The inevitable: Flash vs. HTML5

[Jackolantern]

So I am at a fork in the road, and trying to decide where to focus my technology learning and development energy. I am interested in making a bit more of a rich gaming experience in the browser. The two options I am most interested in are: Flash (I know the basics, but am interested in really buckling down and mastering it) and HTML5 (I of course already know XHTML, CSS2 and current JS, but would need to get up-to-speed with the enhancements, new libraries, and canvas). I can see many pros and cons for each way. In the end I will likely end up learning both, but since each one will take a lot of dedication, and I don't know what tomorrow will bring, I would like to pick one that will work the best for me now. So, here is the list of pros and cons I have so far:

Flash:
-Pros
* It is here and ready now.
* It works absolutely the same in every browser.
* It has many features that will not be supported in HTML5 even when complete.
* Proprietary: New features can be added immediately, and the Flash Player can be automatically updated long before the user would update their browser.
* Many game-related libraries and engines exist.

-Cons
* Proprietary: Completely the opposite of the open nature of the web. True third party tools are few and far between, and open source is almost unheard of in the core Flash tools world.
* Flash's ubiquity is mostly due to its video-streaming capabilities, not its scripted content. With HTML5 offering video out of the box, Flash's installation base will slowly start to dwindle.
* HTML5's canvas is also poised to strike a blow at scripted Flash content (including games).
* Limited use on mobile devices.

HTML5:
-Pros
* Will work out-of-the box in modern and future browsers. No plugin required!
* Open-source, open-standards technology will create a slew of free tools.
* Open file formats will mean eventually every major web development IDE and tool will be able to work with them.
* Compatibility currently exists on some mobile OS' for many HTML5 features, and plans exist to extend support on basically every major mobile OS.

-Cons
* The standard is not even done. Further changes to the standard could potentially break work already done.
* Wild variations in support between browsers is already being seen, and could grow worse as browsers support more complex features.
* HTML5, even in its completed form as proposed, does not include nearly the features that Flash has. Canvas animation can sometimes be crude by comparison.

So what do you think? I am leaning maybe a smidgen towards Flash at the moment just because of the incompleteness and still young support of HTML5. But there is the thought that a complete, complex game could take 2 years or more to finish, at which point it is difficult to say how far along HTML5 will be.

[hallsofvallhalla]

That is a really hard one. It depends on how fast you are wanting to get a game going. Regardless of the plugin Flash will be around for a long long time and nearly everyone has flash. Most PCs, phones, tablets that don't have flash won't have a browser capable of HTML5 anyways.

I am a huge fan of straight browser based games but the fact of the matter is flash is not that much of a difference on loading and playing.

[Jackolantern]

I actually think at the last minute I have decided to focus on HTML5. Mainly due to the same kind of message that was in the long post I made a few weeks ago. Basically, HTML5 is mostly an unexplored continent. Flash has tons of documentation, tutorials, books and techniques, but almost everything really has been figured out about it already. Figuring something amazing out with HTML5 could lead to some amazing opportunities. So I am taking the road (so far) less traveled and going HTML5. I will return later to learn Flash, because as you have said (and the industry shares the same opinion), Flash is not going anywhere anytime soon. Google has a lot of stock in Flash due to YouTube, as well, since changing every single video to HTML5 would basically be impossible at this point. Flash is going to stick around for a long, long time.

[hallsofvallhalla]

i am glad you chose that road. Like you say, it is a road far less traveled and people always look to the adventurers in the desert horizon returning from a great adventure

[Xaleph]

Good choice, would be my choice too.

Anyway, Youtube already supports html video elements. Many, many movies have been stored on default source files. So most video`s can be accessed regularly by most browsers. Modern anyway.

Still, Flash has been cool features, however, if you look at it from a different perspective, HTML5 + JavaScript + CSS3 ( which are all native browser languages ) are fully capable to completely replace Flash. Currently, the only downside is HTML, which has the bad DOM model, and, CSS i suppose. Which does not allow transitional effects as of yet. Other then that, HTML + CSS is a far better ( and faster ) choice.

Anyway, i`m off to bed. Goodnight!

[Jackolantern]

HTML5 in its current form definitely doesn't have everything to completely replace Flash. Flash animation is much more powerful than HTML5 at the moment. As far as I know, HTML5's canvas doesn't even offer a way to manipulate vectors. You just make the primitive and it is immutable, so it would have to be replaced with a series of shapes to give the illusion of mutation. Also, HTML5's video capabilities are probably not even 50% when compared to Flash, and likely not even 20% when compared to Silverlight, the leader in video streaming at the moment. I have looked around and at least as of 4 months ago (about the most recent figures I could get) a fraction of 1% of YouTube's videos have been converted to HTML5. There is an almost immeasurable amount of videos on YT, and that will likely keep Flash installed on people's computers for a long time to come.

That said, I think Adobe has come close to wringing about as much out of the Flash platform as they can. Some of the latest features feel nearly pointless to me at least, such as inverse kinematics. It sounded really cool but you quickly find it is much more limiting than you think and my understanding is almost no developers use it. Basically the advances in CS5 and CS5.5 basically felt like excuses for trying to sell another copy. On the other hand, HTML5 development is moving fast, and I bet even before the specs are done a 5.1 is going to be in the works with huge further advancements. Each new browser release is pushing the envelope further, and it just feels like a more exciting place to be right now for me personally

[hallsofvallhalla]

where HTML5 will soar above Flash is the streaming. Flash games can easily get above 10 megs and you have to download that before you can play. It is one package. With a browser language you will have many more opportunities to stream the media to download as you go. Will keep players much more interested.

[Jackolantern]

Very true! I was thinking about a way to use a small iFrame or similar along with AJAX to invisibly force the browser to load and cache images on the fly to basically load more of the game without having to shift to another page or load them all up front. There are so many options.

[Xaleph]

Hmm seems we need to define HTML5 first. Because, to my knowledge HTML5 is just a language like HTML4/XHTML. ( not true, I know ) where HTML5 is going to offer a videotag and an audio tag. And the canvas, as well as some new input tags. Video streaming however, has nothing to do with HTML5. Implementation is up to the browser makers, and there has not been an agreement on which codec is going to be used, et cetera.

Also, i believe the canvas element is going to get openGL support, which means hardware accelarated graphics. This however, is also up to the browser makers/vendors as well as the OS vendors.

However, if we skip all that, and all cool functions are going to be implemented, then HTML5 >>>> Flash. Period. Primary reason is that canvas is going to be hardware accelerated. I mean, come on, how cool is that? Way better then software driven animations. You don`t want to know how hot my laptop can get when i`m watching some youtube movies.

Anyway, the above is my definition of html5. This does not include Javascript or CSS. Mostly because both of these are an added layer on HTML. And it will be like that for now. And I encourage it. So, html5 is != flash. However, html5+css+js > flash.

[Jackolantern]

Canvas is only considered a "2D context" at the moment. This hints to the future of a possible "3D context", but WebGL may already have that hole plugged first. WebGL is not part of HTML5 at all, as it is not even considered by the same groups that are handling HTML5 (WebGL is being worked on by some non-profit called Khanos, or something like that). WebGL is still basically a distant and only mildly possible future. I think only one browser has moderate support for it, and some browser vendors have already expressed great concern about implementing it.

Plus, Canvas does not have the full support of the hardware. As usual, it runs only in a cut-off browser segment of memory. So at the moment performance of Canvas is similar to, or less than, Flash. However, looking holistically at the system, Flash uses more memory than the JS segments of the browser. So each side still has pros and cons.

The thing that still concerns me about HTML5 above all else is all of the different browsers variations. There is already a storm of hacks and tricks that are forming to try to get it to work in as many browsers as possible. Of course this will get better as more people pick up native HTML5 browsers, and as HTML5 renderers become common (only Firefox has an experimental HTML5 rendering engine).