The Apple Developer Backlash – notes from the field #dropple

?Kirk:

What originally attracted me, like many other technologists, to Apple after Mac OS X was released, was simple: fantastic hardware engineering, a UNIX operating system, with a fantastic user interface. Having been frustrated with the proprietary nature of Windows, and the comparatively poor user interface and application support of Linux distributions, the Mac was an excellent choice for development. Having access to top-tier productivity applications like Office and iWork made the choice trivial: OS X meant I could have my developer cake and eat my enterprise cake at the same time.

Apple as well actively courted the advanced developer market. XCode and Cocoa/Carbon were excellent ways to build platform-specific applications (and far better than Visual Studio and the Microsoft UI libraries). Java support was there for running Eclipse and other Java-based IDEs. At last there was a good platform for those developing applications that would eventually run on Linux and Unix based servers.

However, Lion, the App Store, moves to remove Java from being a viable option for anything on a Mac (although today’s announcement about Apple working with Oracle on OpenJDK may mean, assuming OpenJDK is a reasonable option compared to the J2SE distribution, that this improves at some unspecified point in the future), and the general stance of Apple at this point mean that, as sad as it is to say, my next computer (the first time in 6 years) will likely not be a Mac.

At the end of the day, when I want a computer, I want a computer. I don’t want a phone or tablet with a keyboard. I don’t want a curated application environment. I want a messy, creative, productive ecosystem sitting on top of a rock-solid kernel. I want Snow Leopard, not Lion.

I used to avoid Windows out of an instinctive recoiling at the very idea of supporting Ballmer, the completely ridiculous mess that is XP, and the market disconnect that resulted in the unusable Vista. Windows 7? It’s now good enough. And the fact that it’s now good enough, and Apple are ruining any last vestige of developer/power-user credibility, means it’s now an option for me and most advanced techies I know. And that should worry Jobs far more than Android does.

Ana:

When I was at OpenGamma’s offices recently, Kirk was saying he probably won’t be buying
another Mac. I was agreeing. The backlash is gaining momentum.

The Lion OS is too scary, on top of what has happened to Flash and Java. It’s clear that Apple’s dream is that nobody can install anything on a Mac without Apple getting their cut and having their say. This is not how developers operate, especially open source developers used to autonomy and flexibility. Being told to choose your tools from a pre-approved list is about as evil as it gets to us, and there’s been no credible reassurance that this won’t be the case soon.

I think Apple is counting on its consumer market share to retain developers. The prospect of being an App Store lottery winner will be enough of an incentive to keep developers building apps for OSX and the iPhone, and probably even paying for the privilege.

Well, maybe they’re right. Some software companies probably will be willing to do that. But those individual developers who recognized OSX for what it was in the early days and jumped on board writing incredible apps the likes of which no Windows user could dream of because they ‘got it’, those developers who rushed to sign up to write iPhone apps just because they could imagine what you could do with an accelerometer, those developers who built apps that no committee could invent, in short, those developers who made Apple’s vision actually work. The ones who provided the momentum, the excitement, the first cool toys. The ones who prevented OSX from being the next NeXT. These developers won’t stay. As individuals, they’ll switch platform. As employees, they’ll switch jobs. Creative and passionate people creating awesome stuff on the leading edge. Walled developer garden. Choose any 1.

I’m particularly angry on behalf of the academic community who bought into the OSX vision in huge numbers. Remember the supercomputer built on top of the (now discontinued) XServe server? I watched Apple’s market share at one particular academic conference go from about 20% to 60% at 3 successive conferences (a 2-year period). A huge amount of academic software is written in Java. Indeed, this fact made it possible for many academics to switch to OSX in the first place. Now the future of these investments in hardware and software and sweat equity is uncertain. At least we can move our number crunching into the cloud, maybe with a nice iPad interface to keep an eye on things and view the output, but this golden age of XServes running XGrid seems to have been all too short.

I’ve been doing a lot of my programming work within Virtual Machines these days, even on my Snow Leopard machine with its Darwin underpinnings. It’s a great way to work, I can test software on multiple linux flavours and use whichever is most convenient for a particular project. I can take snapshots and revert my machine to previous states. Working with non-Darwin OSs has reminded me how frustrating it can be to build software on OSX with all its missing header files and non-standard libraries. So, with the popularity and ease of use behind virtualization technology, these days there’s really no need to have an OS which has *NIX underpinnings. So, Windows 7, which I must agree looks rather good, is actually a viable option these days. “Lesser of two evils” has never been so apropos.?Kirk:

What originally attracted me, like many other technologists, to Apple after Mac OS X was released, was simple: fantastic hardware engineering, a UNIX operating system, with a fantastic user interface. Having been frustrated with the proprietary nature of Windows, and the comparatively poor user interface and application support of Linux distributions, the Mac was an excellent choice for development. Having access to top-tier productivity applications like Office and iWork made the choice trivial: OS X meant I could have my developer cake and eat my enterprise cake at the same time.

Apple as well actively courted the advanced developer market. XCode and Cocoa/Carbon were excellent ways to build platform-specific applications (and far better than Visual Studio and the Microsoft UI libraries). Java support was there for running Eclipse and other Java-based IDEs. At last there was a good platform for those developing applications that would eventually run on Linux and Unix based servers.

However, Lion, the App Store, moves to remove Java from being a viable option for anything on a Mac (although today’s announcement about Apple working with Oracle on OpenJDK may mean, assuming OpenJDK is a reasonable option compared to the J2SE distribution, that this improves at some unspecified point in the future), and the general stance of Apple at this point mean that, as sad as it is to say, my next computer (the first time in 6 years) will likely not be a Mac.

At the end of the day, when I want a computer, I want a computer. I don’t want a phone or tablet with a keyboard. I don’t want a curated application environment. I want a messy, creative, productive ecosystem sitting on top of a rock-solid kernel. I want Snow Leopard, not Lion.

I used to avoid Windows out of an instinctive recoiling at the very idea of supporting Ballmer, the completely ridiculous mess that is XP, and the market disconnect that resulted in the unusable Vista. Windows 7? It’s now good enough. And the fact that it’s now good enough, and Apple are ruining any last vestige of developer/power-user credibility, means it’s now an option for me and most advanced techies I know. And that should worry Jobs far more than Android does.

Ana:

When I was at OpenGamma’s offices recently, Kirk was saying he probably won’t be buying
another Mac. I was agreeing. The backlash is gaining momentum.

The Lion OS is too scary, on top of what has happened to Flash and Java. It’s clear that Apple’s dream is that nobody can install anything on a Mac without Apple getting their cut and having their say. This is not how developers operate, especially open source developers used to autonomy and flexibility. Being told to choose your tools from a pre-approved list is about as evil as it gets to us, and there’s been no credible reassurance that this won’t be the case soon.

I think Apple is counting on its consumer market share to retain developers. The prospect of being an App Store lottery winner will be enough of an incentive to keep developers building apps for OSX and the iPhone, and probably even paying for the privilege.

Well, maybe they’re right. Some software companies probably will be willing to do that. But those individual developers who recognized OSX for what it was in the early days and jumped on board writing incredible apps the likes of which no Windows user could dream of because they ‘got it’, those developers who rushed to sign up to write iPhone apps just because they could imagine what you could do with an accelerometer, those developers who built apps that no committee could invent, in short, those developers who made Apple’s vision actually work. The ones who provided the momentum, the excitement, the first cool toys. The ones who prevented OSX from being the next NeXT. These developers won’t stay. As individuals, they’ll switch platform. As employees, they’ll switch jobs. Creative and passionate people creating awesome stuff on the leading edge. Walled developer garden. Choose any 1.

I’m particularly angry on behalf of the academic community who bought into the OSX vision in huge numbers. Remember the supercomputer built on top of the (now discontinued) XServe server? I watched Apple’s market share at one particular academic conference go from about 20% to 60% at 3 successive conferences (a 2-year period). A huge amount of academic software is written in Java. Indeed, this fact made it possible for many academics to switch to OSX in the first place. Now the future of these investments in hardware and software and sweat equity is uncertain. At least we can move our number crunching into the cloud, maybe with a nice iPad interface to keep an eye on things and view the output, but this golden age of XServes running XGrid seems to have been all too short.

I’ve been doing a lot of my programming work within Virtual Machines these days, even on my Snow Leopard machine with its Darwin underpinnings. It’s a great way to work, I can test software on multiple linux flavours and use whichever is most convenient for a particular project. I can take snapshots and revert my machine to previous states. Working with non-Darwin OSs has reminded me how frustrating it can be to build software on OSX with all its missing header files and non-standard libraries. So, with the popularity and ease of use behind virtualization technology, these days there’s really no need to have an OS which has *NIX underpinnings. So, Windows 7, which I must agree looks rather good, is actually a viable option these days. “Lesser of two evils” has never been so apropos.