I spent the last 3 years developing complex data visualization apps. We had strict constraints as to what kind of browsers we could expect(think government agencies) hence our use of Flash/Flex to build the rich apps our customers required. Flex helped us build some really cool stuff quickly, and had a rich ecosystem of tools, resources, and examples to draw upon. I'm really disappointed that Adobe didn't even provide some sort of gradual migration strategy. They basically took the only parachute in the airplane and jumped, yelling, "We are going to make it open source!" on the way out.
I have a difficult conversation waiting for me on Monday with my boss when I explain that our significant code base is now part of a legacy platform that the vendor has jettisoned. The software industry has an interesting gambling component: you bet on which language/technology/platform is going to be popular and hope for the best. Looks like I rolled snake-eyes.
Sorry mate, ever heard of vendor lock-in ? This is what happens adopting closed source technologies, they sounds great to start, then ... here you are.
Migrating to community-supported-free-open-source technologies "suddenly" makes sense.
Wasn't flex also open source at some point? I remember there was some hype about it a few years ago. At least the compiler and SDK were opened.
Back then I played around with it a bit and eventually concluded the new AS was basically a re-do of the Java language, library and ecosystem (even the bytecode was similar). I know there is a curse on Java in the browser, and making it look like Java helps enterprise adoption, but mirroring Java to that degree looked like a bit of a wasted effort.
"""Sorry mate, ever heard of vendor lock-in ? This is what happens adopting closed source technologies, they sounds great to start, then ... here you are."""
So, you didn't really understood that part where he wrote that for their specific requirements Flex was better suited, and enabled them to deliver stuff that they couldn't otherwise (at least without much pain)?
Also, yeah, it's like this kind of thing never happens to open source projects. Like, for example, the community loses interest and moves on, or decides to change the direction of the project to something you can't work with and don't care about, and the old platform you were building on is suddenly left to digital rot (no new features, not adapted to new environment and os releases, no more fixes, etc).
Sure, you still have the source for the old project.
But it's damn near useless unless you have the will, the knowledge, the time, and the manpower to fork it and adapt it to your needs, or the money to pay others to do it.
You have to understand that living at the whims of a volunteer community has it's own drawbacks, compared to living at the whims of a commercial company.
The problem with the Flex community as it is today is that the sponsor of the community has no direct responsibility to the community except as it benefits the company. Obviously this will be fixed as soon as it is handed to a sponsor that, in effect, _is_ the community.
"I'm really disappointed that Adobe didn't even provide some sort of gradual migration strategy". This might not be much of a consolation, but I'm sure they'll be offering a solution in the form of an alternative product that will be familiar to Flex developers.
I built a number of visualization and charting tools for a client of mine using Flex. They worked well, but printing was always problematic and once iPads started emerging, the "this won't work there" problem reared its head. The fix for me was to go back and re-implement what I'd built using Flot ... an HTML5-based visualization library. Migration was simple ... maintenance is much easier, and performance-wise, it loads faster than Flex ever did. For anyone in the same boat, I highly recommend checking Flot out.
In the environment where I used Flex printing was always a problem as well. There was an (understandable) expectation that they should be able to print out a meaningful representation of what they saw on the screen. A _huge_ source of pain.
Flot looks interesting, but the examples aren't qutie there yet. While I know it's all possible using the canvas, where are the data change animations, decent resizing, animated zooming etc.
I believe the title is accurate. As with their last post regarding the death of mobile Flash, Adobe seems to have an unfortunate penchant for ambiguous language. Rather than just coming out and saying that they are putting Flex out to pasture and wish it well, they spin the news by reaffirming their commitment to Flex while admitting that the Flex team is moving to HTML-based projects. It's a confusing, seemingly contradictory message.
That said, I think it's admirable that Adobe is handing over stewardship of Flex to the Apache Foundation instead of just killing it outright. They should be commended for that, since many other companies seem unwilling to so. But given the long-term trajectory of technology trends, it does indeed seem likely that Flex's days are numbered.
Sure, but I think that definitely qualifies as a win/win for those who want to see flex have a future (I'm fairly agnostic on such things, fwiw).
Further, I do think it says something that a large corporation, such as Adobe, is willing to open source products it's leaving behind. Shrewd move for them or not, ultimately, if most/all companies did this with the IP that they had EOL'd, I think it'd make a real difference, fostering grass-roots innovation in interesting ways.
Of course, they won't say otherwise right away. But they're contributing the Flex SDK to an open source foundation. Knowing that the Flex SDK is already open source, why would they do that if not to let the community decide of the future fate of the technology? And if the community doesn't do anything with it, then so be it?
If you let me pick between Apache and Adobe for stewardship of a project, given track records, I would totally prefer Apache to keep a project alive, maintained, community-driven, and open.
The problem is not the title, it´s Adobe´s deliberately unclear response.
Saying they are still committed to Flex and at the same time state that they will be moving their efforts to HTML5 sounds a mixture of cowardliness and hypocrisy.
The truth is that they will support the investment that´s been made in the short-term, but no, they are not committed to the technology anymore.
It seems that Adobe is shifting most of their resources away from Flex and towards other projects. The same thing happens to obsolete operating systems like Windows XP. They are technically supported, but the company has moved on.
I won't comment on the semantics of whether the title is editorializing.
No. More like if you wrote a toolkit for displaying ui items in canvas (I think Mozilla did this with Ace/Bespin/Skywriter/whatever-its-called), then it would be equivalent to flex.
Adding to the other comments, I find it easiest to think of Flex as a Flash-version of HTML, CSS, JavaScript, and pick-your-favorite-JS-libraries such as jQuery and Sencha. Instead of HTML, Flex has MXML. Instead of JS, it has ActionScript. (Though both come from the same language standard.) CSS works, and they bundle a good set of default themes and widgets. I think it's a really nice piece of work, especially when they put out Flex 4, it's a joy to work in compared to raw HTML+JS. The only downsides are Adobe, the flash runtime, and the intellectual heritage of Java sometimes getting in the way.
Flex is basically, a command-line toolkit to build SWF, while Flash is a GUI-based tool. Flex has the builder based on Eclipse and borrowed lots of conventions from the standard web technologies.
As the browser has been heading towards the capabilities of Flash, Macromedia/Adobe long ago added a layer on top of Flash for application development similar to HTML called Flex/MXML.
Pulling out of mobile, which is the present and future, hints a little at Adobe's focus now. Flash seems to have lost with mobile and now Flex being pulled, how long before desktop is EOL or sunsetted?
Flash 11 and Stage3D seemed a promising alternative to WebGL or competitive player, which will they choose there? Google had O3D and chose WebGL. But Flash would have supported IE and Unity and Unreal have committed to exporting to it since it is a low level and compiled render engine, it may be faster than WebGL with just javascript for some time or more steady across all browsers like Unity. Anyways, lots of questions on Adobe's direction... standards or not? and which ones?
Flash 11 and Stage3D aren't just a promising alternative to WebGL: they're the only viable alternative. WebGL only works in Firefox and Chrome (50% of the market) and, further, half of the Firefox and Chrome users don't have modern or secure enough GPUs or drivers to use WebGL.
Said plainly, WebGL only hits 1/4 of the desktop market. Stage3D/AGAL, on the other hand, are well designed and have an automatic software fallback. It will work on every machine, with one of the fastest software renderers known to me (I believe it's SwiftShader, please respond if you know more).
I really hope Flash 11 doesn't go away anytime soon because WebGL just isn't ready for ubiquity.
> half of the Firefox and Chrome users don't have modern or secure enough GPUs or drivers to use WebGL
Fixed with current Firefox, as far as I know. The first release of Firefox with WebGL had an awful whitelist/blacklist system for video drivers, primarily due to them crashing when asked to do relatively basic initialization and operations. Now, Firefox has a sensible probing mechanism for video drivers, and as a result, almost any modern video driver will work just fine. (Certainly anything new enough to run OpenGL 2.0, which means all the graphics chipsets sold today, and those sold in the last few years.) Intel, ATI, and nVidia chipsets all work just fine; those represent the vast majority of the market at this point.
(Also, if a user's graphics drivers don't provide enough stability or security to work with WebGL, Flash won't improve that situation either.)
As for browser support, right now it works in released versions of Firefox and Chrome, and preview versions of Opera and desktop Safari. iOS 5 supports WebGL for iAds, and web content will likely follow soon; people have already figured out how to make it work using private APIs. As far as I know, current Android releases support WebGL as well. So, as usual, that just leaves Internet Explorer. (And even IE users have several possible alternatives, quite apart from upgrading to a better browser: plugins like IEWebGL, Chrome Frame, and even some attempts to use Java or Flash as a fallback.)
So, while WebGL support does need a bit more time to become sufficiently widespread, it seems likely to do so in the near future.
You missed my point entirely. Even if Flash has the same blacklist as Firefox/Chrome, it has a high-performance software implementation, meaning you can hit the entire market with Flash 11. Until IE gets WebGL, everyone out there updates their drivers, and the Intel GMA950 is no longer the top video card, WebGL just isn't ubiquitous enough.
That said, SwiftShader also features an OpenGL ES 2.0 API, making it highly suitable for WebGL implementations as well. In fact Google is currently in the process of integrating SwiftShader support into Chrome (http://codereview.chromium.org/8480015/).
If Stage3D is stillborn, I won't know how to react. As an old OpenGL guy with no Flash experience, I'd personally much prefer to work in WebGL and JS (or even better, NaCl). But, it was looking like Stage3D was going to win. I was expecting WebGL to be the right "open standards" choice, but Stage3D would have better compatibility and security in practice.
Also, with Unreal running on Flash, it's become clear that Alchemy has quietly matured into something serious. I was expecting to see everyone argue endlessly about NaCl's "open standards" issues while watching Alchemy charge ahead to take over without asking for anyone's backing.
As Adobe mentioned, they are going to donate Flex SDK to community. Developing SDK is not a major problem. But development of underlying platform is. Can community trust Adobe that Flash Player runtime will be around for a longer time?
As someone that has worked around and with flash for years, I say good riddance. Though there are portions of the flash 'product' I think are much stronger than HTML5 and javascript (specifically, I feel that actionscript 3 is lightyears ahead of javascript), Flex was always one of the weaker parts. It always came off as half baked and did not blend well with everything else. The whole concept of MXML files was ridiculously obtuse and only led to unmaintainable code.
My only disagreement is with your disdain for MXML. AS makes it a lot nicer to build UIs than, say, Java with Swing et al., but I still prefer some sort of structured language like XML or even S-expressions to build a UI with over the procedural "make new widget, set the attributes, add it to this other widget" pattern.
I did some work with flex/mxml stuff about a year and a half ago, and frankly, it ranks as the worst software development environment I have had the displeasure to use. This includes the 8086 chip with 1 LED for debugging.
As another Flash veteran I completely agree with this. If you ever look at the Flex source code it's so obvious that it has been hilariously overengineered (which probably explains why Flex apps have always run so poorly).
I don't think Flex ever really had a suitable target audience - the premise was being able to create cross-platform/cross-browser widget applications, but that was hindered by the fact that the entire app ran on the AVM (with all of the limitations that entailed).
Anyone that wanted to deliver a great user experience (or had performance constraints to consider) would end up writing a native app anyway.
It's a mature open-source framework - the Flash Player is getting some significant upgrades (Unity + Unreal engine with Stage3D), allowing many things not possible now, tomorrow or next year with HTML5/Js. I know I'll be using it on our projects this time next year, as will numerous other codeshops like ours.
This is all before you look at what becomes possible with executables created with 3rd party tools like SWFStudio & Zinc (where we roll our own native code into the release).
Sorry, but even this time next year, there's going to be nothing around that approaches what we are doing with these tools now, in the timeframes and budgets we are doing them in (this part is key for our customers) - believe me, we've evaluated all of them long & hard.
Just because it might be the end of Flex for you and your projects, it's not going away overnight for many other developers.
With both Silverlight and Flex gone, the roadmap for future projects is clear: It will be HTML5. Unfortunately, what is not clear is which HTML5 runtime. In many ways, Safari enjoys the greatest reach - runs on Apples and Windows. Presumably Windows 8 will lock out Safari, which means we still have to support Safari and IE10. And we still haven't figured out where Android browsers fit in.
Perhaps Adobe's tool will be able to account for browser quirks better than humans can. We can all hope.
In Metro, IE is tied in pretty well, and currently there aren't any other browsers that work in Metro. On tablets (only Metro), the browser would have to be compiled to run on ARM processors (assumedly), so the barrier to entry is perhaps higher than now.
I think that it's likely there will be alternative browsers (assuming MS doesn't lock them out), but it may well be a similar situation to the iPad, where these browsers are really just a skinned UIWebView, rather than a different rendering engine.
IE was always tied to Windows, but that never prevented other browsers from running.
On tablets (only Metro), the browser would have to be compiled to run on ARM processors (assumedly), so the barrier to entry is perhaps higher than now.
Sure, but that's hardly locking browsers out, and besides Mobile Safari already runs on ARM (as does Firefox).
I think that it's likely there will be alternative browsers (assuming MS doesn't lock them out), but it may well be a similar situation to the iPad, where these browsers are really just a skinned UIWebView, rather than a different rendering engine.
But why? Just because Apple locks other browsers, MS will too? I don't get why you assume that.
I just wish people wouldn't criticise Adobe for making these difficult (but correct) decisions. They see the writing on the wall as clearly as everyone else. HTML5 and WebGL is what the rest of the computing industry is moving toward, and Flash is long in the tooth and the wrong development model for the future of the web. Flash served its purpose as a stop-gap techology until the web caught up. It's time to move on and we should thank Adobe for not prolonging this transition any further.
Heck, I'll go so far as to praise Adobe. While it's easy to "see the writing on the wall" from out here, history seems to say that it's often really hard for companies with dominant/historically strong positions in a market to see that revolution before it completely passes them by... I think Adobe has been more quick to change than many large/dominant companies in tech and other industries in years past.
Everyone is better off with Adobe on the "same team". They were trying to swim upstream and they knew it. The web dev world was marching forward with or without them and I'm glad they made the difficult but correct decision to embrace what is clearly the future.
Despite the widespread hatred for the Flash runtime it's almost universally agreed that Adobe makes great tools to create Flash things, and now we have them making those tools for the web. Which is also widely acknowledged as a giant hole in the modern web dev process. This really is a good thing for almost everyone as far as I can see.
I'll second that. Most of the time, major transitions happen through the death of the companies promoting the old models, and the rise of successful ones promoting the new models. Worse yet, quite often the old companies have enough power to strangle the new models for a while. Always nice to see a company actually listen to where people want to go and help them get there; they might just manage to survive the transition.
I agree, and it's unfortunate that the web hasn't actually caught up yet. Uploading files via the browser, working with sockets, and consistent video / audio playback are still seriously crippled in the current browser landscape. I hope that these features may finally improve across all browsers now that our universal crutch has begun throwing in the towel. I still think it's a bit too soon.
Note to mention the infamous JavaScript cryptography and it's problems, and it happens that Flash 11 introduced a secure random number generator for crypto.
The SDK will continue to be developed, the roadmap for the commercial product will be outlined.
What I'm missing here?
Pro-Apple troll. Go away. Who was the idiot that crafted the headline?
Does anyone else have the feeling that Adobe was just waiting for Steve Jobs to die before they killed Flash, just to avoid giving him the pleasure? He would have been delighted that he won. ;)
I have a difficult conversation waiting for me on Monday with my boss when I explain that our significant code base is now part of a legacy platform that the vendor has jettisoned. The software industry has an interesting gambling component: you bet on which language/technology/platform is going to be popular and hope for the best. Looks like I rolled snake-eyes.