Why HTML5, and the future of exporters

Hello,

I want to speak my mind about something. I follow your products since construct classic and I have to say that you make the best game maker with the worst choice of technology (directx then, html5 now).

I'm making a game for the rotary competition, while making an iOS and android version, trying in the process every single library and "game maker" out there (impact/lime, gamesalad, stencyl, game maker and corona).

Apart from several issues that in my opinion keep costruct2 from being a professional tool (debug tools, if conditions inside events and reusable functions with object referencing would really make a difference) I can safely say that construct2 is bar none the best "Game maker" out there, it's intuitive, extremely powerful and has incredible features that here are taken from granted but that are hard to "code" in other solution (pushing towards other objects, pinnning, layer rotation and coordinates relative to the screen as opposed to the sprite comes to mind). Sadly html5 is a joke and I'm sorry to say this.

No other "game maker" or even library (impact/limejs) has the same performance as construct2 for this I can vouch, you guys are light ahead performance wise, but still.. the game I'm making for the rotary competition has been stripped of almost everything effect wise I had in mind at the beginning because my pc (that can run crysis in ultra) couldn't get past the 30fps mark with incredible lag spikes. The same exact game in corona runs at 60fps on an iphone4 with physics and particles (as opposed to custom movement behaviour in construct).

If I try to run the same game on an iphone it never gets past 2fps, to be honest I've yet to find even a simple example of those included that can run at more than 20fps, but this is not your fault it's html5 which is a joke. Not only this, but with html5 I can steal the entire content of a game with a couple of clicks, exposing code and especially assets that can be recycled in other games (and then good luck in court).

Meanwhile INFERIOR products like stencyl or even gamesalad (which is the most limited piece of software I've ever seen) gets loads of subscribers simply because they have an iOS exporter, while charging obscene amounts of money for their licences.

The way I see it the future of casual gaming is in the mobile market be it android or iOS and in cross platform development for indie games (just take a look at all the indie bundles) but absolutely not html5 that it's looking more and more like a passing fad in my opinion, too many ideas that drift away, too many different way of handling things across browsers too many issues, granted it may be that in the next year things will stabilize and html5 will be the best thing ever but I'm having an hard time believing this and even if it'll be the case construct2's competitors will have the upper hand by offering mobile support.

The way I see it you are standing literally on a gold mine, you have the best tools I've ever seen and Ashley really knows what he's doing under the hood, you should just consider creating other exporters.

The solution is not appmobi (which by the way is better than nothing) that is another hack on top of an hackish technology but a real iOS and/or Android exporter which, not only me, but a large portion of your users would be more than glad to pay for. If this is not possible (I understand that it's not an easy job) at least make an exe exporter like construct classic, not directX but SDL/openGL to allow for multiplatform development: you'll gain thousand of subscribers.

You have the best community, the best documentation on top of the best tool, what are you waiting for?

I hope I haven't offended anyone and I would like yours and the community honest opinion if I spent so much time writing this is because I believe in you and I want you to succeed with all my heart.

Peace

This has been beaten to death in the forums by now. More exporters are coming down the road. Also html5 is still a new technology and is always getting faster, it's already much faster than it was when construct 2 beta first arrived. Maybe stick to one of the other tools for now if you can't wait.

Realistically, I don't see how they can write those exporters now. Development of C2 itself would either temporarily halt or be too slow to be worthwhile. This would happen because writing an exporter will take a lot of time and because each update of C2 would necessitate an update of each exporter. Once they reach "1.0" I'm sure they'll turn to exporters. What would (probably) be more useful would be to release an EDK (exporter development kit) so ambitious people like yourself can start writing exporters (and charging for them, if desired). It's hard to imagine, but important to remember Scirra only has two developers and only one of them works on C2. One person can only write so much code.

what danny said. Also wanted to add, I doubt html5 is a passing fad. html is the basis for everything on the web, and has been since the beginning. I don't see any browsers (mobile or desktop) reverting back to not supporting the latest version of html, but I do however see several different browsers all competing in this new arena for speed, where there's plenty of room for improvement.

also, the detailed answer to your question is here.

edit: I agree with trevor10. I don't see it being possible to create in the near future, but I do hope the edk arrives before the next exporter. it's something I would definitely be interested in trying my hand at. We could even create open source community exporter projects, recruiting several forum members to work on different parts of an exporter

I can't say i don't agree. Right now the situation for HTML5 tech on games is worrisome at least. Canvas just doesn't cut it. Not suitable for games at all like it is now. WebGL lacks support on most platforms. Only works great on Desktop/Chrome as far as i can see (update: it appears there's initial support on some Android devices). I won't even mention IE. The truth is that Construct is a hope for the future. You won't get your fancy game running at 60fps on all platforms right now. From what i have seen Appmobi and similars are not the solution definitely. Yet i have great hopes for the future. But in the present i will have to use other engines/frameworks (Unity?) to get my games on iOS and Android with max performance. I'm a seasoned programmer/coder/whatever so that's not a problem for me but still... T__T

The solution is not appmobi (which by the way is better than nothing) that is another hack on top of an hackish technology but a real iOS and/or Android exporter which, not only me, but a large portion of your users would be more than glad to pay for. If this is not possible (I understand that it's not an easy job) at least make an exe exporter like construct classic, not directX but SDL/openGL to allow for multiplatform development: you'll gain thousand of subscribers.

You have the best community, the best documentation on top of the best tool, what are you waiting for?

I understand the frustrations and concerns, but consider it from Ashley's point of view. He already has a mountain of work to do! The IDE is incomplete and missing a lot of features like functions and containers - those need to get implemented first before any work on other exporters can happen.

I think HTML5 was the best choice to start with. If he had started with an iOS exporter first instead of HTML5, it would've been a vastly more limited audience that could have used C2 from the start. With HTML5, everyone can use it even if it currently isn't as fast as it could be on some platforms. Also keep in mind the increasing speeds of hardware and improvements of software - iOS 5 was a 2000% speed increase for HTML 5 over iOS 4!

I know Ashley makes it look easy with the weekly releases, but remember he's one man coding C2 all by himself! He still has to sleep! Patience! :)

I completely agree with the OP. HTML5 is basically useless. I too have been very disappointed with the performance of HTML5. I have many Mobile devices, tables, chromebook, PCs and Mac and there is never a common performance between them to make 1 game for all devices and expect it to work well.

Appmobi is our best bet if we do not get real exporters. I honestly say there is no decent game creator for Android, there are a few for IOS.

If C2 could export to Android, it would be a huge success because Ashley would own the market.

I used to be a teacher teaching kids how to make games with Game Maker [another software by this actual name] for five years, and I have to agree with 0plus�s opinion that Construct 2 is one of the finest game creation software ever.

How I wish I could go back in time and teach my dear students Construct 2 instead.

Construct 2�s platform and solid alone makes a mockery out of Game Maker�s.

Things like fade and destroy, small things like this that really saves a lot of time and clear memory leaks automatically.

The Touch object that have the ability to interface with the accelerometer and gyroscope, all these I hope users will not take for granted.

The creator of Construct 2 is the salvation to the ideal of a real game making tool in my eyes.

Take your time my dear Construct 2 developers.

I know you have a lot of things to accomplish, we see your grant idea and we will support you all the way.

When you are ready, exporting to ios will be nice ;-p [but not the wrapper kind that jail breakers can still access game asserts]

lucid Currently I work as a web developer (not designer) and I've been doing it for 6 years now. When html5 and css3 were announced I was excited, I made tons of proof of concepts, but after a while the reality started to sink in: it was internet explorer 6 all over again. Too many discrepancy between browsers (which still remain now, just take a look at the changelog of construct and how many bug fixes specifically to a platform are there). Also html5 is not a standard yet (http://en.wikipedia.org/wiki/HTML5#Standardization_process) and the more time passes, the more the situation worries me. I think that html5 is perfect for video (vimeo) and audio (grooveshark) and to some extent some interesting web interfaces (which will be limited like the fancy flash one are) but games? Mmmh I'm not so sure.

Even (and it is a big even) if we'll be able to get a stable platform (performance and compatibility wise) which I doubt as javascript is interpreted and thus inherently slow, the assets problem still remains.

You invested time and money to develop a game, you hired a musician and an artist to make your music and graphics, do you really want someone to steal your assets?

Again I don't want to offend anyone, I'm just stating my opinions because as I said believe in construct and I really really want it to succeed.

HTML5 is for sure not the best for gaming as of right now. Too much unnecessary discrepancies on what the standards should be and as I see it's all corporate stupidity. Also, to be fair, it is still a baby and much has to be done for it to be a full fledged gaming platform.

Ashley has made it clear, that he first wants to be done with the IDE and the HTML5 exporter before delving into exporter hell. I too, am an EXE exporter supporter, as I believe it is the best platform, for many reasons that cannot be denied ( granted not for cross platforming or portability ), but know that this is not an easy task and with one programmer alone, as of right now this is not possible. And if the EDK isn't released yet, nobody else can do this.

Patience is key factor, and it is clear that only ONE exporter is not, under any circumstances, a good choice. Remember that Both C2 and HTML5 are not complete yet and still in development. When C2 supports more exporters, I am sure it will dominate every single one of them. After all, it already makes every other tool look like either a joke or a hassle.

One way to deal with the exporter hell would be to be able to use C2's projects in CC?

No big loss, as CC is EXE only, DX9 is enough for the software's capabilities, unless 3D comes into the game, which I highly doubt.

HTML5 is spanking new, so relax. Remember that computers increase their firepower all the time, tools improve performance with each new release. Give it a year or so for Construct and HTML5 to mature.

I'd prefer that Ash would give us EXE output as well. I'm not yet able to "think" in gaming through browsers, so I visualize my games being played through keyboard/gamepad.

Either way, C2 is still an infant, so let it grow.

I agree with C2 and HTML5 being a baby as of now. C2 can't really do heavy games yet seeing as it runs on a browser and a platform that's not yet "standardized" or whatever. But I believe that completing the HTML5 exporter is easier compared to other exporters (I think) so I'm patient enough to wait for it to be done.

I dont think you offended anyone with your post Oplus1, Its important for us to know were using the right tool for the job. Im glad your sharing your opinion. Here is mine :)

I see a lot of people getting frustrated trying to make mobile games at the moment. Correct me if I am wrong - but the mobile platform just isnt ready for C2 yet (its really not C2's fault). Maybe Ashley could spend a lot of time tweaking it and get slightly better performance - but at the end of the day until the mobiles change the way they handle html5 its like beating your head against a wall. Why put all the work in when mobiles will catch up on their own?

Dont get me wrong - I feel your pain! When i first started using C2 - I was using it with the sole intention of making Iphone games. However, its time for a new game plan, the App store is not the only way to make money with your creation.

I find it strange that people dont seem to be as excited as I am about Kongregate. They will potentially sponsor your game paying between $250-$2000 as a one off payment - and then you still generate money from add revenue as well. If you dont get a sponsorship - then your free to post your games on all the portals you can find (increasing your ad revenue).

Right now - there is nothing stopping anyone from making a really nice game in HTML5 with C2. Your right about the fact that not all platforms support WebGL (and that it is absolutely needed to really be able to play any intensive games). What I intend to do for my game is simply add a note at the beginning of the game saying that unless you are playing with a WEBGL enabled browser your likely not experiencing the game as intended.

We cannot control everyones experience in our games like you can with a flash game for instance yet (we need to accept this). HOWEVER - Our games arent going to get dusty! (From what i understand web games have an extremely long shelf life). We have the tools NOW to make fine games! Maybe only X percent of the market can truly experience it the way we want them to, but things can and will only improve (and of course your game will still be there when it does).

IMO we should be focusing on using C2 to its maximum potential (IE WebGL) - Not stripping projects back to bare minimum - and then even still, frustrating yourself in order to get any little performance boost you can.

Thanks for sharing your thoughts! It's always interesting to read analyses like this, and thank you for your kind words about Construct 2. However, I disagree with some of your points :)

the game I'm making for the rotary competition has been stripped of almost everything effect wise I had in mind at the beginning because my pc (that can run crysis in ultra) couldn't get past the 30fps mark with incredible lag spikes.

Was it running in WebGL mode? WebGL usually solves any performance issues. If not, updating your graphics card driver can help, but yes, it's a problem that you have to do that. With WebGL enabled, it's going to be pretty close to what the native performance is, so WebGL really is the future of high-performance web games. One of our WebGL benchmarks on my computer got over 14,000 sprites on-screen at 30 fps in Chrome. That's vastly more than most games need to keep a steady 60 fps.

f I try to run the same game on an iphone it never gets past 2fps

Is that iOS 5? We're aware of performance issues on mobile and we're working on some things to make it work better with HTML5 (not whole new exporters). Hopefully some things will be appearing in the near future. Performance tips also has some advice, and since mobile devices have much weaker hardware and currently have weaker software too, you should not expect a game that runs well on desktop to run well on mobile - you have to design it for mobile from the start. With some careful tweaks, I could get Space Blaster working at about 20-30fps on iPad, but admittedly it still wasn't great. The performance tips article has some advice specifically for mobile.

ot only this, but with html5 I can steal the entire content of a game with a couple of clicks, exposing code and especially assets that can be recycled in other games (and then good luck in court).

A common misconception is that the code can be stolen - if you enable 'Minify script' on export the code is mangled in to an unreadable mess, object names are removed, and so on, making it near-impossible to reverse engineer your project. Yes, it's easier to steal images, but your events and logic and project setup are probably safe, and there's always copyright legislation on your side if anyone nicks the images.

he way I see it the future of casual gaming is in the mobile market be it android or iOS and in cross platform development for indie games (just take a look at all the indie bundles) but absolutely not html5

"A good hockey player plays where the puck is. A great hockey player plays where the puck is going to be." We're certain HTML5 is the future of gaming on both desktop and mobile. Yes, it's not there now, but it will be. When we started the HTML5 exporter no browsers hardware-accelerated it so it was really slow everywhere. Now, with WebGL, it absolutely flies, it's just mobile is a little behind desktop. For example, Google just released a beta of Chrome for Android, complete with the super-fast V8 javascript engine and GPU-accelerated canvas, and I'm sure they're going to add WebGL in future too. That's where the puck is going. It's no more of a fad than Flash was.

ou should just consider creating other exporters.

This one is pretty easy to answer, not just because we believe HTML5 is the future, but because we can't. For some idea of the time scales involved, we started the HTML5 exporter in January 2011, over a year ago, and it's still not done. We can't yet afford to hire anyone else, so taking on another huge project like another exporter would be suicide for us - we'd have lots of stuff being worked on, and nothing finished, which would make us uncompetitive. Also, given the time scales, by the time we're done, HTML5 would probably have become really solid great platform on mobile anyway! So we'd have wasted our time. We're more than happy to focus on HTML5 only for the time being, since we believe in it 100%, and can't actually afford the time to work on anything else. It's where the puck is going.

Thanks for the reply Ashley, yes the game renders in webGL, in the end it was the physics that were creating problems, I currently can get around 60fps with almost a thousand objects on screen, but my computer is a beast.

My iPhone is still ios4 because if I'll be able to get it to run on my iPhone it's fair to say that it'll run better on ios5, I know that it lacks in performance, but I don't think it's a good idea to target only ios5 (although possible).

The minified scripts can be reverse engineered (http://jsbeautifier.org/) but I agree that can be said also for flash.

I understand what you are saying about html5 and while I don't agree I respect you belief in it, the problem is that currently the "hot" trend in games are mobile apps and without proper support you risk losing a big market share to inferior products.

I mantain that construct2 is the best game maker ever made and I'll never stop using it (if ever only for prototyping), the day I'll be able to publish game on the app store with it I'll be the happiest man.