Seems like we are basically in agreement. Though I would like to posit the following. At one time BSD contained AT&T UNIX code and was therefore UNIX. At the time it was viewed as a “branch” of UNIX. People spoke of AT&T UNIX and BSD UNIX. As time moved forward, the BSD folks replaced the original AT&T code. Why does that suddenly make it not BSD UNIX?

At the time that HPUX, SunOS, Xenix, etc were created, they were not UNIX. For the simple reason that they were not allowed to use the name. However, everyone knew that they were in fact UNIX.

I think the broader issue is that this article attempts to give a simple answer to a fairly complicated issue, and makes a huge misstep in assuming that Linux is derived from the UNIX code base.

Once cannot begin to answer the question, “what is the different between UNIX and Linux” without discussing quite a bit of history. What did UNIX originally mean? What do people generally mean when they use the name “unix” now? Etc…

All true.. I agree with you.
but… the BSD code that got incorporated into SYSV and during ‘unix wars’ of the ‘80 today is NOT part of free *BSD systems. The original vi written by Bill Joy is in my HP-UX system but on my other FreeBSD system I got only nvi written by Keith Bostic. Some of the original BSD code written in late seventies and eighties is now part of the AT&T UNIX but this code is not present in todays *BSD systems like FreeBSD, NetBSD or OpenBSD. Why? Because it was so heavily dependent upon original AT&T UNIX code that it had to be rewritten from scratch to allow for free redistribution under BSD license. It is all deeply discussed by McKusick in his essays on BSD history.

So once again, STRICTLY speaking *BSD is not UNIX because it does not contain any proprietary AT&T UNIX code, but at the same time *BSDs and Linux
are Unix in the sense they are built according to Unix principles.

I am the last person not to give FreeBSD due credit (I’m contributing code to the Project) but truth must be told regardless. I consider Unix as a family of operating systems built according to the ideas first devloped by Thompson and Ritchie.

Last thing - the development of *BSD systems is not really different from the development of Linux - both projects started as an effort of amateur programmers, not professionals like CSRG or AT&T Bell Labs. There is no real continuity of BSD development from the ‘80 to the era of Internet and free *BSDs. BSDs now are being developed by mostly amateur programmers from around the world, same as Linux…
You look for continuity of code and talent? look for OpenVMS (now owned by HP)
developed by DEC in late seventies and still going strong, developed mostly by the same people… I also think that HP-UX, AIX and Solaris are true descendants of the original AT&T UNIX codebase.

Thanks for reading it :)

Being that the UNIX code base (including SYSV) includes lots of code ported from the BSD code base, which AT&T calls UNIX, we can only conclude that the same code being in BSD makes it UNIX. Your view that the only “real” UNIX systems are HP-UX, AIX and Solaris isn’t accurate and isn’t even shared by The Open Group, the owners of the UNIX name.

BSD and AT&T UNIX were cross pollinated for a long time. BSD eventually shed all AT&T code, but to this day, all major UNIX variants still include BSD UNIX code. To say that BSD is not UNIX is to deny history and to fail to give credit to the hard work done on UNIX at Berkeley.

**MS Windows would be a bad example since MS intentionally diddled with it to cause problems with people running MS-Windows on DR-DOS.

Doug, don’t worry. It’s not all that important what your metaphor is. I could as easy talk about the difference between cats and dogs: Dogs have owners; Cats have staff.

If you’re interested in what I consider the “Full Story”, Dr. Peter Salus wrote _The Daemon, the GNU & the Penguin_ which is available online at Groklaw:

The Daemon, the GNU and the Penguin

(just in case linking doesn’t work, Groklaw.net is not too hard to search)

Highly recommended reading.

In the metaphor department, I like Neil Stephenson’s. They’re all cars. Microsoft is the big, expensive, full-service dealership which sells the cars with the hoods welded shut. Apple has another dealership across the road, much the same but more expensive and very pretty in comparison to the somewhat utilitarian MS vehicles.

Across the intersection are the upstarts, who are giving away their vehicles for free. They’re tanks. Not lumbering awful stinky things, but modern M1-Abrams tanks with atomic engines that sip fuel, ready with the keys in them for anyone to take. Variations in colour, style, size, all there for the taking or design your own. (this is linux, bsd, etc)

Yet, regardless of how much hollering, shouting and hand-waving the people at the field of uber-tanks do, other people still go into the expensive dealerships and spend hard-earned money for less. Much, much less.

http://www.cryptonomicon.com/beginning.html

You failed to do your UNIX history homework. There is no AT&T SYSV code in *BSD. All those parts were rewritten - read some history on the BSD. So todays derivatives of Jolitzes’ 386/BSD are no more UNIX than Linux is. Both of them are Unix-like and not Unix, strictly speaking. But whereas Linux kernel was written from scratch by person who wasn’t coming from the ‘hacker’ community of the US universities, BSD was written/rewritten by people who got access to the original AT&T SYSV codebase.

The only still relevant today real UNIX systems are HP-UX, AIX and Solaris (they all still contain the original SYSV code).

“All version of Unix are exactly the same, except for the parts that are different.”

The following passage is located at: http://e2fsprogs.sourceforge.net/ext2intro.html

“In its very early days, Linux was cross-developed under the Minix operating system. It was easier to share disks between the two systems than to design a new filesystem, so Linus Torvalds decided to implement support for the Minix filesystem in Linux. The Minix filesystem was an efficient and relatively bug-free piece of software.
However, the restrictions in the design of the Minix filesystem were too limiting, so people started thinking and working on the implementation of new filesystems in Linux.”

As you know, cross-development is used regularly.

I write this to add something to the mix, not to claim any great expertise. No, not at all like Ken Brown who wrote the book, “Samzidat” to tell us all about how Linux Torvalds could not have possibly created Linux. The book was apparently printed but never distributed. I have written to him at the Alexis de Tocqueville Institute http://www.adti.net and asked him to put it on the Internet for all to read. I have neither heard nor seen anything except an excerpt I was able to locate many moons ago. If he is as proud of his work as he claimed, I would really appreciate him making it available to all. I suppose he libeled too many people to take the chance.

Its all good, except for the SCO i really hate that guy.

UNIX® is a registered trademark owned and enforced. For something to be UNIX® it has to pass a barrage of compatibility tests. For instance, Mac OS X 10.5 is UNIX®.

The Unix community uses Unix (not all-caps) to refer to the family of operating systems descended from v7 Unix at Bell Labs. That includes the commercial Unixes descended through AT&T System V (SYSV), and the mostly non-commercial Unixes descended through BSD.

What complicates the picture is the question of what it really means for something to be based on something else. Saying it needs to have code from it doesn’t really work, as there’s very little v7 code left in OS X.

In the case of Unix, the best guideline for whether something is Unix or not is POSIX compliance. Both SYSV and BSD are moving towards the same POSIX standards for everything. Since Linux and Minix are too, and since they’re both modeled on the core Unix design elements from v7, it makes sense to say that they are Unix, even if they aren’t UNIX®–just like FreeBSD is Unix, even though it isn’t UNIX®.

Like a lot of people, you seem to think that commercial UNIX = all UNIX. This just isn’t the case. Solaris, HPUX, AIX, etc are all commercial UNIX variants. They cost a lot and are designed to run on high end, expensive hardware.

You fail to consider the free variants of UNIX. FreeBSD, NetBSD, OpenBSD, and DragonflyBSD are all examples of freely available decedents of UNIX that run on all sorts of inexpensive hardware.

Further, your statements about UNIX applications and Linux applications is about 6 or 7 years out of date. The big commercial UNIX applications (Oracle etc) are now available for Linux. Open Source “Linux” applications (apache, etc) have always been available for UNIX.

You state: “Mac OS X Leopard is built upon UNIX code.”

Actually, Mac OS X Leopard is built upon the BSD code base (with a little GNU thrown in here and there). While it is true that BSD was originally based on the UNIX code base, the BSD developers replaced all the original UNIX code long before Apple based its OS on BSD. Therefore, it is incorrect to state that Mac OS X Leopard is based on UNIX code as it does not include any UNIX code.

Just for verification, I check with my personal trainer. He had to look it up, but then he agreed.

Like the Dunkin guy, I think you are struggling with the difference between UNIX-like and UNIX variant. UNIX variant = derivative of the UNIX code base. UNIX-like = Uses an API and userland which in many superficial ways resemble UNIX.

By the way, Minix is UNIX-like and not a UNIX variant. Education is fun!

Second, to say that Linux could not stand on its own with UNIX is false. There isn’t any UNIX to remove from Linux. If your point is that Linux resembles UNIX, well, yes it does. If your point is that Linux is a derivative of UNIX, no it isn’t. Try doing a little research next time.