Slashdot Mirror
What Killed Adobe Flash? (daringfireball.net)
An employee, who claims to have worked on the development of Flash, writes: Apparently, the world settled on the "One True Cause" for why Flash "died". Take for example this blogpost by John Gruber about FedEx... it ends with this consideration on Steve Jobs' "Thoughts on Flash": "If it had been an angry rant, it would have been easily dismissed without needing to be factually refuted -- "That's just Jobs being a prick again." The fact that it wasn't angry, and because it was all true, made it impossible to refute."
Impossible to refute. There's no doubt that this was the beginning of the end for Flash, right? Except that this is utterly wrong. I worked on Flash, and I worked on the thing that actually killed Flash. It is my strong belief, based on what I observed, that Steve Jobs' letter had little impact in the final decision -- it was really Adobe who decided to "kill" Flash. Yes, Flash was a bad rap for Adobe, and Steve's letter didn't help. But ultimately, what was probably decisive was the fact that developing Flash cost Adobe a ton of money. John Gruber, responding to the blogpost: To be clear, I don't think Jobs's letter killed Flash. But I don't think Adobe did either. Eventually Adobe accepted Flash's demise. What killed Flash was Apple's decision not to support it on iOS, combined with iOS's immense popularity and the lucrative demographics of iOS users. If Jobs had never published "Thoughts on Flash", Flash would still be dead. The letter explained the decision, but the decision that mattered was never to support it on iOS in the first place. It's possible that Flash would have died even if Apple had decided to allow it on iOS. Android tried that, and the results were abysmal. Web page scrolling stuttered, and video playback through Flash Player halved battery life compared to non-Flash playback.
Impossible to refute. There's no doubt that this was the beginning of the end for Flash, right? Except that this is utterly wrong. I worked on Flash, and I worked on the thing that actually killed Flash. It is my strong belief, based on what I observed, that Steve Jobs' letter had little impact in the final decision -- it was really Adobe who decided to "kill" Flash. Yes, Flash was a bad rap for Adobe, and Steve's letter didn't help. But ultimately, what was probably decisive was the fact that developing Flash cost Adobe a ton of money. John Gruber, responding to the blogpost: To be clear, I don't think Jobs's letter killed Flash. But I don't think Adobe did either. Eventually Adobe accepted Flash's demise. What killed Flash was Apple's decision not to support it on iOS, combined with iOS's immense popularity and the lucrative demographics of iOS users. If Jobs had never published "Thoughts on Flash", Flash would still be dead. The letter explained the decision, but the decision that mattered was never to support it on iOS in the first place. It's possible that Flash would have died even if Apple had decided to allow it on iOS. Android tried that, and the results were abysmal. Web page scrolling stuttered, and video playback through Flash Player halved battery life compared to non-Flash playback.
No, Flash was a resource hog. We told them for years to quit using WaitNextEvent() and update to using the Carbon event handling APIs, but they kept on dragging their feet.
-jcr
The only title of honor that a tyrant can grant is "Enemy of the State."
One thing that layout and UI designers miss about Flash is WYSIWYG. Web (non) standards display differently under different browser and OS brands, versions such that you either have to test under a gijjillion client variations, or live with rendering mistakes. I HATE THAT and it makes me scream bloody murder. I want WYSIWYG dammit! Even slashdot often gets it wrong, as menus overlap when they shouldn't, etc.
And WYSIWYG doesn't mean that you have to settle on one screen size, it just means that if you test under size X it renders the same way under a client set for size X. Essentially the server does any resizing so that one doesn't have to rely on an inconsistent client. The client just sends it's preferred screen size and the server renders it and sends "dumb" coordinate-based vectors back: no client-side auto flow or "float" shit. Floats can float up my ass; floaters are what you find in the john.
It's probably the dumbest invention I've seen in my many decades of IT. Great job security perhaps, but sucky productivity as we fight with the plague of fat-client versionitus. Makes DLL-Hell look good in comparison. Now we got Client-Hell.
Damned humans! Its like a mom or wife that randomly rearranges your room while at work or in the basement. Whoever invented auto-flow deserves to have wake up one day to find that one of their girlfriend's tits are on her crotch and another on her back, but her snatch is now where her nose used to be. Hell, the client-floaters probably WANT it that way, sicko Picasso pervs!
I hope something like Flash with WYSIWYG comes back, as an open standard. The schizophrenic client problem is the main reason PDF's still live. Managers, customers, and/or designers decide where the put stuff and it STAYS there; imagine that. It stays where they actually want it and you satisfy their request as they sketched it. No drifting, screwy overlaps, or surprises when browser version N + 1 comes out. It's like magic! Imagine a Beowulf cluster of shit that stays where you actually PUT it. Imagine all the people living in WYSIWYG harmony, like God, I mean the Matrix admin, wanted it. Shifty shifters go to the basement to be Picasso BBQ.
Table-ized A.I.
iOS didn't kill Flash. Nor did Steve Jobs. Nor did Adobe. MACROMEDIA KILLED FLASH!... because they allowed it to become what it was never intended to be.
Way back in the mid 1990s, Macromedia acquired FutureSplash -- an ANIMATION product used by Disney, FOX (for the Simpsons), and others -- and renamed it Flash. I used Flash 2 for ANIMATION and it was a great tool.
Along comes Flash 3 and the introduction of MovieClips and transparency. Transparency was pretty straightforward, but MovieClips were not. MovieClips contained an animation (and timeline) that could then be placed in the main animation timeline. So, if you had an animation of a character dancing in a MovieClip, you could add that MovieClip to the main animation timeline and make the dancing character move up, down, sideways, whatever.
The introduction of MovieClips also brought some basic programming beyond the even more basic timeline actions that previously existed (solely for the purpose of starting, stopping, etc. an animation). You could now add your (stopped) dancing character MovieClip to the main timeline, and then add a button to the main timeline and add a "Tell Target" action to tell the MovieClip to start playing. This "Tell Target" programming was VERY basic, but it was sooooooooo confusing to most Flash animators because the FAR majority of them were truly animators, not programmers. In fact, MovieClip programming was so confusing to the animators' mindset that the "macromedia.flash" user group was constantly inundated with questions about "Tell Target." The concept of targeting "_level0" or the "_parent" or such made absolutely no sense to most animators. As a regular contributor to macromedia.flash, I eventually made a small website of "Tell Target" FAQs that was quite popular at the time.
What happened after that is what eventually killed Flash. Some people are great animators. Some people are great programmers. A very rare few are great at both. The ones that were great at both and using Flash started making some of the best Flash websites around. They were getting accolades left and right and being featured everywhere Flash was talked about. Gabocorp, 2Advanced, Der Bauer, etc. were thrust into the spotlight with their ability to combine great animation with great Flash programming to make jaw-dropping Flash websites.
With these kinds of websites garnering a lot of attention, the ever-increasing demand for more/better Flash programming started. Flash 4 add variables, input fields, the first real ActionScript, and other programming-based enhancements. Read through the list of versions after Flash 3 and most include more and more and more programming enhancements. Flash 5 introduced ActionScript 1.0 and Flash 7 had ActionScript 2.0 and on and on and on... until Flash died.
Security issues? Not a problem if Flash isn't a programming platform. Resource hog? Not (as much of) a problem if Flash ins't a programming platform. Unable to run on a mobile device? It's VECTOR GRAPHICS!!! Not a problem if Flash isn't a programming platform.
The interesting part is that in the wake of Flash's death, Animate survives... as an ANIMATION platform. Want to meet Flash developers who aren't looking for work right now? They're the ones who never stopped using it for Animation. Personally, I used Flash for a LOT of programming, but I also used it for a LOT of animation. With the shift of branding from Flash to Animate, I'm happy to see the return to the core purpose of Flash 1.0: ANIMATION!
If you look at the enhancements for Animate 2015 and 2017, you'll see a lot of items related to animation and graphics and not a lot related to programming. This is the way it should be... and probably the way it always should have been. Flash as a programming platform always should have been a separate product, like Flex, so it could live/die on its own merits, or lack thereof.