ewinemiller said:
This is so totally wrong. The only reason OS/2 lasted as long as it did was because it could run Windows. Nobody was buying OS/2 for OS/2 they bought it because it was a better Windows. If it didn't have that going for it, it's market share would have been even worse.
I have to chime in here. Not only was I an avid OS/2 user/fanatic, I worked at IBM as an intern during the OS/2 2.0 release.
People were buying OS/2 for OS/2 in the business world, in the 1.x days. Windows 3.x support was added in late in the development cycle, once they saw Windows 3.0 taking off. Much like the case the Mac has been through with the 68k->PPC, OSX, and Intel transitions, it gave the DOS/Windows compatability as an option but it was always preferable to have a native app. Yes, it added value by offering true multitasking and virtual machine protection. But the best way to go was to get a native application.
Where Apple is caught in the same position as OS/2 is this: By making your operating system a superset of another (what IBM tried with OS/2 vis-a-vie DOS/Windows), then yes you do get the benefit of running the other guy's applications PLUS your own applications. However, you get caught in two traps:
(1) You are not in control of the platform you are encapsulating. IBM did not control what Microsoft did with Windows. So MS came out with Windows 3.1 and their applications started "requiring" features in it, breaking OS/2 compatibility; and then Windows 95, which OS/2 could not handle and fizzled out. The problem is the platform you're trying to be compatible with is always moving. You end up reacting to the other guy's changes with significan lag time.
(2) Application vendors will not write native applications. In the case of OS/2, why would software vendor X bother to write an OS/2 native version of their application when the Windows version works on OS/2? The developers look at it like this: I can write one version that services 95% of the market and can run under a virtual machine on the remaining 5%; or I can write TWO versions with the second hitting a very very small portion of the market. What do you think they're going to do?
IMO Apple IS caught in this trap w.r.t. Windows. Let's face it, the fact that applications are missing on the Mac is why those who want to dual boot are interested in it. Even Microsoft doesn't port all of its large applications to the Mac. Many vendors don't.
The idea of having an OS is to get people to write applications for it. This is one of the fronts on which the Mac has lost, it's always going to come 2nd to the larger Windows market. You said so yourself.
But I think Apple does have a trump card that gets them out of these traps. Yellow box. Get developers to write "Mac applications" that also happen to work on Windows. Get 100% of the market with one application build. It also capitalizes on MS's relative complacence in the compiler market as of late, if Apple can offer superior development tools.