I am following this this discussion carefully as well as all what is
going on on Apache Flex from the very beginning.
Clearly if the future of Flex attached to Flash Player would be bright
this thread should not exist at all.
But is here, and there is a lot of concerns surrounding it. I have
recognised that long time ago and started my own project called BixBite.
If you are not familiar with it yet here is a little presentation I am
just getting ready for public.
http://bixbite.org/BixBitePresentation.swf (hope everyone can play swf
directly from the browser ;) )
There was a lot of talks here about approach to be taken to bring Flex
to the next level.
One of them was the fact that in order to target multiple platforms you
need to have a descent compiler to convert AST into native language.
That requires a lot of work and effort. Nobody here familiar with a
scale of effort that needs to be put into such project have confidence
it can be done in reasonable amount of time and resources due to lots of
platform inconsistencies and the fact, that some of the platform
specific languages and features just missing.
I took a different approach. Instead of relying on generic compiler, my
goal is to implement a framework with a consistent set of rules, in
native languages and platforms.
Feel a gap in missing feature territory and encapsulate it in one place.
On top of that very abstract scripting language can be developed as well.
But my approach is a bit different from what HaXe can offer. I am not
trying to create Esperanto language and force people to learn it.
I propose to hire native interpreters that sitting in one room, for
dispose of all people who can speak their native languages.
This room is my framework, so and Flex can also be that kind of room.
If I will see Flex going in that direction, leverage the best practices
of the target platforms instead to try to be a Swiss knife, I would be
happy to share my vision in details and contribute to the project. But
most of all, did you consider or discussed that very option well enough?
Best
Dan
On 11/21/2012 9:59 PM, Carlos Rovira wrote:
Hi Alex,
I'm strong advocate to make POCs in different directions so we could end
getting more knowledge that could end turning all efforts in a real next
generation framework. For that reason I think we'll end over the next year
with various groups targeting different points of views: AS3, Haxe and so
on, as well maintaining actual codebase
For me, in a full rewrite, the reason not to go AS3 is:
* AS3 will be killed by its own evolution AS4
* AVM2 will be killed by its own evolution AVMNext
* If we could get rid off Adobe's technologies, the better. It's not only
because we, as an open source project, should not depend heavily on
propietary technologies (if we can), it's because we have the experience on
how Adobe throw the towel, and that makes a precedent, so they could make
it again. Confidence in a future depending on Adobe is something that I
would try to avoid if possible.
* Haxe has the key points we are asking for: One language (OOP) - multiple
targets.
* Haxe will serve us next AS4/AVMNext without the need of change the
language.
* They already has the HTML5/JS output, the actual Flash AVM output, and
all the mobile platforms output.
...of course all this have sense since the proposal is "a new Flex from
scratch".
Your point of "Haxe is not in Apache" is not a point for me. Take into
account that we already use other open source projects that are not in
apache. More over, Apache is an instrument right now and even for such new
project, we could event think in go directly to Github, if after a few more
trys we don't get git-github support or Apache bureaucracy is not as agile
as we need.
I'm with you that we should start coding and making more POC and not only
talking, but In my case my next efforts as you saw will be in Git support,
and need to learn more about Haxe to be able to start playing with all this
ideas.
2012/11/21 Alex Harui <aha...@adobe.com>
On 11/21/12 12:53 PM, "Kevin Newman" <capta...@unfocus.com> wrote:
But if we are to change languages, why not go with a language that,
looks a lot like AS3 (and ports easy), addresses the language
scalability issues of JavaScript (lack of classes, typing, a compiler,
etc.), and can compile to JS as well as other languages? Haxe can be
compiled into JS, ABC/SWF, C++, C#, etc.
My angle for now is not to change languages. We can write in AS3 and
cross-compile to JS and maybe other languages. Apache Flex effectively
owns
AS3 because it owns a compiler for it.
Why NOT use Haxe?
-Haxe is not in Apache.
-There are lots of existing AS3 code libraries I think we should try to
leverage.
-I know how AS3 behaves on Flash.
But again, none of these, even in aggregate, are strong enough reasons to
a-priori say that some other group of folks shouldn't pursue a rewrite on
Haxe.
--
Alex Harui
Flex SDK Team
Adobe Systems, Inc.
http://blogs.adobe.com/aharui
