flex-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Frank Martin <fm4...@gmail.com>
Subject Re: Porting Flex to the GPU
Date Mon, 11 Mar 2013 13:15:19 GMT
You should check out Feathers more thoroughly. MXML is being worked on
(check the git repository) and the skinning is actually far easier that in
Flex with it's half-baked CSS BS. Feathers lets you style ANY property, not
only those that are supported by CSS.

I can only advice you to give Feathers a try. I never looked back a single
day cause after more than a year of Flex mobile development I was
completely full of it. A half-baked, bloated framework with tons of bugs
and way too much baggage that is too slow for anything but the simplest
mobile applications. Just terrible.
Feathers boosted my velocity considerably, absolutely no comparison to Flex.

However, as with any Stage3D framework that emulates the native display
list (this is what Starling does) there are some drawbacks. These
applications just don't scale well, which means that if you've got lots of
display objects, the application will slow down significantly, which makes
the 60fps goal that some people have extremely difficult to achieve,
especially on tablets.
Not sure if this is something that is being worked on at Adobe (they don't
seem to put any more substantial effort into AIR anyway) or if these kind
of frameworks need to change something, but I'm sure there are guys here
with far more background knowledge that can elaborate on this.

If Flex is ever coming back to the mobile platform, I don't think that a
simple (I know that this isn't a simple task) port to the GPU won't do.
They'd basically have start from scratch with a more streamlined framework
and personally, I don't see that happening.

  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message