Termination of financial support starling and feathers

[денисг22484613]

Guys, apparently the AIR is no longer an Adobe priority. Tell the developers when you are planning to end AIR support ?

[ASWC]

There are differences that are measured and difficult to contest, as I said for example the performance of javascript/Canvas2D/WebGL in any context is just awful when compared to Flash/AIR. Performance of Flash/AIR is also bad compared to other tech like native Ios for example. Those are measured and objective facts. As David corrected himself, it takes HIM longer to develop something in any other tech other than AIR and I have no problem with that kind of subjective statement. When comparing techs you often can get measured facts and some way more subjective, development times are often very subjective. In truth, with a few exceptions, most tech dev times are about the same.

As for vector vs raster, it's just a fact (measured) that vector graphics, while they are easy to deal with at runtime, are CPU hoggers (even with a GPU mode) and not that easy to deal with when dealing with multiple tech/cross platforms since not many tech handle vectors. For that reason even when working with Flash/AIR years ago we eliminated vectors completely and not only we got ok performance on weak mobile device, we were also able to share those assets with other technologies without any conversion issues (xamarin, native ios/android, WebGL, etc ...). When every little bit of CPU counts, you really don't want vector, you want raster.

[Phasor1]

I was not comparing AIR vs native performance. That was not the talk. I compared the development time between AIR and canvas to arrive at a similar result.

It is logical that to get the maximum possible performance it is better to work with bitmaps / textures, but we return to the beginning of the conversation: this "need" depends on the type of project.

For us the key is the balance between result (smoothness) and speed of development.

We have our own framework based on AIR gpu mode in which we get very good results (between 50 and 60 fps even in slow devices). And we do not need to take so many steps to get there. But that workflow is not usable in javascript / canvas. You have to work harder to get a closer result.

For us, it does not make sense to waste more time on a workflow based on textures if we have a shorter path to get very good results.

Conclusion (the same): it is possible that in AIR the development time can be shorter than in html5 canvas, because there are other ways to do it.

Our results are also an objective fact. We have tried it enough times with different developments.

[ASWC]

Some people have the luxury to choose what they work with, some don't. We knew working with Flash/AIR will make us lose customers so we prepared for it and we made the switch. Converting our AS3 codebase to typescript was fairly easy, converting our apps as well and we didn't have a vector to raster messy conversion to deal with. So we were good to go pretty quick. Now I compile for web/ios/android/desktop, do I notice a speed difference in dev time? Nope, not one bit. Working with bunch of languages and techs gives you a good perspective on the pros and cons of each one and development speed is rarely a factor except maybe for javascript which is the only purely evil language ever made. Would i stick to Flash/AIR if I had the choice? Sure, nothing wrong with that.

[davidrenelt]

"As David corrected himself, it takes HIM longer to develop something"

I corrected you, not myself. At no point i claimed anything but that. You took it somewhere else, because .. who knows .. i really have no idea what your motivation in this is. It's mostly "look how smart I am" and "adobe is dead" .. both things can be true, but at this point everyone heard you. So, thanx, bye.

Your style of argumentation is at best annoying i must say. I think your contributions to the discussions are unwanted from both sides of the isle at this point. Why waste your time?

[ASWC]

My motivation is to correct people that are misinformed or simply make false statements. In YOUR case, your statement that somehow, magically, developing in Flash/AIR is FASTER than ... is simply wrong.

Then in your second reply you acknowledged the subjectiveness of your statement by saying YOU find it faster (just an opinion) which I then recognized to be a valid (subjective) statement.

Then in your next reply you simply start to attack me personally which I guess reflects some kind of frustration on your part. Sorry, keep your statements factual and you won't have to deal with me.

[davidrenelt]

"My motivation is to correct people that are misinformed or simply make false statements. In YOUR case, your statement that somehow, magically, developing in Flash/AIR is FASTER than ... is simply wrong."

Doubling down are we huh? I never said anything like that. You make your own "false facts" to correct them. You have too much time it seems.

[Phasor1]

ASWC  escribió

My motivation is to correct people that are misinformed or simply make false statements. In YOUR case, your statement that somehow, magically, developing in Flash/AIR is FASTER than ... is simply wrong.

...

I have shown you that it could be true. But you remain stubborn in that your way of developing things is just as fast on any platform. What is true, but only in your case, and in your way of developing and in your type of products.

As I said, in AIR there are other ways to develop it faster than yours for purely interactive products (not UI based applications - which is a totally different thing and there are alternatives -).

To achieve a result similar to what we do on html5 canvas (at an acceptable execution speed), much more time has to be spent. It does not matter which method or framework you use.

[davidrenelt]

It's is very tiresome to "argue" with you, which is something people point out time and time again to you, which you then interpret as attack. But since you are all about "correcting misinformed" people, you should be able to appreciate when someone points out an error. Firstly, read back what i wrote initially, my statement was always the same, I did not need to change anything, i was just pointing out that you misread what I wrote.

Because you decided to double down on the false narrative that I said something wrong and you are corrected me, in the spirit of things i felt like i should correct you, the misinformed in this matter again.

No attacks, just helping you along on your quest to perfecting your argumentation skills, so you can "correct the misinformed" better in the future.

This is all very petty .. and I'm a little sorry about that since everyone else need to at least skip over this drivel, but hey .. this thread has not too much information value anyways, might aswell have a little fun.

[Fede_Rivas]

I dont like this kind of news and of course, it doesn't sounds good.

Personally we're very happy developing with Air, we reuse 99% of the code between platforms, we've a stable and reliable platform and the perfomance using starling and feathers it's simply perfect for the kind of projects we develop ( business apps ).

We've coworkers using other platforms ( don't want to discuss details in each one ) but they simply can´t deliver the developments as quick as us, and they can't reuse as many code as we reuse, so maintenance costs once the app is in the market are much better with Air.

It's very sad that Adobe doesn´t seem to find a way to monetize this platform or revamp it. Many others try to reach the level where Air is right now and for me Air is still better in many ways. The only simple diference is that Air "is not cool" anymore, and being cool or not it's easily changed with marketing and new ideas.

Air doesn't need great changes, it's very mature and stable, have a good ANE collections for almost everything but needs quicker bug repairs, better marketing and the most important "A COMPANY THAT SUPPORT IT", because the worst sensation for us the developers is that silence from Adobe that happens many times. We simply dont know whats gonna happen.

I dont know the costs of the support that Adobe gives to Starling and Feathers but they're by far the best, most compromised and active community they have.

From this kind of things sometimes emerges good opportunities, perhaps open sourcing Air will be the fresh "AIR" we need.

[hferreira80]

Facts:

Adobe does not respond to these forums

A new release of AIR comes out every quarter

The Android 64-bit requirement was timely enough

The dead-line of the Android 64-bit requirement is close

Currently AIR does not support 64-bit on Android

Looking back (eg 64-bit iOS), we're likely to see the 64-bit version for Android almost dead-line

Why instead of wasting time discussing the sex of the angels, don't start using the time to study alternative frameworks to at least know they exist and what you can do with it and eventually even decide to move (even if 64-bit AIR for Android sees the light of this days).

[zwetan_uk]

Here what I know
even if Adobe AIR stop updates for mobile, I will still keep using it for the desktop

Now if people really want to waste time on alternatives oh let's look at Dart and Flutter

even with the hummingbird project announced

https://medium.com/flutter-io/hummingbird-building-flutter-for-the-web-e687c2a023a8

the official stand of Google is that Flutter does not support desktop, but still they provide that

Desktop implementations of the Flutter embedding API

now look into the examples

https://github.com/google/flutter-desktop-embedding/tree/master/example

https://github.com/google/flutter-desktop-embedding/tree/master/example/windows_fde

https://github.com/google/flutter-desktop-embedding/tree/master/example/macos_fde

https://github.com/google/flutter-desktop-embedding/tree/master/example/linux_fde

and tell me you would rather compile your desktop app from C++
instead of using Adobe AIR and ANE, I double dare you

for all the bad mouthing about Adobe AIR I'm saying it again
you don't know how good you have it with Adobe AIR

[ASWC]

Sure a quick easy desktop app I'll do it with AIR every time still today. However the power of native desktop developing (VS2019 and XCode) is sure slower but quite satisfying for big desktop projects.

[ZachBJ]

Thats why I always said over and over and over and over and over again to those kind of stubborn people:

If you love it live it

If you hate it leave it. (and sh*t up)