Hacker News Reader

Comment by derefr | Hacker News Reader

So, what's this other thing we need to have to be able to replace flash? What else is flash currently doing for everyone that's so unique?

View on HN

Delivering a cross platform environment that provides the same functionality with more features.

Don't get me wrong here, the first sentence may sound as Flash "fanboyism", but i think it is rather objective. Let me explain:

- Flash Apps work on Windows/MacOSX/Linux/smartphones/tablets with the exact same API and no difference in implementation of that API (in contrast to IE/Webkit/Opera/Firefox that implement different features and even differ in details when it comes to the implementation of HTML5/CSS3)

- Flash offers additional APIs that are not available in HTML5 (Webcam access for example)

- taking this further, Air enables cross platform installers, integration into the OS, even more APIs

And this can be explained really good, because Adobe can implement new features much faster than HTML standard committees can agree on new HTML features (plus a variety browsers that need to support those). So this shouldn't surprise anyone. It's wishful thinking when people claim that HTML5 can replace Flash in _every_ regard (and here i mean 100% alternative, which can't be true if you just take the webcam access). HTML5 can replace the majority of todays Flash applications. At least i can't think of many web sites that depend on Flash that wouldn't be possible with HTML5 (Websites that make use of Webcams being one of them). And i hope all those navigational elements, advertisings and whatever people chose Flash for, will be replaced with proper HTML in the future! To implement a full fledged multimedia website is harder with HTML5 than with Flash, imo (think of cross browser compliance and difference in performance between browsers).

Just to make my standpoint clear: I would not create a website with Flash nowadays, even if it means less effort. HTML5 is for the web. Flash should die. The "fact" that it can be extremely difficult to create those web sites remains, though.

Regarding the article:

Am i missing something or isn't this just another JS-framework? There are already hundreds of them out there, why is this labeled as "Flash alternative"? Unless i am missing a point this looks like nothing fundamentally new. Seen in the light of the headline every JS-framework is a Flash alternative.

It seems like the majority of things people use Flash for, then, could be replaced with a cross-browser plug-in that exposes a unified API for accessing resources outside the browser (i.e. webcams, microphones, whatever Air does, etc.) from Javascript, and is updated by some process quicker than a web standards committee—a "Cool New Thing engine" of sorts. If it allowed browsers to override its features with their own, each feature could "graduate" from the plug-in's codebase when it became implemented natively by all browsers.

So to get rid of the Flash plugin you are proposing a new plugin which is not a HTML standard. Who is supposed to write such a plugin and how would it differ from Flash, then?

Chrome Frame comes to my mind, this covers IE6,7,8 but not Firefox. May be a good candidate, but Chrome Frame has been around for a while now, and i haven't heard of web sites using it. Chrome Frame could lead the way to "one standard implementation of HTML5" which could at least eliminate some of the major pain (performance and cross browser differences). It's a nice thought, but somehow i have my doubts about Chrome Frame becoming that popular. Too much politics in the browser war. Also, many enterprises don't allow users to install that, which in turn makes it a no go for most web sites.

I'm suggesting, I think, that Adobe drop the current "target" of their flash plug-in as a client-side embedded viewing frame for a special filetype that's actually an executable running in a proprietary VM, and just make it into said plug-in, basically a support framework for browsers. Expose everything that's currently exposed to Actionscript to Javascript instead, in other words.

It's not something that Adobe would ever want to do—but it might be something they eventually will be forced into, after they realize no one is really using Flash all that much any more.

That bridge already exists ... Flash and JavaScript can communicate with each other and in the form you're describing - http://www.happyworm.com/jquery/jplayer/latest/demo-01.htm will use OGG if it can or MP3+Flash behind HTML/JS/CSS.

I'm skeptical about Chrome Frame's adoption. It sounds like just a way to trick IE users into using Chrome.

If they wanted the benefits of Chrome, why not just run Chrome in the first place?

"could be replaced with a cross-browser plug-in that exposes a unified API for accessing resources outside the browser (i.e. webcams, microphones, whatever Air does, etc.)"

...aka Flash?

Capturing webcam and microphone input.

So why isn't anyone pushing to include that in HTML5? (That is, HTML5 the let's-do-new-stuff-with-browsers movement, not specifically the markup language standard.)

Because you already mentioned the difference:

HTML is all about a standard for everyone.

Implementing new features on your own is not much different from what Adobe does. For example: Let's say Firefox implements an API for Webcams. Will MS implement the exact same API in IE? Unlikely. First of all, the have Silverlight (i am supposing that Silverlight can access the Webcam as well, but don't know for sure). MS wants to push Silverlight and defend IE marketshare. With MS blocking this feature, it essentially means that noone will use it because every website is developed with the first requirement "Has to reach the vast majority of users", which in the end means "IE support". They would eventually develop an own API, which turns out to be incompatible to Firefox' API if pressure is high enough.

Now, if you have the thought "but if MS/Apple/Google/Mozilla/Opera would agree on that new API", that's what the standardization process is all about.

I did actually mean "the standardization process", I was just differentiating from the part of HTML5 that's about, well, HTML. This is more of a feature that belongs in the DOM exposed to ECMAScript, which the HTML5 movement seems to be all about pushing forward the standards for as well.

What you describe is exactly the opposite of how HTML evolved up to this point. Usually precisely what has happened has been that browsers implemented things that the others didn't support, got some traction, were widely copied, and then, finally, standardized.

It wouldn't matter if they did, there's always going to be another that that they coulda/shoulda/woulda done that Adobe, MS, Unity etc will do first.

Another area where HTML5 is really weak is games, at the forefront is some great examples (http://www.kesiev.com/akihabara/ seems to be popular) but they're really only admirable because they're "not Flash" - consumers only care about the result, and the result is sub-par for modern casual gaming - http://www.kongregate.com/games/IriySoft/cursed-treasure-don....

HTML5/JS/CSS is obviously going to do a lot more than the previous, current antique standards and it's going to be equivalent or even better than Flash in some cases. And if Flash/Silverlight/etc stopped developing right now that might still be true next year.

See http://dev.w3.org/html5/html-device/ -

"This specification defines the HTML device element, to enable video conferencing from HTML applications."

My guess is that the flash tool-chain will be a lasting reason that it won't get instantly replaced by any of these new frameworks. There are a ton of developers and designers out there that know how to build flash and don't want to start learning new tools in unproven frameworks.

That said, I think the more and more bloated nature of adobe's tools will become a huge weakness. Developers will be able to build tools for much more specific work-flows that feed directly into specific use cases. Sproutcore and this one seem to be aimed at the 'desktop feel' app development for html5, I'm guessing there will be other frameworks and tools for creating games, vector animations, silly interactive websites, etc. Each of tho uses should grow it's own tool/framework specific for the task

The scene graph? I think that is what they call it - but that is probably about all (that and the tools that go with it)

vector animation engine? I don't know much about SVG though.

http://raphaeljs.com/

and streamable. I remember in the 56k modem days the main Flash's feature is streamable Web multimedia.

Reply on news.ycombinator.com