July
17
2009
17
When I found out Palm released the webOS SDK yesterday, I immediately downloaded it and took a look at the documentation. I'm a pretty dedicated iPhone developer but was excited about new possibilities on the Pre and its successors. As much as I enjoy developing iPhone apps, the iTunes store is horribly crowded and it's really challenging to get a toehold with new apps. The iPhone app market has turned into somewhat of a beast. Ironically, Apple has been touting the size of the iPhone app market and App Store as a major accomplishment, which it is, but there's always a downside when things get "big". I'm not sure big is good myself, at least not when Apple is the only game in town.
We can't go back to the salad days of the iTunes App Store, but the same thing is poised to happen on the webOS platform. Or so I hoped. I knew that webOS development was based on HTML, JavaScript, and CSS, but I was hoping there was a way, some way, any way, to tap into advanced hardware features and software technologies. Chief on my list is OpenGL, which is a requirement for serious games. GL even became necessary for some of my simpler apps, like Kaleido and Butterfly Collection, since basic software rendering just isn't responsive enough for smooth animations at decent speeds. You need to tap into the graphics hardware with OpenGL ES.
Sadly, my suspicions were confirmed -- there is no way for developers to tap into OpenGL ES using the webOS SDK, despite the fact that the hardware supports it. So that's a major blow. Then I took a look at the accelerometer capabilities. The accelerometer is desirable for games that use tilt control of course, but is also key to apps based on the equations of motion, like my gMeter (vehicle performance) and greenMeter (eco driving) apps.
Well, strike two -- while the webOS SDK allows access to raw accelerometer data, it's limited to a 4 Hz sampling rate (that's four samples per second). Applications like gMeter and greenMeter need 50-100 Hz to even be practical, and most games need at least 20 Hz for smooth inputs that won't lag too far behind typical graphics framerates. A low rate of 4Hz is not usable for dynamic motion where high fidelity is desired. Accelerometer support in the webOS is suitable for detecting basic movement of the phone for interface rotation, but that's about it.
In my quest to re-experience the early days of app development on a new mobile platform, I didn't rewind one year like I was hoping -- I went back two years, back to the days when we could only develop web apps on the iPhone. I wrote this piece back then, bitching about the limitations of web apps on the iPhone. It seems we're right back in the same boat with webOS.
This wouldn't be so bad for Palm if we were still in 2007, but in the age of sophisticated iPhone native apps here in 2009, web apps just don't cut it anymore. With such amazing software capabilities flourishing on the iPhone, Palm can't afford to wait a year while they make the transition from web apps to native apps in their SDK. Palm might have had a chance against the 2007 Apple SDK, but not the 2009 version. Not even close. With this limitation, webOS will not be taken seriously by consumers who place importance on games or sophisticated third party apps. The iPhone has raised their expectations too high.
That's a pity, because I find Palm's webOS to be a pretty nice operating system that is very well done for a 1.0 release. The Pre hardware is mediocre, but that can be overlooked because of the nice OS and user interface. But when it comes to the SDK, I think Palm stepped up to the plate and laid a solid bunt. Unfortunately, nobody was on base, and getting to first isn't going to win the game for Palm. They need home runs at this point. They need to swing for the fences.
From what I can see, it looks like Apple is going to hold on to the mobile app pennant for a while. The real question is, will Palm even hold on for another season?
We can't go back to the salad days of the iTunes App Store, but the same thing is poised to happen on the webOS platform. Or so I hoped. I knew that webOS development was based on HTML, JavaScript, and CSS, but I was hoping there was a way, some way, any way, to tap into advanced hardware features and software technologies. Chief on my list is OpenGL, which is a requirement for serious games. GL even became necessary for some of my simpler apps, like Kaleido and Butterfly Collection, since basic software rendering just isn't responsive enough for smooth animations at decent speeds. You need to tap into the graphics hardware with OpenGL ES.
Sadly, my suspicions were confirmed -- there is no way for developers to tap into OpenGL ES using the webOS SDK, despite the fact that the hardware supports it. So that's a major blow. Then I took a look at the accelerometer capabilities. The accelerometer is desirable for games that use tilt control of course, but is also key to apps based on the equations of motion, like my gMeter (vehicle performance) and greenMeter (eco driving) apps.
Well, strike two -- while the webOS SDK allows access to raw accelerometer data, it's limited to a 4 Hz sampling rate (that's four samples per second). Applications like gMeter and greenMeter need 50-100 Hz to even be practical, and most games need at least 20 Hz for smooth inputs that won't lag too far behind typical graphics framerates. A low rate of 4Hz is not usable for dynamic motion where high fidelity is desired. Accelerometer support in the webOS is suitable for detecting basic movement of the phone for interface rotation, but that's about it.
In my quest to re-experience the early days of app development on a new mobile platform, I didn't rewind one year like I was hoping -- I went back two years, back to the days when we could only develop web apps on the iPhone. I wrote this piece back then, bitching about the limitations of web apps on the iPhone. It seems we're right back in the same boat with webOS.
This wouldn't be so bad for Palm if we were still in 2007, but in the age of sophisticated iPhone native apps here in 2009, web apps just don't cut it anymore. With such amazing software capabilities flourishing on the iPhone, Palm can't afford to wait a year while they make the transition from web apps to native apps in their SDK. Palm might have had a chance against the 2007 Apple SDK, but not the 2009 version. Not even close. With this limitation, webOS will not be taken seriously by consumers who place importance on games or sophisticated third party apps. The iPhone has raised their expectations too high.
That's a pity, because I find Palm's webOS to be a pretty nice operating system that is very well done for a 1.0 release. The Pre hardware is mediocre, but that can be overlooked because of the nice OS and user interface. But when it comes to the SDK, I think Palm stepped up to the plate and laid a solid bunt. Unfortunately, nobody was on base, and getting to first isn't going to win the game for Palm. They need home runs at this point. They need to swing for the fences.
From what I can see, it looks like Apple is going to hold on to the mobile app pennant for a while. The real question is, will Palm even hold on for another season?