|< Free Open Study >|| |
If you go back far enough, you find that the history of Linux is closely tied to the history of Unix, and even to the history of computing. The history of Unix is an interesting one, as Unix has had a subtle yet mind-bogglingly large impact on computing in general. All roads lead to Unix, so to speak. Around 1971, AT&T's Bell Labs group developed the first incarnation of Unix (which was originally known as Unics or Uniplexed Information and Computing System) Unix' name was a pun on Multics, a more complex forerunner of Unix. For a while, Unix had very limited exposure. It was used primarily within AT&T, for internal purposes.
Eventually, it was released as a product to the world at large, and started to build momentum. AT&T's Unix reached its pinnacle in the version dubbed System V (as in "System Five", and sometimes abbreviated as SysV). Perhaps inevitably, though, offshoots and derivatives started appearing.
In 1978, the University of California at Berkeley released an operating system known as the Berkeley Systems Distribution, or BSD, which was very much like Unix. The reasons for this divergence varied. There were issues with the license for the source code that AT&T required for Unix; some people argued that AT&T was not being proactive enough in enhancing and improving key aspects of System V Unix. But whatever the reasons, BSD became tremendously popular, and became the basis for many other offshoots.
When the dust settled (if it can truly be said to have settled, even today), most of the offshoot Unix "flavors" provided by various vendors were derived from either SysV Unix, BSD, or a hybrid of both. Additionally, a specification known as POSIX was created by the IEEE (Institute of Electrical and Electronics Engineers, Inc.) to attempt to codify the common features of these various offshoots into a single standard.
Today, there are a great many flavors of Unix competing with one another. Each major hardware vendor generally has it's own flavor, and there are several free flavors as well. Among commercial vendors, IBM has AIX, Hewlett-Packard has HP-UX, Compaq has Tru64 Unix, SGI has Irix, Sun Microsystems has Solaris, and this list is not even comprehensive. Even Microsoft had a flavor, called Xenix, and there are tools to provide a Unix-like environment on top of Microsoft's Windows operating systems.
The free versions of Unix include several offshoots of BSD (FreeBSD, NetBSD, and OpenBSD) and, of course, Linux. Linux and the free BSD systems are similar in origin and intent, but vary widely in focus and ideology. Linux systems appear to have the greater industry momentum, but otherwise they are fairly similar (until you get under the hood, at least). Just to add even more confusion, there is a commercial cousin of the free BSD operating systems called BSDi. You can make a few interesting observations about these offshoots. Perhaps most noteworthy is that the commercial vendors' flavors generally run exclusively on the vendors' own hardware architectures. (Even Sun, which had an Intel version of their Solaris flavor, has not had great success with it, by most accounts.) The free flavors, in contrast, have each been ported to several different platforms. Linux, for example, has been ported to the Intel, Alpha, PowerPC, SPARC, MIPS, ARM, Motorola m68k, IBM S/390 architectures, and even other more exotic platforms. NetBSD, meanwhile, runs on even more architectures than Linux!
Many pundits have described this colorful bag of flavors as the fragmentation of Unix. It is hard to decide which of these "fragments" is the best—which operating system is best for a given application is pretty subjective and variable. Some platforms win at database applications, others win at web server performance, and still others win at numerical scientific computing.
However, the free operating systems are really starting to shake things up. Initially dismissed as irrelevant, or cited by the pundits as simply further examples of the fragmentation of Unix, these operating systems (and Linux in particular) have actually shown signs of reversing the fragmentation of Unix. That is, almost every major vendor of Unix flavors has begun to embrace Linux to a greater or lesser degree. At the time of this writing, the only holdout appears to be Sun Microsystems, which is perhaps not surprising, since Sun's Solaris is arguably the most successful commercial Unix flavor available.
Of course, it's by no means a foregone conclusion that Linux will "defragment" the Unix landscape. However, the rate at which traditional Unix vendors are adopting Linux strategies is striking. Curious readers may justifiably ask why. Well, the primary reason is, of course, the same reason this book exists: because Linux represents quality, flexible, powerful, and free software, and businesses are just as interested in these things as are readers of this book. However, it's also possible that the Unix vendors have begun to realize that developing and maintaining a Unix-like operating system is a very expensive proposition, and is usually not profitable. (Most hardware vendors make their profits on the systems and services they sell, not the operating systems they sell with them.) Rather than compete with each other in a sort of computing arms race, many businesses are choosing to reduce their development costs by adopting Linux, and focusing their competition where it matters: on the hardware. In the end, though, it's hard to generalize the market strategies of large corporations as we just have, so you can find as many opinions on the matter as there are pundits.
The term "Unix" is actually a trademarked name. Only the original, AT&T version of Unix can properly be called "Unix"; the offshoots (including BSD) are "Unix-like." The actual owner of the "Unix" trademark is the Open Group (at http://www.opengroup.org). Among other things, they attempt to promote and certify compliance of standard Unix systems. However, the most direct descendant of the original AT&T Unix is probably Caldera's UnixWare product.
|< Free Open Study >|| |