The HTML5 and Flash Debate from a Flash Developer's Perspective | Hacker News Reader

The HTML5 and Flash Debate from a Flash Developer's Perspective

youtube.com

31 pointslancer38345 comments

The most important part is when he shows a <canvas> demo running poorly and says something along the lines of "in my day to day work, I can't use this. It won't work as a solution."

The debate right now is one side saying you should use "HTML5" because it's more open than Flash while Flash is buggy, bloated, and proprietary. The other side is saying "I need Flash to give my clients what they want." Just like people started using CSS over table-based layouts when it was pragmatic, people will stop using Flash when Flash stops being the best tool for the job. There's nothing wrong with that.

We are facing exactly the same thing with graphs, we're using a custom canvas based solution but we found out it just isn't up to the task anymore when we need them to display lots and lots of data. Browser lockups are a result because it is just that Javascript intensive.

We are looking at Flash solutions right now to solve that problem.

Yes, they will. It's a big "when" though.

I work with Flash and I don't think it's buggy or bloated - that's overblown. It's not perfect, but it's definitely much better than people paint it to be (specially people who don't work with it).

In the same vein, HTML5/canvas is unreliable both in terms of performance, adoption, and consistency among different environments. People like to think of "standards", but when it requires as much hacks as any other 'standard' work nowadays do, and when it's not even available in half the browsers out there, I wonder how can anyone think it's the golden goose people like to claim it is.

> I work with Flash and I don't think it's buggy or bloated - that's overblown.

Depends if you're using Windows or not. On Mac and Linux, the performance and reliability are pretty bad (compared to other modern software, anyway).

> Just like people started using CSS over table-based layouts when it was pragmatic,

And vice versa. There's a nice table-based layout on the Google front page still!

I don't know about the old Google home page but the current one I see only has one <table> and it's wrapping what I would consider to be tabular data.

I'm looking at google.ca (it redirects me there) and it has the blank space to the left of the search field as a <td>.

Look here: http://www.google.com/advanced_search

Do you consider that mass of table usage to be for tabular data?

People (in general) started using CSS over table-based layouts way after it was pragmatically beneficial, because they didn't want to learn new stuff or were stuck in the old mindset (e.g. the design of Hacker News).

And all those people not using CSS layouts for non-pragmatic reasons actually held up progress because it stifled the market for tools (and skills and frameworks etc.) that would work in the new manner.

So good analogy, but I think it argues well for the opposite of what you intended.

I'd argue CSS took hold early: it was in an era where you needed a lot of hard-to-maintain hacks to get it to work in all browsers. Today as in the past, many people get CSS layouts just usable in IE 6, while they look as intended in standards-compliant browsers. If you want a site to look pixel-perfect in all browsers, it's still not pragmatic to use CSS rather than tables (or Flash). What's changed are the constraints: more people started using browsers that weren't IE 6, all the other browsers became more consistent with each other (or use the same rendering engine, Webkit), and clients have become more OK with websites looking different in different browsers as they've become more used to the way the web actually works.

But this will eventually boil down to a difference in opinion, and I agree with your point that it would be better for everyone if more people were quick to adopt technologies that are better in the long-run.

Seems humorously apropos that it's in 10 minute long Flashy (errr HTML5y) video format rather than skimmable text as I'd prefer.

Admitting he/she doesn't normally bother to read points in entirety - upvoted.

Ignorance is bliss.

Woe to the author who fails to provide content in the exact format you require.

I don't think he would be able to show the performance of HTML5 and Flash demos using text, even when coupled with some images. For that it requires video, and I prefer a single video showing all the "article" and demos together than clicking play on five or six different embedded players.

Useless FUD. Yes, let's compare a 15-year-old mature platform to one in its infancy, and dismiss the latter on grounds of the performance of its more esoteric features. We get it, Flash developers fear change; cry me a river. Considering the relative rates of progress of Adobe, the only company that can advance the Flash platform, and every other tech company in existence currently working on HTML5, I know where the safer bet five years from now is, barring Adobe doing something smart and opening up the Flash runtime. So HTML5 is still rough around the edges, big surprise there.

The fact is, for video playback, it does just fine today, even on limited mobile devices, and if sites were to only replace Flash with HTML5 for video and audio playback, we're talking about 90% of Flash's popular use right there. As HTML5 matures, including content creation tools and the performance of various browser runtime implementations, it'll eat into Flash's share for other tasks as well. Apple is creating a market demand for HTML5 development tools and expertise, which is the first necessary step in order to bring the platform to maturity. After that happens, Flash developers will jump on board and forget this thing ever happened. Until then, just keep coding your crappy Flash ads and leave the rest of us who actually want some progress for the open web alone.

Yes lets compare them. Why shouldn't we?

The mature platform and an immature platform.

HTML5 holds promise but it's simply not capable of delivering anything close to what the flash platform can.

Just because something is new doesn't mean it's better. Let it naturally outperform flash once it's better, nothing hinders that.

So if you believe HTML5 will be better then let it prove itself.

So far it's no where near.

My point is that HTML5 does do SOME things better, some things that Flash is quite commonly used for, like video playback without pegging my CPU quite as much. To say that all of HTML5 is useless because it doesn't replace Flash for games or because Canvas and SVG browser support aren't quite up to snuff, is idiotic. That support will come, and HTML5 will become a practical replacement for those functions as well at some point. For now, people will keep using Flash for that, and figuring out something else for iDevices. Adobe Flash CS5 can already export to Canvas, so the content creation support is obviously on its way. Flash isn't the end-all web platform either, and webGL can do many things in 3D that Flash developers could only dream of. Flash has its place, but that place is at the periphery, not the core of the web; Flash-only sites deserve to die in a fire. To think that HTML5 will stand still now that there's a market pressure to support it is incredibly short-sighted.

You should really learn to put forward your claims without having to resort to strawmen. Who said HTML5 was useless? No one did.

That you get better performance with video using HTML5 does little to remedy the fact that playback of video is really all you can do.

You try and seamlessly integrate video, with logic, dynamic text, vectors in HTML5 and you will be pretty disappointed.

The Flash isn't the be all end all. But what it does it does so many times better than HTML5 for quite some time.

I agree that Flash is better for many things, that the tools that support the format support greater capabilities, especially for things like access to webcam and mic hardware, and obviously 2D and extremely basic 3D interactive games and presentations, and Flash will likely be better for some or all of these things for a while longer. But I also think Flash is widely abused, with ads and other site elements doing things like playing videos or animations and consuming resources (precious on mobile devices) without user intervention, and worst of all designing entire site interfaces in Flash with no HTML fall-back. I support what Apple is doing because it really forces site operators to reevaluate how deeply they rely on Flash for their web presence, and because it drives the development of content creation tools that support HTML5 content, as well as progress in browser support. I acknowledge that it's painful to undertake a transition like this, especially as you've got clients asking why they have to pay this much more to get their content on iDevices now. But in the end, we get a more open web, which means more user control over the experience, which is a good thing for users like you and me.

Given the CPU usage of the HTML5 examples shown in the video it seems like Flash is not alone with that problem.

With regards to designing entire sites in flash. Go to thefwa.com check out some of the sites on that portal. Then tell me why you would want an HTML fallback for that? That is kind of like saying that all video should have transcription of what is being said.

Sometimes there is no substitute.

With regards to the open web argument. Since we are in agreement that Flash isn't the be all end all of the web, then what is the problem?

If HTML5 is better then it will eventually outcompete FLASH. Until it is, support both. That is what I don't understand with Apples claims.

HTML5 is right now rather resource intensive as shown in the video. It actually going to be worse for the battery life of an iPad than flash is.

It's a hollow argument.

At least any browser developer can do something about the HTML5 performance characteristics, including giving their users greater control over activation of certain elements on platforms where that may be especially critical, such as ones where you don't want to be wasting batteries on content you didn't care about seeing in the first place, like ads. Only Adobe can fix Flash's issues, and they've not done a great job so far.

As for that site you linked, you may want to find a better example. All that stuff seems relatively straightforward to implement in CSS/javascript and Canvas, and that would probably be much less painful to actually use, if possibly more effort to create. That site was a terrible experience, and a perfect example of why encouraging this kind of behavior by web developers a bad idea.

So you actually believe that HTML5 wont be used for making ads once it's matures?

Not the site itself. What it is linking to.

There are HTML5 ads today, and they're already extremely annoying, possibly even more annoying than Flash at the moment, because they're so badly designed, often completely broken. But eventually, there will be some kind of backlash, and browser makers will block the more annoying behavior or lose ground to competitors that will. I never said the transition was going to be easy, and there's pain all around right now, but at least it's exciting, and it feels like we're finally making some progress towards a better destination. This pot needed to get stirred a bit, and that's what the HTML5 platform excels at. With the Flash platform, there is no competition, and it has drifted away from the user's control.

sorry but how do you expect browser makers to block more annoying behavior? Thats nonsense.

Regarding competition.

So your idea of competition is how apple does it by not allowing Flash on the iPad and iPhone.

That's not competition.

WiseWeasel, dude, you're all over the place. Like many of the "vs." crowd. I am gonna hit all of your points, so you don't Republican the debate.

ThomPete - you're a voice of reason and contemporary fact.

To the "vs." crowd: try and read all the points; not skimming and then commenting without paying the respect to do some legwork.

* ads - HTML5 ads will exist. The easier they are to make, the more of them there will be. Advertisers will use whichever has the widest audience. In fact, advertisers will be more encouraged by all the circle-jerk mob-mentality surrounding HTML5.

* video playback - for one, HTML5 doesn't play better or with less CPU. I think about 10 or so YouTube videos have proven this (visually, not blog opinion rhetoric). I'd link a few to you, but let's face it...you're not gonna look at them. You'd be forced to see proof. (If you do manage to be overcome with a neutral state of mind, do some research)

* cpu - another red herring. So far, I've not seen 1 canvas+js example (of which has a close equal to a Flash equivalent) that outperforms Flash. I wouldn't expect it to, either, since Flash is compiled and JS is interpreted. Bottomline: an immersive experience is going to take some CPU. Realize it applies to all technologies and get over it.

* ipad/mobile - Flash depletes battery. I agree. Equivalent canvas+js experiences use similar CPU. Taking that into consideration...the battery specifically hates Flash? Please.

* greater control/poor coding - Do I really need to point out that this applies to everything? Not just Flash?

Let's consider that HTML5 is being implemented in 2010. I've been on the 'net since 1995. That's 15 years for video to make its way into the browser -- without the "dreaded" plugin tech.

You know what's dreaded? Not being able to move ahead at the speed necessary to make great, immersive experiences. Being tied to a set of slow, archaically-driven, standards.

Flash (or more specifically plugin arch) is a petri dish for the web's innovative future. It can't and shouldn't die. If anything, it should become more open, more accepted and more accessible.

I'd be fine with it being open. I'm just very happy to see a vendor push website operators to not completely rely on Flash, a proprietary plugin controlled by Adobe. If Adobe were to open up the Flash runtime, I would have zero issue with it, and I'd even wager that Apple would turn themselves around on the issue. To have the entire web at the mercy of Adobe is a bit much for me to swallow. Flash is great at a few things, but websites completely reliant on it needed a swift kick in the pants.

I agree with what you're saying.

However, I am alright with a website being 100% reliant on Flash just as I am alright with a website being 100% reliant on video. The key issue here is openness, not platform.

To give a more practical example of the benefit to users of an open standard that browsers code to versus a proprietary plugin that a single vendor controls, I can point to the fine-grained control users have over javascript. Popup-blocking is now a standard feature for browsers, and most give users the option of controlling various other features which have been abused by web developers. The way this works is one browser will support the feature allowing users to limit javascript behavior, they start stealing marketshare from the others, and then the others are forced to adopt the capability to keep up. With proprietary browser plugins, however, the content is a black box handled entirely by the plugin runtime, over which the browser has little control. The user can customize what the plugin vendor says they can customize, and there is no drive to cater to users' needs instead of content partners, as all browsers must use the same implementation of the platform. Flash is all-or-nothing, and it's what Adobe wants it to be. Web standards are piecemeal, and they are whatever the user wants them to be.

I'm not going to comment on the subject matter, but this guy is pissed. You can tell by his nervousness -- shows that he is passionate about his tools. Good for him

Passionate about his tools or really angry because he thinks the other arguments are stupid. Either way, it makes no difference in his argument.

Or it shows that he's freaking out, because some people are telling him he's hitched his cart to a dying horse, and trying his best to rationalize them away.

Flash developers (read: not designers hacking together an intro/banner) don't care if the Flash technology dies.

Flash developers want the development experience that Flash offers. HTML5 doesn't offer it...yet. When it does, I personally would love to migrate to it.

Your assumptions are pitiful and illogical.

Exactly my thoughts.

Let HTML5 prove itself instead of insisting that it's better because it's open.

That isn't an argument anyone outside academia can afford making.

The "vs." crowd won't watch/care for this. It has proof in it.

wasted 10 min of lifetime. Really useless ...

Why?

because it does not tell anything about flash vs html. It's just a statement from a flashdeveloper that is speaking about e.g. performance and showing cpu usage. It would be great if flashdeveloper would read Tinic Uro's Blog and understand boundaries on systems and browser. The performance of a flash animation/website also depends on how it was done. I've seen "BLING BLING" made by a flashdeveloper that uses 90-100% CPU. The same "BLING BLING" running on the same machine/browser done by an other flashdeveloper uses around 50% CPU ... Bla bla and no standardized tests.

> Bla bla and no standardized tests.

You mean like anything else? Thank you for admitting you didn't have a point. Most people can't do that.

A simple HTML5 demo ran many times slower than a much more complex Flash demo. And the HTML5 one was made by people who are trying to convince everyone to follow them.

Do you really think their code is that bad?

I found it to be pretty interesting.

I tend to fall on Apple's side when it comes to not wanting to be forced to support a technology that it is against. That being said, seeing the CPU usage on these simple HTML5/JS games was a bit shocking - no idea if that is due to sub-standard coding, or if it's inherent in JS, but it seems to blow a hole in the "Flash is a resource hog" argument.

I am wondering if better developer tools for creating graphical HTML5/JS and/or JIT compilation will help with this. Clearly an iPhone is capable of amazing things graphically, but that's all been on the native app side.

> I tend to fall on Apple's side when it comes to not wanting to be forced to support a technology that it is against.

How is that Apple's side? Allowing something to be installed is not by any means the same as being "forced to support" it.

Apple does allow Flash to be installed on Mac OS X, and actually includes it with the OS.

Right now there is not a way to "install" something on iPhone OS that would then become available to Safari and other apps. To do that, they'd have to build a plug-in architecture for iPhone OS.

That kind of CPU usage in HTML5/JS games is common, even on higher end systems.

so.... flash developers are to 'cool' to write things down? Are the people upvoting this actually watching it???

It is more of a (biased?) demonstration than a debate, and obviously video is a better choice.

I've watched it and also read several accounts in blogs. I'm one of those fools who looks at both perspectives, rather than being lampooned into one popular opinion.

Reply on news.ycombinator.com