AIR is now from HARMAN, what to expect?

[clouds40936465]

It is good or bad?

[zwetan_uk]

https://forums.adobe.com/people/Leo+Kanel  wrote

If it was Adobe and not Harman, you will be crucified by the well known fan boy right now whose name starts with Z =P

while you artificially try to raise FUD talking about things you don't know

I (the name starting with a Z) work on something that all of you are overlooking for many years

it's here and it's opensource GitHub - Corsaair/redtamarin: AS3 running on the command line / server side

so Leo, you asked me once what I was contributing to the community, among other things that's what I do

I write cross-platform C++ source code based out of the AVMplus source code, which is approximately 1/5th of the size of the AIR code base, and I do that so ActionScript dev in general can easily build command-line tools the same easy way they build AIR apps.

Also why I can deeply appreciate the amount of efforts Adobe developers been providing the last few years.

[Leo Kanel]

Ha of course you had to answer that lol. So please tell me what you think I dont know?

If you refer to redtamarin, I do know you are doing this although I find it of no use. So from your saying, you basically not contribute anything to mobile, yet you always have an opinion about it (especially android). Kudos.

Joking aside, to settle this once and forever, no one (including me) is poking you for your knowledge or what you do with as3. We all (especially me) about your clinical fanaticism with AIR.

[zwetan_uk]

https://forums.adobe.com/people/Leo+Kanel  wrote

Ha of course you had to answer that lol. So please tell me what you think I dont know?

everything that is discussed in this thread about HARMAN

people assume stuff, guesstimate other stuff, but simply they just don't know

the same way I trusted Adobe developers on their track record, I will wait to see a track record from HARMAN to start to comment

so far I'm seeing good news and announcements, I really don't see what purpose it serves to criticize things that did not even had time to start

let them work on AIR, let them release new SDK and runtimes, it gonna take few months to find a rhythm for them
they are taking over a project which is far from being trivial, it is unfair to judge them that soon

If you refer to redtamarin, I do know you are doing this although I find it of no use. So from your saying, you basically not contribute anything to mobile, yet you always have an opinion about it (especially android). Kudos.

well... I did provide a link to redtamarin, so off course I refer to it

so it's no use for you, ok
but that's maybe because you are too self-centered

the community is larger than just you or your specific projects

if you look around, all programming languages allow dev to build their own command-line tools,

for building, for testing, for doing many many things

also for mobile apps, many of them need a backend to talk to

you may not find use for it but I guarantee you plenty of AS3 dev do find use for that

Joking aside, to settle this once and forever, no one (including me) is poking you for your knowledge or what you do with as3. We all (especially me) about your clinical fanaticism with AIR.

it is not fanaticism
I have been called fanboy as if I don't know the tech I'm talking about
but if you actually do read my comments my angle is always related to the technology and the engineers working on it

whoever they are, Adobe or HARMAN, when those developers spent countless of hours working on a fairly complex code base,

that I happen to use and to like because it is tremendous help me to build my own software

sure I will defend them against less than unfair comments

[Leo Kanel]

zwetan_uk  wrote

it is not fanaticism
I have been called fanboy as if I don't know the tech I'm talking about
but if you actually do read my comments my angle is always related to the technology and the engineers working on it

It is fanaticism when you blindly try to defend them even when they do something stupid, wrong, especially harmful for the community (this is Adobe I am talking about).

whoever they are, Adobe or HARMAN, when those developers spent countless of hours working on a fairly complex code base,

that I happen to use and to like because it is tremendous help me to build my own software

sure I will defend them against less than unfair comments

Harman I accept it and I have also defended them as they didn't have enough time to prove themselves. Adobe..we all know what adobe thinks and how they treat their community.

Long story short, chill your tits. AIR is not a holy grail. There are many good technologies for Mobile, (React Native, Flutter, Native Script, Xamarin, AIR, Ionic, Titanium, Unity, GM, Corona just to name a few), there are others for desktop, and there are even more for backend (Firebase, Azure, AWS). I am sure that most of us have at least use 5 or 6.

[KevinF]

ajwfrost75  wrote

I'm interested by the "less mature than modern cross-platform frameworks" comment though, as I see it as a lot more mature than some technologies that keep changing so dramatically and that come and go quite a lot.

Try a .NET-powered framework like Xamarin or Unity. C# is a far more sophisticated language than AS3 will ever be, and .NET is a far more powerful framework than AIR will ever be. It's faster and easier to write better code that accomplishes more and has access to more native features. And since .NET is JIT-compiled to native code on Android, and pre-compiled to native code on iOS, the performance vastly outpaces AIR. Not to mention the far more sophisticated tools offered by Unity and Microsoft, or the extensive developer ecosystems.

React Native is another rapidly growing option widely supported by major tech companies, though I'm less familiar with it overall.

[ajwfrost75]

Yes I agree - although of course, ActionScript is also JIT-compiled to machine code on Android and AOT compiled on iOS. When I meant "mature" I was more thinking about things like Angular, where most of the tutorials and stackoverflow responses seem to be out of date because the framework itself is evolving, APIs are being updated etc etc. Adobe have done a very good job of keeping the APIs always compatible so that you can still compile and run old code on a new runtime.

[ASWC]

Let's keep everything in check here:

Xamarin, yes it runs faster then AIR and Unity and it's fun to work with but the downside is that it has only 2 targets, Android + Ios. It does just that and it does it the best way possible. Very difficult to create a cross platform code that can be used to output to desktop and/or web. I was able to create a base code that could run on all platforms but that base code did represent about 30% of the whole project. Maybe you can reach 40 or 50% but that's far from what we could do with AIR (when it had a web target) where without ANE you would get a near 100% compatible code base.

Unity, no it's not .NET powered, it uses C# (it used to allow Python and javascript as well), it's faster than AIR in 3D but in 2D results vary from one developer to the next. It's also very good at what it does and does deserve its success.

You could make a case that AIR is less attractive than Unity or flutter especially once it becomes subscription based since as of now AIR supports LESS platforms than those 2. As far as performance AIR might have fallen behind a bit but it can catch up. It's worth noting also that out of all cross platform tech AIR has the worse SoundAPI with latencies at above 60ms while industry standards accept 30ms max (and all cross platform tech except AIR stick to 30ms max or below).

Yes AS3 is lacking features in comparison to modern languages. Adobe gave up on ASNext and that doesn't help. Not sure what HARMAN can do about this but a more attractive language can make a big difference.

[KevinF]

> although of course, ActionScript is also JIT-compiled to machine code on Android and AOT compiled on iOS.

I don't know details of how the ActionScript VM handles AOT/JIT when compared to .NET, but in my experience there's a staggering difference in CPU performance between .NET and AIR on mobile devices. Coming from AIR, I'm frequently surprised at how much more I can get done on the CPU per frame in Unity.

> Unity, no it's not .NET powered, it uses C#

C# is a language, not a framework. Unity runs on Mono, the open-source .NET framework.

If you're basing your performance comparison between AIR/Unity on forum posts from 2016 or earlier, bear in mind that Unity is rapidly evolving and metrics from several years ago don't mean much today.

[ASWC]

If you really think for one second that I implied C# was a framework then I can't wait to see what else you are gonna post on these forums .... It's gonna be worth the read.

[hferreira80]

I have many years experience with .NET Framework and C# language and Flex/AIR.

I have 0 years of experience with Xamarin.

I started a new project with Xamarin this weekend (or better, a new App to totally replace a current one built with AIR) and I can say that, first, I see my self, searching on web for trivial things and a few hours latter I have a complete MVVM working with a blazing performance that I could never got with Flex/AIR on mobile device (I tested on emulator and also native Android device, I didn't have time iet to test on native iOS device).

Today I tested the publish to APK and again the performance it's outstanding (no more time to coffe, I can say).

I didn't have time to generate IPA so far, but I can say that blowing my mind, with a so simple process about the Visual Studio for Mac. I can go to options, link to my Apple account and it's all done !

With Flex, I have MXML and AS3 that I know very well.

With Xamarin, I have XAML and C# and seems that the learning curve is less dramatic than I thought will be, so I can say that I'm glad with my decision.

I agree in one thing, is fantastic for mobile devices, however, I can't reuse my code for Desktop and that's the reason I will stay around with AIR for now.

[ASWC]

Absolutely, Xamarin is the best option after native for mobile dev especially in term of perf. Now AIR plays (or is supposed to) in a different market so we shouldn't try to compare the two really. It's worth noticing that flutter claims to get similar perf as Xamarin but I haven't got any precise information about that. C# is definitely a beautiful and very modern language and while I like AS3 for its simplicity it really feels way behind in comparison with C#. I really don't like Dart on the other hand so that's why I still resist doing a deep check on flutter but it's definitely getting momentum and is becoming harder to ignore.

[hferreira80]

"different market"

I don't agree on that for Apps (not games).

AIR + Flex could be compared with Xamarin for Mobile Apps development.

I really don't like Dart either.

Dart was a wrong choice and we are talking about it only because of Flutter.

Seems more like an alpha product, comparing with Flex and Xamarin simply because there is no UI separation and see a lot of people complaing about the same.

After that evaluation, Xamarin win it BUT I believe with a lot of investment it's possible to Flutter get better however Dart it's unattractive and the future about Flutter it's uncertain for now.

[ASWC]

What I mean by different market is that Xamarin focuses only on mobile (that's 2 OS only) and that's the main reason they can offer such a good performance because they compile to native with a small trade of in perf. AIR cannot do that but does offer more platform support. So they both are a perfect fit for 2 very different needs.

[JoãoCésar]

there is no UI separation and see a lot of people complaing about the same.

however Dart it's unattractive and the future about Flutter it's uncertain for now.

I agree. I try to like Flutter but this Dart language and that confusing formatting are just not for me.

[KevinF]

ASWC  wrote

If you really think for one second that I implied C# was a framework then I can't wait to see what else you are gonna post on these forums .... It's gonna be worth the read.

I didn't mean any offense. My point is that Unity does run on .NET (in the form of Mono). You're correct that scripts are written in C#.

[_falcan0]

I think that diving into web and things like that is pretty premature if the main goal is making money off AIR. You need to get back the wider appeal and confidence of devs. The most important things to gain that IMO:

o Build up the ecosystem / tighten the community- this is super-overlooked right now. Look at Unity or even GameMaker. Give AIR a central hub / website like others mentioned. A dedicated forum. Asset store. This might take some time to build up, but at least you could provide a place to share / sell native extensions to start with. Community is super important to drive this tech, when someone new arrives it's nice to see there are community driven resources at your fingertips.

o Focus - Have a clear market, for example alternative to Unity 2D for building up cross platform games / apps. Focus on simplicity of use and speed of deployment, for people who are not exactly professional programmers, just someone who can develop okish in JS/HTML5 and/or struggling in Unity 2D / C#.

o IDE - Right now I think what is hampering the mass appeal (where lots of Unity devs are being recruited from) is just integrated environment which is a combination of visual and coding. People who are not hardcore devs(prefering to use C# or Java) are I believe your best bet. That used to be the core audience of Flash / Flash IDE anyway. Right now, you need Stage3D to deploy anything  sensible performance wise to mobile platforms, but you have no way to layout your UI / Sprites etc in the Flash / Animate IDE. There might be attempts by 3rd parties to supplement this, but thats not the same as integrated solution. You shouldn't know about Stage3D even(at first at least). It should "just work" like in the past: Open up the "AIR IDE", drag & drop some sprites / bitmaps, add code, and it runs at 60fps. This would be very powerful for people trying AIR for the first time.

o Commitment & updates - unfortunately, in the past we've seen this handover pattern from Adobe multiple times, which ultimately resulted in slow and painful death of various technologies. I experienced this firsthand with FlasCC, cross compiler from C++ to AS3 which was turned into OpenSource(a nice way of letting it bleed out, though they said of course this will be a super positive step). This is raising expectations that this pattern is repeating yet again with AIR. It would be good to counter this with pretty frequent updates in the upcoming X months before deciding what to do next.

The ecosystem buildup and just making sure it works accross Androids might be very powerful by itself rather than doing some crazy stuff with the AIR runtime itself like adding web targets(and then choking on it). That might surely come later but I think changing the general vibes of death into the vibes of growth is much more important right now, and there are pretty simple ways to achieve that which for some reason Adobe was avoiding like plague for all of these years.

P.S.(Having a roadmap would help as well, but I relaize you're still gathering feedback)

[Applauz]

I work in an innovation lab for a company...  we are always trying new things. 

I started building all functional mobile prototypes in AIR because I can build in 2 days what takes a team to build in 2 weeks.

Today I have 40+ AIR built apps in each of iOS and Google Play.  There is no way I would have been able to build those apps at the speeds I did had it not been for AIR.

This is what is powerful.  I’ve gone from wireframe to App Store in 1 week. (Both platforms).

I tried porting my apps with react native, etc.  All of them had poor performance. They were ok but not great.  AIR is the next best thing to building in Swift or Android Studio in my experience.

All I hope for is that this tech remains for me to build with. I am happy to pay for it.

I wish there would be more features built in that get covered by 3rd party ANE companies.

On a side note...  if you don’t have anything positive to say. Please leave.  You aren’t helping anything. You‘re a part of the problem now. We need positive people to assist with pushing this forward.  If you truly want AIR to survive use your time to let everyone know how great it is.

[hferreira80]

@ASWC

Indeed, in the past, but without Web (dead of Flash) and constant issues related with Mobile, I'm not so sure and yes, Xamarin is focus on Mobile and it's match better on that.

@JoãoCésar

Flutter makes remind me LISP and the way they release seems that they did in a hurried way (no xml UI).

[ASWC]

I agree without the web target I'm not sure how AIR can be really attractive to new comers. Some here say the web target doesn't matter but I do disagree. It's harder to find AIR attractive as a cross platform without it imo.

[hferreira80]

Microsoft is doing that.

They call Blazor.

I didn't have investigated so much (it's a not final product iet), however seems that you will be able to use C# on the client side and compile to WebAssembly.

How cool is that ?

Flash was a plugin for the browser, opening a runtime (desktop power) on the browser.

WebAssembly seems that same that already embeded on the browser, so I believe that this may be the future of web programming.

Compile to Web Assembly would be a winner even without marketing.

[ASWC]

I haven't heard of it and just checked it out and it does indeed look pretty cool. C# directly in a HTML page? What? Gee! I wish I had a project right now that could use it just so I can find an excuse to try it.

[hferreira80]

The project is called Blazer.

It allow to compile C# code to WebAssembly and also interact natively with HTML/JS.

They were doing internally as experience/evaluation of the project during an year and recently became official with Preview 1, so it's not more an idea and it's the next target for .NET.

The buzz word is WebAssembly but for me it's Flash 2.0 accept by community and pre-built with the browser.

Seems the future.

MS did a good job with Blazer and I would not surprised if we see in the future something like Java/WebAssembly and other languages.

So, AS3 (or AS4) compiled to WebAssembly it's the only way that I can see hope to became relevant again but time to market define who we will remember tomorrow and with the focus on specific Mobile and so specific on Android, it's hard to believe on this.

AIR updated more for Android, removes the strengths of AIR.

[Phasor1]

With all my respect, I do not understand the people who want linux with some sort of "priority" destination in AIR ...

In the business world, very few people work with Linux. The very few who work with that OS, will most likely have a PC at hand.

On the other hand, "hardcore" Linux users (programmers and system people), in general, care very little about the products that can be developed with AIR (in many cases interactive multimedia) ...

You have to be realistic and propose realistic things (and what really are priorities). Agree that everyone has their priorities, but within a general vision, there are a series of common priorities for all.

The first is to keep AIR running and updated for windows, mac, Android and iOS. This obviously includes Android 64bis (which is what they are working on now).

Adding features to desktop is not a priority right now (it is very powerfull as it is). It will be necessary in the future, and  Harman has announced that they will also take care of that part in the future.

The next priority is to create a website with information, tutorials, examples (and a forum). Make the community involved and participatory. And advertise it. This is totally fundamental, since easily accessible tutorials are what will bring people closer to AIR. Above ANY NEW feature. Because if you cannot find information easily and clearly online, nobody will know what can be done with it ...

Next is to move forward with everything (possible) that is a real lack ("fundamental") within the technology. Little by little (there will be nothing "revolutionary" in a year).

Another feature that some see as fundamental, is that AIR can compile to web destination ... Of course it would be very good, but this is not realistic at this moment in time. It is impossible that Harman (or anyone else) can implement this in a few months ... when they will also have other realistic priorities.

Personally for us, a web output does not seem so fundamental, for several things:

The type of products for which AIR is more optimal (interactive multimedia), no longer fits as much in the web as in the era of the Flash Player. Today, what I see most on the web, are "responsive" html5 pages with vertical scroll. For this there is absolutely nothing better than developing it in pure html5 + css + javascript. They are informative content with some small animations, but they are not interactive multimedia products. It has nothing to do with the type of products for which AIR has more potential ...

Yes, it is possible to talk about developing "webapps" (which I suppose is what some refer to). But to develop pure webapps of a certain level, from my point of view, it is better to do it within a pure html5 / javascript framework or a specific UI app framework that compiles for web (DOM, not canvas).

AIR is not as optimal for such "browser output projects", and probably never will be. Because, I insist, actually Flash / AIR started as a multimedia platform, and not as a specific platform for (web) UI apps.

For me the global problem is more for developers who work on a "complex product". In those cases, they have to make a determination whether or not they continue with AIR.

In our case, luckily, we work more in services with a series of small "products", which are easily portable to other technologies at any time. Of course, what we work with AIR (once again) is pure interactive multimedia, and for that there is absolutely nothing that comes close in any other technology. Anything (in that area) that we port to another technology, has fewer features (or is less "optimized").

An example: html5 canvas, as "output" is crap. We only see problems, slowness, defects ... And this cannot be "fixed" by any framework or environment, because we all depend on that "runtime" (which will not improve in the short term).

In short, what I have said other times. AIR is very good for a number of project types, but it is not optimal or perfect for "anything". If we want to develop for all this other type of destinations we must learn other environments and ecosystems. There's no more...

For our part, we will continue working on AIR for presentations, events, and interactive multimedia in general, because our situation allows it and because it is the best technology for all this. Consider other technologies as an option for these things, nowadays is a waste  of time. We also work with them, but for "other types of products".

Finally, comment that what is happening now with AIR, will happen to you with any other type of development environment in the future. You will have to recycle, because the technology will be "abandoned" (or will have less relevance) at some point.

[zwetan_uk]

Phasor1  wrote

With all my respect, I do not understand the people who want linux with some sort of "priority" destination in AIR ...

In the business world, very few people work with Linux. The very few who work with that OS, will most likely have a PC at hand.

...

If you target user software, Linux, sure is not a priority considering the market share
but when it come to the business world it is a different story

when you do projects like business touch screen, interactive display, orientation, navigation, etc.
there Linux make a huge difference

it is simply cheaper, more reliable and easier to update an operating system image based on Linux
especially when that OS run a software supposed to run 24/7 in public space like mall, airport, hospital, building etc.

you can do it with Windows too, but the imaging part is a nightmare

so AIR on Linux for the desktop, not only allow to install an AIR app on a Linux OS for the imaging part
but also provide the luxury to provide small'ish update where you can only update the SWF

think scenario where you have to update all your app software over a closed business network

I know many business case where people would rather pay a AIR Linux licensing fee vs Windows Licensing cost

now in term of upgrading AIR for Linux, it is a small effort compared to produce a web AIR target
it is not crazy, and it seems HARMAN already developed this kind of Linux builds for their own

[ASWC]

I think for HARMAN at this point it's more about resource management and short term success. That's what their next moves are going to try to answer.

They said they want to make money and bring new user to the tech so they need to figure out what's the right type of "new features" that's gonna attract new developers and that their own resources can handle. It's probably more about overall strategy than about adding this or that to the tech.