The various Berkeley Software Distribution operating systems, having been nearly destroyed in an ugly lawsuit with AT&T, have had a challenging past. FreeBSD has been the only non-commercial Berkeley unix to have had any measure of real success and, to this day, the hangover of its early legal woes still stymies its popular use outside of industry despite improvements to its codebase.
The infamous AT&T/BSD lawsuits caused FreeBSD to jettison two-thirds of its codebase and start over from scratch, knocking its feature-set back several years, causing it to be bought out and divested by hopeful investors and only reaching robustness with code infusions from two commercial unix systems, BSD/OS and Mac OS X.
So as FreeBSD heads toward its milestone tenth release, it is Apple's platform that has been the center of gravity for the demonic operating system, largely in part by Apple's open-source efforts that have been more and more contributing code back into the main FreeBSD development branch. It obviously won't be long until the two are completely merged—so when thinking of 2014 and FreeBSD 10, starting thinking “FreeBSD X.”
But first, how did we get here? What did BSD/OS contribute or, more importantly, how did it take away from FreeBSD? And what exactly does FreeBSD 10 look like? What is Apple's role in the operating system? Important questions to be answered, picking up from after the lawsuits…
In 2001, just as Apple was releasing its first consumer-oriented version of Mac OS X, WinDriver Systems bought FreeBSD. WinDriver had focused, until this point, exclusively on writing third-party drivers for high-end video boards that ran under Windows, but was making a concerted effort to capture the embedded systems market. News of their FreeBSD purchase was much-maligned at the time by pundits, but FreeBSD developers supported the move since it would allow them access to BSD/OS source code.
BSD/OS was the oldest 4.4BSD derivative, preceding FreeBSD and even NetBSD, a little-known fork of FreeBSD focused writing drivers for obsolete hardware. Because of its vintage and pure 32-bit C codebase, BSD/OS source was considered to be pristine, efficient, and secure. With FreeBSD and BSD/OS under the same roof, code convergence started in earnest.
Though WinDriver announced sales of BSD/OS 4.3 alongside the pre-existing FreeBSD 4.x series in 2002, their other announcement about FreeBSD 5 and BSD/OS 5 was a much bigger deal. The new versions of each operating system would essentially be the same product, representing the merger of the two operating systems, but for BSD/OS offering value-added (re: proprietary) software to the distribution package.
Unfortunately, being so new to embedded systems, WinDriver's development coordination and marketing were disastrous. FreeBSD and BSD/OS development dragged and missed deadlines, with FreeBSD developers complaining publicly about the “hellish fascism” of WinDriver's development model. After many months of uncertain development hell, WinDriver cancelled BSD/OS and spun FreeBSD off into its own company, FreeBSD Mall, Inc. After two torturous years, and not without some assistance by Apple, FreeBSD 5 was finally released fully integrating BSD/OS.
Parallel to the WinDriver wreck, Apple had been working in earnest to mate its newly purchased OPENSTEP operating system with the Mac operating system in what it called Rhapsody and, later, Mac OS X. Core to OPENSTEP was a full 4.4BSD implementation running on the Mach kernel, which Apple upgraded. Mac OS X Server 1.0 included a full FreeBSD 3.0 implementation; Mac OS X v10.0 included FreeBSD 4.2. Apple continued this pattern for years, calling the FreeBSD-on-Mach portion of Mac OS X Darwin and released it as its own BSD operating system.
Over time, Darwin began to lead the way with Apple code migrating back into FreeBSD and contributed significantly to the FreeBSD 5 release. Some developers even went so far as to thank Apple for FreeBSD, insisting that, without Apple's help, FreeBSD 5 would have never left WinDriver's development hell.
Fast-forward several years later and, with the release of FreeBSD 7.2 (for use in Mac OS X Snow Leopard), the significance of Apple's contributions to FreeBSD was so great that developers began saying that FreeBSD owed Apple big and called on Apple to merge Darwin and FreeBSD.
Since then, that's just about happened. FreeBSD 9 was released about nine months after Mac OS X Lion, integrating updates from the underlying Darwin 11. Going further into the release timeline, Apple will be release OS X 10 at the same time as FreeBSD 10, the two operating systems having finally come into true parity and FreeBSD X focused on high-end enterprise serving and point-of-sales. Apple will finally move its website and Apple Store from a hodgepodge of Linux and Solaris to FreeBSD X, and can finally say that it's eating all its own dog food.
It's also likely that, with the FreeBSD and Darwin codebases having been merged, FreeBSD will gain the Open Group's UNIX® certification that Darwin—and therefore iOS and OS X—already has. Wrap your head around that one: it only took about twenty years for FreeBSD to become a real Unix operating system.
So what will an Apple FreeBSD release look like? Good question. While predicting two years ahead in the computing industry is a dicey undertaking at best, certain things can be counted on. For instance, FreeBSD's lackluster, anemic installer will go straight to the open-source trash heap. Expect a fully graphical installer using Quartz. Likewise bet on the full OS X Aqua interface, making FreeBSD X like Mac OS X Server before it was discontinued after Snow Leopard.
|The FreeBSD X installer app.
In the end, FreeBSD X will be available as a 2 GB download from the Apple App Store via an installer app, just like Mac OS X Lion and OS X Mountain Lion. It will drop support for all microarchitectures except 64-bit Intel (x64) and 32-bit ARM (A32) and Apple will run a development program for it exactly like like those for iOS and OS X, with FreeBSD X development licenses running $99 a year.
From there, FreeBSD will become one realization of the most widely shipped unix or unix-like operating system in the world, with an installed Mac user base of some 120 million today and a predicted 300 million worldwide by 2014. This will almost certainly cause a massive sea-change in how open-source operating systems are viewed by the general public, and the iOS/OS X/FreeBSD X trifecta will work to create the most pervasive unix use ever. GNU/Linux should get ready to kiss its ass goodbye.
So, the last chapter of FreeBSD itself, from its early beginnings as a limping, feature-anemic unix that just barely survived a massive lawsuit to it receiving a code transfusion from the more professional BSD/OS will finally be written, and a new story about an Apple-infused modern juggernaut of long, hard uptimes will be told. Thanks to Berkeley unix and Apple quality, the future of FreeBSD X is only just beginning.