to cross-compiling with JS (and
possibly other end-points).

--
Jonathan Campos

[El Koro]

From: "Raju Bitter" <[hidden email]>
>How much interest is
> there in the Flex community to have JavaScript generation as a future
> feature in Flex?

I think there is a lot of interest, but also a lot of doubts about its
future success.
>From what I heard, the problem is not really about the language but about
the rendering performances.

One of the most talkative community person on this subject is Mike Labriola,
committers of Apache Flex and technical lead of Spoon, who talked during
various User Group / Flex shows both about his believes but also about the
experiments him and his team has already done on the subject.

He is really convincing, and you can for example hear it on this episode of
the Flex Show (start at 53:50):
http://www.theflexshow.com/blog/index.cfm/2011/12/23/Flex-Summit-Roundtable-The-Flex-Show-Episode-158

A more succint text version on the Spoon blog:
http://www.spoon.as/2011/flex-forward/

Dimitri k.

[Raju Bitter]

Thanks for your response, Jonathan!

2012/1/10 Jonathan Campos <[hidden email]>:
> Long story short. I wouldn't hold my breath for what Adobe releases as
> Falcon JS. Not to say that the devs that have worked on it are doing a bad
> job, but it is incomplete and won't be intended as a final product.
That's what I thought, it's going to be a lot of work to do full
cross-compilation for the Flex components with CSS support.

> Based on discussions with many parties there is some real work needed to be
> done on this to make it work. 1st, some pretty significant updates to Flex
> to support the JS cross-compile. 2nd, the cross-compiler itself.
Agreed.

> This is 100% my guess, but I would suspect that a rouge group would need to
> work on a new Flex standard... let's call it Flex 5... that is remade with
> this plan in mind. Then as the cross-compiler comes the Flex code will be
> created in a way that it will be friendlier to cross-compiling with JS (and
> possibly other end-points).
Sounds realistic. I've been a contributor to OpenLaszlo, and have a
good understanding of how the cross-compilation works. I believe pure
ActionScript 3 to JavaScript compilation is not the problem. The
current Flex components are much more heavy-weight (and powerful) than
what OpenLaszlo has. If cross-compilation of Flex apps is the goal,
the components would have to be modified, which is a larger effort -
which would result in a new Flex standard. If you have a large 3.x or
4.x Flex app, I don't think it's realistic to have cross-compilation
working for the existing apps. But for future apps based on a "Flex
5", it's a different story.

Projects like Jangaroo already have ActionScript 3 to JavaScript
compilation working (http://www.jangaroo.net/home/), but they use
ActionScript as the better JavaScript to create JavaScript apps
(please correct me if I'm wrong, I've never used Jangaroo).

- Raju

- Raju

[Raju Bitter]

Thanks for the info/links, Dimitri.

2012/1/10 Dimitri k. <[hidden email]>:
> From what I heard, the problem is not really about the language but about
> the rendering performances
I still believe that it's possible. Laszlo Systems has created a very
large Flash based RIA called Laszlo Webtop. They can cross-compile the
app to JavaScript, an the performance is about 70-80% of the speed you
have in Flash (Webtop is about 2 mb as an SWF, and around 2.2 mb as a
JavaScript file, if I rememer correctly). That was about 2 1/2 years
ago, and I haven't been following the recent improvements.

> One of the most talkative community person on this subject is Mike Labriola,
> committers of Apache Flex and technical lead of Spoon, who talked during
> various User Group / Flex shows both about his believes but also about the
> experiments him and his team has already done on the subject.
I wasn't aware of the work the Spoon group had done in that area.

- Raju