Nitpicking the infographic: Canvas should be listed as "partially supported" at best on those iOS devices, as it won't get hardware support until v5.
Interesting mix of data. Wonder why Android wasn't listed? The list of Android devices that even partially support flash in this time snapshot is a sliver at best as well.
It's not going to be any one thing that's going to kill flash. Instead, it's going to be a range of dedicated tools that each do one thing better than the Frankensteinian mess that is Flash. Flash video is already being replaced by HTML5 video, for example, and shiny hovering things at the top of the page are being replaced by CSS.
> Flash video is already being replaced by HTML5 video, for example
Care to provide some numbers to back that up? How many of Youtube's, Vimeo's and other big video websites are using an HTML5 player?
And please don't say that's because of H264 video penetration over FLV. Because H264 video (which is sometimes confused with "html5" video) has been the standard for quality Flash video for a few years already, and a 100% penetration of H264 over FLV would have no correlation to the penetration of "Flash video" versus "HTML5 video".
They may not default to it, but Youtube[1], Vimeo[2], and a few other places have HTML5 players available; I have them all turned on. HTML5 video is still in beta, but it's somewhat easier to build into a page than Flash and its development environment doesn't cost $500. The fact that it edged into Flash's market in the first place says something.
The fact that they're available, and you have voluntarily opted in, doesn't mean it's being replaced. Unless your count your household as a sole representative of the web in general.
My point is that the two largest video distributors both provide and maintain html5 players. The fact that they're putting in so much effort means that there is demand for HTML5 video and that people are actually using it. Furthermore, Apple mobile browsers don't have Flash, and their browser market share is steadily increasing. I don't have any hard figures, but even that says something; Adobe has every reason to release stats that support Flash, which suggests that it those stats don't exist.
Either way, HTML5 video adoption is certainly not descreasing, and there are several signs that it is increasing.
I agree with the author that Flash still makes sense for some projects. "HTML5" isn't an attempt to obviate Flash, it's a collection of specifications that pick away at some of the use cases of Flash like basic video and audio playback.
With that said, I think it's great we're now being more careful about when and where we use Flash. For performance and stability reasons alone, I'm happier it's on less and less sites.
This is my issue with the whole discussion. The pundit class has latched onto the "HTML5 will kill Flash" story so a lot of people now think it's gospel. But in truth there are a lot of features that HTML5 makes no attempt to duplicate.
The same is true in .NET. It's great that Microsoft folks want to say HTML5 is their primary development platform so they can sound cool. But unless they are seriously extending HTML5 there's no way it could replace everything available in Silverlight and WPF. Because that's not what it's designed for.
HTML5, as you said, is an attempt to deal with the most common usage cases in a standardized manner and that's great. But on the specialized functionality front Flash and Silverlight still easily have it beat.
flash is terrible for blind people. to be public sector compliant you essentially need to be able to tab through elements. you need html of some version.
With mobile requests on their way to outstripping desktop browser requests in the next two years on a lot of sites, it's the mobile site that's going to drive decisions about what you put on your web site and how you choose to let users interact with it. And Flash is essentially dead in the water from that perspective.
I'm perfectly willing to accept that we'll continue to use Flash to gussy-up the experience at the desktop and allow for more complex interactions with media. But frankly, most sites don't need (and users don't really want) that complex a set of interactions.
How many of us have gone to a Flash restaurant site and sat there cursing because it was hard to find what we really needed - the menu, hours, address or phone? Worse, tried to hit that from a mobile phone?
That's why flash is going to get increasingly marginalized, not because HTML5 is allowing similarly complex effects. Flash largely goes in the category of those old DVD ROM authoring softwares: harder to implement, harder to test, harder to use.
Most sites would get a lot more mileage out of a serious study of basic information design principles than tacking on the latest whiz bang effects, be they flash or HTML 5. Demos are fun but on the web I'm after information, not some kind of half-assed video game.
I think the Flash-based restaurant example has had it's time. Let's try and move past the fallacy that most Flash-based sites are restaurant sites. They're not. YouTube, Grooveshark, blog.tv, etc. are all possible because of Flash.
The dream is that we'll be able to harness webcam, microphone, great audio API, and so on, without some proprietary plugin.
The reality is, we won't be able to. As soon as standards are all caught up, Flash will have included 5+ years worth of advancements that will be absolutely necessary for some types of interaction with the user.
Flash may be proprietary, buggy, closed, etc, but life without YouTube and the like would really feel like the stone age more than an improvement.
Html5 youtube is starting to work quite well. Grooveshark is entirely html5 except for the audio, that is only because the html5 audio components are a little lacking at the moment (but they are making leaps and bounds, and within a year there should be good support in the firefox and chrome). Standards have pretty much caught up, we just need to wait for people to upgrade to browsers that support these standards.
> The reality is, we won't be able to. As soon as standards are all caught up, Flash will have included 5+ years worth of advancements that will be absolutely necessary for some types of interaction with the user.
The next big feature for the next flash player is full 3d support. WebGL, the html5 mirror of this already has support in firefox and chrome (although, it can't be used on XP or linux with intel due to bad driver support). In this case, the open standards are ahead of Flash.
> Standards have pretty much caught up, we just need to wait for people to upgrade to browsers that support these standards.
If we live today and today only, Moore's law wouldn't apply and we could all say for certain that HTML5 and it's successors will be the lead indefinitely. I don't see that happening. If Flash didn't exist today, given the rate that standards are drafted and implemented, would we even be here today talking about audio API or reliable video streaming? Who knows.
Flash is the necessary evil, you see. Standards are driven by demand. But how can demand exist for something that a user cannot experience? Flash. It is the petri dish of web innovation. Perhaps not the best petri dish but a petri dish nonetheless.
> In this case, the open standards are ahead of Flash.
They're only ahead if you consider one idea with poor implementation and compatibility in around 3% of the browsers to be "ahead". And by that same logic, Unity3D is already way ahead of both of them, and it doesn't mean anything either.
Flash already has OpenGL/DX enabled on Flash 11 (currently in an open beta), which should be released later this year. Given the quick turnaround we have for Flash Player penetration (6 months for 80% penetration; 12 months for 95% penetration) I can bet Flash's 3d will be at 90% penetration on all desktop platforms and browsers before even WebGL reaches 20% market penetration. There's no helpful headstart here. Like it or not, a Flash Player upgrade gets installed much faster than a browser upgrade. Look at the number of people who are still using old IE versions.
There are no "leaps and bounds" in the HTML world. Even if standards came out as quickly as a single company putting out their own platform, actual implementation and adoption takes a much, much longer time. This is the real world, and that some people ignore it is baffling. There are many reasons NOT to use Flash, but trying to sell HTML5 as a more capable standard is missing the point entirely.
And considering Apple's stance on not allowing WebGL on iOS either, plus Microsoft refusal to implement it, can't see any real advantage in using WebGL coming in the future. Quite the opposite.
With Apples position on flash (especially on iOS), do you think they are going to leave browser 3d stuff totally alone? It does make sense though, they want people to put that stuff in an app, and not a webapp.
Goes to prove that Apple's problem is not Flash, but the web. They don't want the web to have too many features and anyone who thinks they truly support HTML5 is mistaken.
"I think the Flash-based restaurant example has had it's time. Let's try and move past the fallacy that most Flash-based sites are restaurant sites. They're not. YouTube, Grooveshark, blog.tv, etc. are all possible because of Flash."
I'll stop using that example as soon as I stop having that problem once a week. It's even worse in fashion, where designers for tiny brands continually get suckered into making flash-only sites that are essentially useless beyond being brochureware.
Are there sites that use Flash to great benefit? I have no doubt, although it seems to me the examples you're using are just using it as a way to deliver media. But for 99% of web sites, Flash is a bad idea and provides little benefit. I use Flashblock on my browsers, and the only time I seem to need to unblock anything is to watch a video. That's a bad sign for folks looking for a long-term use of Flash outside of players and video games.
My point with Flash reaches a few thoughts into the future, while most are perfectly happy to have their ignorance contemporary. Let's just say that tomorrow Flash DIES. It's gone. No more Flash. Do you think those same shitty restaurant sites are suddenly going to become standards-loving, keyboard-jockeying HTML 5 + JS sites? Again. The issue isn't Flash. It's the industry. The same restaurants are going to call up the same vendors and ask them to provide them with, what they consider, the new hotness.
We place the blame on the customer, or the user, or in this case, the technology. Perhaps all along it's just us. And our uncanny ability to misinform the client and leave them without information on accessibility, mobility, scalability.
It seems unlikely those sites have a very representative audience. The numbers you cite are perhaps large slices of tiny slivers of the overall group of web users.
Indeed. In fact, Gruber has, on several occasions, encouraged readers to go Flash-free, which I'm sure has an impact on the number of his visitors that don't have Flash installed.
I suspect HTML5 won't displace flash, but standards in general eventually will. Not any time soon, mind you, but eventually.
I've been trying to come up with a way to articulate how I think about this whole non-debate and here's a first stab:
One of the primary reasons Flash will eventually be unseated is not that Flash is closed -- standards, to a certain extent, are constructed in a relatively closed manner -- but because standards and technologies built on top of standards are more agile.
Flash is an awesome platform, but you can only build on that platform. Standards are great because you can combine platforms and build things not explicitly envisioned. Flash's single-platform approach makes it so that it will eventually have to play catch-up to a whole set of features that are being built by mixing and matching standards.
CSS3 + Javascript, for instance, can be used without Canvas to build really awesome functionality.
Canvas + Javascript can be used to make really engaging content.
Flash will eventually have to compete not with another product on features, but an entire web of organic technologies that don't have to wait for other components to advance in order to release the next stable version.
I think that doesn't contribute to much. Flash, in a way, is also built on a set of smaller technologies and frameworks that vastly contribute to its popularity, and those are pretty independent. Look at, for example, tweening/easing libraries, which have been pretty popular in Flash for almost a decade and only started showing up in the JS side of things a while ago.
What will displace Flash is the lack of necessity for Flash, which is (or should be) obvious in 80% of what we see online. There are many things HTML do better, and easier, than Flash, and ignoring that is the real problem. People all of a sudden are trying to make HTML5 be like Flash, and failing at it, instead of remembering why Flash has never displaced HTML in the first place. I'm all of a sudden seeing a bunch of completely inane HTML5 websites that could have worked much better were they built with simple HTML technology, and it's infuriating to think we're back to the insanity of the Flash 5 time, where people would build stuff in Flash just because (something that today's Flash-based restaurant websites are an evolution of).
People need to remember that making snappy, simple, straight forward websites is totally a thing. That's what's gonna keep Flash at bay, in its little corner, as a technology best suited for lots of interactivity and media and embeds or whatever it is. Not trying to make all websites be Flash-like but just with an alternative but barely supported technology.
One minor point: The iPhone and all other iOS devices have an excellent screenreader and are in general very accesible out of the box. Judging by the stories you hear on the web at least some blind people use iOS devices and are quite happy with them.
Here is my view: If what you want to do can only be done with Flash then – of course – you have to use Flash. If you can't reach some people you want with a HTML5 solution (because their browsers don't support it) then you have to use Flash. If you can't reach some people you want to reach with Flash then you have to use HTML5. If you want to use HTML5 at all you probably have to use a Flash fallback since it's quite often the case that some people you want to reach can't use HTML5.
Tradeoffs are always involved and the landscape is changing too slowly to care about anything but current (and not predicted future) usage statistics – in most cases. It always depends on the necessary work and supported features. Ideally you would support both but that's either sometimes not possible or too much work for too little gain.
You might not currently want to use canvas but, I think, video has become a pretty obvious case. Today it's eminently possible to use HTML5 with a Flash fallback with little additional work. MediaElement.js (http://mediaelementjs.com/), for example, is a beautiful and elegant solution to that problem.
The nature of Flash is changing significantly. During the previous decade Flash was a multimedia plugin for websites. Now Flash is focused on a write-once deliver-everywhere solution for app developers.
Don't, for the love of human decency, put 45 degree pinstripes behind light-weight text. I sat there waiting for several seconds thinking that some background image would load to spare my bleeding eyes. It never did.
Meanwhile:
"I didn’t have the heart to ask him what blind person he knew that could sufficiently use a web browser on a touchscreen smartphone."
34 comments
[ 1.5 ms ] story [ 78.8 ms ] threadInteresting mix of data. Wonder why Android wasn't listed? The list of Android devices that even partially support flash in this time snapshot is a sliver at best as well.
Care to provide some numbers to back that up? How many of Youtube's, Vimeo's and other big video websites are using an HTML5 player?
And please don't say that's because of H264 video penetration over FLV. Because H264 video (which is sometimes confused with "html5" video) has been the standard for quality Flash video for a few years already, and a 100% penetration of H264 over FLV would have no correlation to the penetration of "Flash video" versus "HTML5 video".
[1]: http://www.youtube.com/html5 [2]: http://vimeo.com/blog:391
Still waiting for facts and figures.
Either way, HTML5 video adoption is certainly not descreasing, and there are several signs that it is increasing.
If you're developing a site with interactive features, shouldn't you look more to where things are going than where they're coming from?
Is the IE8 userbase growing or shrinking day by day? What about HTML5 enabled devices, mobile, etc...?
With that said, I think it's great we're now being more careful about when and where we use Flash. For performance and stability reasons alone, I'm happier it's on less and less sites.
The same is true in .NET. It's great that Microsoft folks want to say HTML5 is their primary development platform so they can sound cool. But unless they are seriously extending HTML5 there's no way it could replace everything available in Silverlight and WPF. Because that's not what it's designed for.
HTML5, as you said, is an attempt to deal with the most common usage cases in a standardized manner and that's great. But on the specialized functionality front Flash and Silverlight still easily have it beat.
I'm perfectly willing to accept that we'll continue to use Flash to gussy-up the experience at the desktop and allow for more complex interactions with media. But frankly, most sites don't need (and users don't really want) that complex a set of interactions.
How many of us have gone to a Flash restaurant site and sat there cursing because it was hard to find what we really needed - the menu, hours, address or phone? Worse, tried to hit that from a mobile phone?
That's why flash is going to get increasingly marginalized, not because HTML5 is allowing similarly complex effects. Flash largely goes in the category of those old DVD ROM authoring softwares: harder to implement, harder to test, harder to use.
Wait until full screen <canvas> ads hog your mobile CPU and you can't even fire up a task manager to end the browser process.
The dream is that we'll be able to harness webcam, microphone, great audio API, and so on, without some proprietary plugin.
The reality is, we won't be able to. As soon as standards are all caught up, Flash will have included 5+ years worth of advancements that will be absolutely necessary for some types of interaction with the user.
Flash may be proprietary, buggy, closed, etc, but life without YouTube and the like would really feel like the stone age more than an improvement.
> The reality is, we won't be able to. As soon as standards are all caught up, Flash will have included 5+ years worth of advancements that will be absolutely necessary for some types of interaction with the user.
The next big feature for the next flash player is full 3d support. WebGL, the html5 mirror of this already has support in firefox and chrome (although, it can't be used on XP or linux with intel due to bad driver support). In this case, the open standards are ahead of Flash.
If we live today and today only, Moore's law wouldn't apply and we could all say for certain that HTML5 and it's successors will be the lead indefinitely. I don't see that happening. If Flash didn't exist today, given the rate that standards are drafted and implemented, would we even be here today talking about audio API or reliable video streaming? Who knows.
Flash is the necessary evil, you see. Standards are driven by demand. But how can demand exist for something that a user cannot experience? Flash. It is the petri dish of web innovation. Perhaps not the best petri dish but a petri dish nonetheless.
They're only ahead if you consider one idea with poor implementation and compatibility in around 3% of the browsers to be "ahead". And by that same logic, Unity3D is already way ahead of both of them, and it doesn't mean anything either.
Flash already has OpenGL/DX enabled on Flash 11 (currently in an open beta), which should be released later this year. Given the quick turnaround we have for Flash Player penetration (6 months for 80% penetration; 12 months for 95% penetration) I can bet Flash's 3d will be at 90% penetration on all desktop platforms and browsers before even WebGL reaches 20% market penetration. There's no helpful headstart here. Like it or not, a Flash Player upgrade gets installed much faster than a browser upgrade. Look at the number of people who are still using old IE versions.
There are no "leaps and bounds" in the HTML world. Even if standards came out as quickly as a single company putting out their own platform, actual implementation and adoption takes a much, much longer time. This is the real world, and that some people ignore it is baffling. There are many reasons NOT to use Flash, but trying to sell HTML5 as a more capable standard is missing the point entirely.
And considering Apple's stance on not allowing WebGL on iOS either, plus Microsoft refusal to implement it, can't see any real advantage in using WebGL coming in the future. Quite the opposite.
I'll stop using that example as soon as I stop having that problem once a week. It's even worse in fashion, where designers for tiny brands continually get suckered into making flash-only sites that are essentially useless beyond being brochureware.
Are there sites that use Flash to great benefit? I have no doubt, although it seems to me the examples you're using are just using it as a way to deliver media. But for 99% of web sites, Flash is a bad idea and provides little benefit. I use Flashblock on my browsers, and the only time I seem to need to unblock anything is to watch a video. That's a bad sign for folks looking for a long-term use of Flash outside of players and video games.
We place the blame on the customer, or the user, or in this case, the technology. Perhaps all along it's just us. And our uncanny ability to misinform the client and leave them without information on accessibility, mobility, scalability.
A year and a half ago Lifehacker was up to 6% not installed and DaringFireball claimed 32% of its users didn't have Flash.
http://daringfireball.net/linked/2010/01/30/flash-lifehacker
I've been trying to come up with a way to articulate how I think about this whole non-debate and here's a first stab:
One of the primary reasons Flash will eventually be unseated is not that Flash is closed -- standards, to a certain extent, are constructed in a relatively closed manner -- but because standards and technologies built on top of standards are more agile.
Flash is an awesome platform, but you can only build on that platform. Standards are great because you can combine platforms and build things not explicitly envisioned. Flash's single-platform approach makes it so that it will eventually have to play catch-up to a whole set of features that are being built by mixing and matching standards.
CSS3 + Javascript, for instance, can be used without Canvas to build really awesome functionality.
Canvas + Javascript can be used to make really engaging content.
Flash will eventually have to compete not with another product on features, but an entire web of organic technologies that don't have to wait for other components to advance in order to release the next stable version.
What will displace Flash is the lack of necessity for Flash, which is (or should be) obvious in 80% of what we see online. There are many things HTML do better, and easier, than Flash, and ignoring that is the real problem. People all of a sudden are trying to make HTML5 be like Flash, and failing at it, instead of remembering why Flash has never displaced HTML in the first place. I'm all of a sudden seeing a bunch of completely inane HTML5 websites that could have worked much better were they built with simple HTML technology, and it's infuriating to think we're back to the insanity of the Flash 5 time, where people would build stuff in Flash just because (something that today's Flash-based restaurant websites are an evolution of).
People need to remember that making snappy, simple, straight forward websites is totally a thing. That's what's gonna keep Flash at bay, in its little corner, as a technology best suited for lots of interactivity and media and embeds or whatever it is. Not trying to make all websites be Flash-like but just with an alternative but barely supported technology.
Here is my view: If what you want to do can only be done with Flash then – of course – you have to use Flash. If you can't reach some people you want with a HTML5 solution (because their browsers don't support it) then you have to use Flash. If you can't reach some people you want to reach with Flash then you have to use HTML5. If you want to use HTML5 at all you probably have to use a Flash fallback since it's quite often the case that some people you want to reach can't use HTML5.
Tradeoffs are always involved and the landscape is changing too slowly to care about anything but current (and not predicted future) usage statistics – in most cases. It always depends on the necessary work and supported features. Ideally you would support both but that's either sometimes not possible or too much work for too little gain.
You might not currently want to use canvas but, I think, video has become a pretty obvious case. Today it's eminently possible to use HTML5 with a Flash fallback with little additional work. MediaElement.js (http://mediaelementjs.com/), for example, is a beautiful and elegant solution to that problem.
It's all about simplifying publishing.
Don't, for the love of human decency, put 45 degree pinstripes behind light-weight text. I sat there waiting for several seconds thinking that some background image would load to spare my bleeding eyes. It never did.
Meanwhile:
"I didn’t have the heart to ask him what blind person he knew that could sufficiently use a web browser on a touchscreen smartphone."
http://www.apple.com/accessibility/iphone/vision.html
I mean, I don't care how jaded and cynical you are, iOS's accessibility support is amazing for the blind.
http://mattgemmell.com/2010/12/19/accessibility-for-iphone-a...