Congrats to the Rocketpack guys in Finland! I met them over a beer a few month ago - they definitely deserve all the success they're having. Maybe I should have accepted their Job offer back then, but I had (and still have) way too much fun with Impact :)
My understanding is that RocketEngine is a complete IDE to develop games - so you write your code directly in the browser. I haven't personally seen it in action though.
For Impact you write your code with whatever editor or IDE you want. It's just a bunch of JS files. Impact also features a tile based level editor that runs in the browser and (surprise) generates level data in JSON format.
Impact is somewhat more targeted at arcade games and not so much at FarmVille-like stuff. It also uses the HTML5 <canvas> element to draw everything, instead of moving around a bunch of DIVs - and while I believe that canvas is the "right way" to do it, it means that it doesn't run in IE prior to version 9.
Impact also has a growing community who creates plugins for all kinds of stuff: Pathfinding, Tweening, Physics, WebSockets, etc. You can see some of that in the forums:
Wow, thanks for the details! I'm definitely going to have to grab a copy and take some time to try out Impact, it looks like a ton of fun. It's pretty clear we're going to start seeing a lot of real games coming out in HTML + JS soon.
As for canvas, the way browsers are competing on improving speed these days I don't think slow rendering is something we'll have to worry about for much longer, and with IE9 almost out it will be possible to drop support older IE versions soon enough too. Fun times to come!
Will Impact work in Android? And will it be able to play sounds and music soon? These are the only 2 reasons I haven't purchased yet, but it looks great.
However, sound is not supported on iOS or Android devices yet. The <audio> implementation on those platforms is unbelievably broken and there's not much I can do about it other than wait for Apple and Google to fix it.
This blog post does a good job at mentioning the problems with sound on iOS and how to work around a few of them. Android has basically the same problems:
Still, even with all those workarounds, you have a lot of lag (>0.5s) before a sound plays and you can only play one sound at a time. That's hardly suited for games :/
I had trouble with the multi-touch on my android, so the demo game didn't really work, but it does seem to work, yes, sorry about that.
I see what you mean about sounds. I just went and tried some of the HTML 5 games I like on my iPhone and just realized they too didn't have sounds.
I'm thinking that if I create a native Android app and wrap the HTML 5 game into a browser, then I may be able to use SoundPool to play some sounds if the game and the app can communicate.
Question: any ideas how something with that many assets and concurrent processes runs smoothly down to the iPhone?
Jquery and CSS transforms start slowing down when I stack them in each other. If I'm the problem here, and not the browsers, could someone point me to resources to make js with images run smoother in the browser? Anyone else have this problem?
Btw, awesome product with rocketpack. Wish I could've used it before Disney got it paws on it.
Rocket Engine doesn't use Canvas. I know the guys behind it and and I think their insight that canvas isn't needed for this was really the cool part of the engine.
IE9 doesn't support WebGL, but it does support 2D canvas (and it's accelerated, unlike e.g. Chrome, which only accelerates WebGL). 2D canvas is going to be around for a while.
Plus, WebGL is not easy to use. The 2D canvas is simple to get started with, and developer ergonomics is an important part of the uptake of any technology.
I was under impression that iOS apps are pretty often just HTML5 applications packed in iOS app format. Can't that mechanism be used to enable these games to be played offline? Of course it's some extra effort for the publisher, but I think iOS is big enough platform to justify some.
I was under impression that iOS apps are pretty often just HTML5 applications packed in iOS app format.
Typical native iOS apps are not like that. They are written in Objective-C against the CocoaTouch framework and are compiled down to native code.
You do get apps in the appstore that are written in html/javascript but even those will come with any extra resources they have shipped as part of the app itself.
If you're doing a purely web-based app that is meant to be installed directly from a web site, you can, as you say, instruct the browser to cache resources for offline use. However I'm not sure if there's a limit to the amount of resources that can be cached. Presumably there will be some kind of limit and games (which are heavy on graphics and sound) may easily cross that limit.
If you're aiming at a broad consumer audience, the "Login with Facebook" button is by far the lowest-friction way to authenticate users (for better or worse).
With no alternative, and nothing but a splash screen? Log in with X is certainly low-friction, but without alternatives it's still a massive barrier to many of their target users, especially when you consider this is a) Disney, and b) Facebook doesn't allow people under 13.
I suspect that the Facebook market penetration is high enough in their target demographic as to make this the most effective way for them to go. Sure, they could provide some secondary log-on technique for the non-Facebook-using minority, but would it really be worth the additional customer support headaches?
Don't know if it's 13- crowd - it's Disney, so at least partially, yes. I'm also assuming it's heavily Pokemon influenced, in which case absolutely, regardless of any stated aims.
As to customer support headaches: they're running an entire custom-built web game. Login problems would be the least of their difficulties, support or technical.
Facebook has something like 500 million active accounts, right? If we assume that the overwhelming majority of those have a 1:1 mapping with real-world people, then somewhere between 15% and 25% of Internet users have a Facebook account.
So, the lowest-friction way to authenticate users still cuts out 75-85% of people on the Internet.
Interesting! That's exactly what my startup side project is: a HTML5 game creator IDE: http://www.scirra.com/forum/viewtopic.php?f=35&t=8468
Nobody's offering me $10m, too bad :P