My thoughts on Flash CS5 and iPhone
I’ve been holding my tongue on this since the announcement, and slowly putting my thoughts together in what I hope is not considered to be a knee jerk reaction.
First of all, a big dose of respect to the engineering team that made this happen. You guys are God-like. I think it’s an amazing feat of technology that you got this working to the point where apps developed in Flash not only run on an iPhone, but were submitted to and accepted by the app store. So anything negative I say is not in any way directed to you guys. Unfortunately, technical awesomeness does not always directly correlate with practical usefulness. God knows I’ve created some things that I considered pretty cool feats of technology, but that were absolutely useless in the long run.
My first reaction on hearing the announcement was that people were going to be awfully disappointed. Yes, it’s an amazing announcement and everyone was very excited about it. But remember when you were a kid and you saw all those commercials on TV for such and such a toy, and you waited and waited for your birthday or Christmas or whatever, tingling with excitement, and then you opened it up and started playing with it and realized that after all, it was just a cheap piece of plastic, with only a fraction of the awesomeness you envisioned? That kind of disappointment. I expected that for one, that disappointment would be in the APIs. The iPhone has a rich set of APIs and Flash has its own rich set. I expected a small subset of both. To my surprise, Adobe has actually done a really good job of getting most of what’s in Flash functioning on the iPhone, and most of what’s natively available on the iPhone accessible from Flash. I haven’t used them yet, but from the way they were described, they seem surprisingly complete. So again, more respect on that end.
However, I still think people are going to experience that Christmas morning disappointment when they start working with this, based on performance. Take a look at the resources Adobe has listed on Labs and the recorded MAX session. And check out the apps that have been released. Almost all of them have comments complaining about performance – slow, choppy graphics and animations, unresponsive UIs, etc. From my own observations, those of others, and even from those who created them, the consensus is that they are very choppy on an iPhone 3G, somewhat better on an iPod Touch 2nd Gen, and “acceptable” on an iPhone 3GS or the latest iPod Touch.
Personally, out of the several apps I have tested on my iPhone 3G, none of them are at the level where I would have released them. Now this isn’t necessarily a bad thing in itself. There are other high performance 3D racing and sports games that take advantage of the advanced features of OpenGL ES 2.0 and therefore require a 3GS. But the things that are choppy and not performing well are the simplest of simple animations and transitions. These are not the kinds of apps and games that should require lots of processing power. If you’re going to specify that your game requires the highest end hardware to function adequately, that carries with it the concept that it is using that hardware’s extra power to deliver an over the top experience. To demand the top end processor just to make simple animations not look like crap is not acceptable.
Now, of course, this is a very early beta or even alpha of a brand new, budding technology. I have NO doubt that it is going to improve. I do have doubts on how much it is going to improve. I would assume or at least hope that it would get to the point where the games such as have been released are at least acceptable on an iPhone 3G.
But my initial point was that people are going to be disappointed. Yes, if Adobe pulls it off, you’ll be able to do some basic animations that perform relatively well, but as soon as you start doing anything remotely complex, you are going to be back in a world of choppiness. That’s my prediction, and I’m willing to eat crow on it if Adobe proves me wrong. But I can imagine people looking forward to converting their PV3D-based stuff to the iPhone, or something with multiple, parallax backgrounds scrolling around and dozens of enemy characters flocking here and there. These are things that perform fine in Flash natively, and things which, if you are using the correct tools, are not a problem at all to accomplish on the iPhone. But I am highly skeptical that the Flash to iPhone route is ever going pull off things like this with any kind of decent performance.
So a lot of people have said stuff like, “Well, yes, I understand this is just for simple, basic games. I’m OK with that.” Well, that’s cool. I’m not going to rain on your parade. If you have a realistic expectation of what you can do, and that’s all you want to do, and you are happy with that, then be happy. Particularly, if you have been developing Flash applications for mobile, you probably have a good idea of what you can do there, and as long as you keep those expectations about the same for the iPhone, you should be pretty happy. However, I will say that a trivial application or game like that is the ideal kind of project to learn Objective-C with. Why not make the leap? You’ll learn something new and useful and when you want to make something more complex, you’ll know how and won’t be arbitrarily stopped.
There’s one thing that really concerned me after the announcement. It was the number of Flash developers in the community saying, “We are all now iPhone developers! Yay!” and “Whew! I don’t need to learn Objective-C now. I’m throwing away all my Objective-C books!” I still believe the bottom line is, if you are serious about developing for the iPhone, you must learn Objective-C. For God’s sake, it’s NOT THAT HARD! Be a professional. Learn a new language. If you are even moderately skilled with AS3, you will be able to pick up Objective-C inside a week. Yeah, the memory management stuff will probably continue to challenge you for a while – I’m still coming to terms with it – but in general, you can be up and running in Cocoa in no time.
OK, so you are a Windows person and don’t have a Mac. So what? Get a Mac. If that’s your sole reason you are holding yourself back from doing iPhone development, and you think that Adobe has solved that for you, I’m sorry, but you are delusional. I wanted to play with Silverlight a bit. What did I do? I bought a PC. Hell, if you are developing for Flash, you are developing for both platforms and you really should have access to both platforms anyway.
Moving past performance issues, there is a huge amount of control you lose in the Flash to iPhone flow. Yes, memory management is a pain in Objective-C, but it has to be done. How does the Flash compiler do it? No idea. Hopefully really well, because that’s what you are stuck with. And it’s not just memory management. Every aspect of an application gets filtered through the conversion process and the Adobe API. There are only a handful of developers using it on just a handful of games so far, but wait until it gets into the real world. I can imagine major pain. As an example of what I mean by this, and one I made on Twitter, check out any of those apps or services that make Flash web sites for you – you know, where you drag and drop some media, and fill out some forms in some wizards and it spits out a Flash web site. There are plenty of them. If what you want is exactly what it spits out, or the changes you want to make are part of one of the wizard forms, you are all set. But the moment you want to have ANY further control over it, you are out of luck. Granted, you will have more control in Flash to iPhone, but I have no doubt there are going to be major points where you wish you had more control. At any rate, if someone came to you claiming to be a “Flash Developer” and showed you a site made with one of those templates, you would either be pissed off, or just laugh at them. And I think this is essentially why some members of the established iPhone development community are pissed off or laughing at this announcement.
Then there is the Apple reaction. As of yet, completely undefined. There has been speculation about the legality of what Adobe is doing, whether it breaks the developer agreement, whether Adobe reversed engineered anything, whether they are using non-public APIs, etc. Of course, Adobe’s stance is that it’s not a problem. But it’s really up to Apple to adjudicate what a problem is. I’m not even saying that Adobe has done anything illegal or even unethical, but I expect that if Apple can find ANY reason whatsoever to stop it, legally, technically, or otherwise, they will. Yes, several apps got approved under Apple’s nose. I’m interested to see how many will get approved now that Apple know’s what’s what. Some say that they won’t care – that they still get the revenue from the sales, so they will be happy to have more apps coming in. I don’t know. Time will tell. Especially if this does result in a flood of low quality apps, as many are predicting. They have many options. They can just disapprove any app that is made by Flash – maybe let the odd one through here or there to appear fair (and don’t think they can’t tell which apps are made with Flash and which aren’t). Or they could attempt outright legal action against Adobe. Or they could just change some API or app signing procedure or whatever, causing Flash based apps to not be submit-able, and forcing Adobe into a constant game of catch up. Or any number of other responses. I don’t think they will simply do nothing.
Finally, this is just my personal speculation, but I somewhat feel that this is more of a political move against Apple, to somehow force some kind of action on getting the Flash Player onto the iPhone. I don’t know how that is supposed to work, but progress on that front seems to be at a standstill now, and this is no doubt going to stir things up at least. So as technically amazing as this is, I predict that this technology is not a long term strategy for Adobe and I doubt it will be part of CS6.
