By John Gruber
WorkOS: APIs to ship SSO, SCIM, FGA, and User Management in minutes. Check out their launch week.
Tom Yager writes on Apple’s decision to at least temporarily close the source code to the x86 version of the Darwin kernel. I think he takes things much too far with his conclusions, though:
Apple’s retreat to a proprietary kernel means that all users must accept a fixed level of performance. The default OS X kernels are built for broad compatibility rather than breakneck speed and throughput. That doesn’t matter at present, because all Intel Macs are built on the same Core Duo/Core Solo 32-bit architecture. But Apple’s workstation and server will be built using next-generation 64-bit x86 CPUs. […] Macs will inherit the benefits of Core Microarchitecture’s evolution, but OS X is limited in the degree to which it can exploit specific new features without creating branch after branch of OS code to handle each tweak to the architecture.
The insinuation here is that if the source code to Darwin’s kernel were available, users of high-end Power Macs (or whatever they’re going to call them) and Xserves would be able to run their own hot-rodded kernels and yet still run Mac OS X. I suppose technically that might be possible, it certainly doesn’t sound plausible. I highly doubt one could just plop down a recompiled kernel and expect the rest of Mac OS X to continue “just working”.
The truth is that other than serving as a nice symbolic gesture, I’m not sure there’s been any practical benefit to the fact that the Darwin kernel has been available as open source. In other words, I think it’s hard to make a case that most Mac users should care whether the Darwin kernel is open source or not.
★ Wednesday, 17 May 2006