From owner-freebsd-questions Mon Nov 27 20:22:59 2000 Delivered-To: freebsd-questions@freebsd.org Received: from guru.mired.org (okc-65-26-235-186.mmcable.com [65.26.235.186]) by hub.freebsd.org (Postfix) with SMTP id 0333137B4D7 for ; Mon, 27 Nov 2000 20:22:54 -0800 (PST) Received: (qmail 38526 invoked by uid 100); 28 Nov 2000 04:22:53 -0000 From: Mike Meyer MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <14883.13084.936215.485266@guru.mired.org> Date: Mon, 27 Nov 2000 22:22:52 -0600 (CST) To: Cliff Sarginson Cc: questions@freebsd.org Subject: Re: Linux vs. FreeBsd (reposted) In-Reply-To: <105953745@toto.iv> X-Mailer: VM 6.75 under 21.1 (patch 10) "Capitol Reef" XEmacs Lucid X-face: "5Mnwy%?j>IIV\)A=):rjWL~NB2aH[}Yq8Z=u~vJ`"(,&SiLvbbz2W`;h9L,Yg`+vb1>RG% *h+%X^n0EZd>TM8_IB;a8F?(Fb"lw'IgCoyM.[Lg#r\ X-Message: You should get a better mailer. Sender: owner-freebsd-questions@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG Cliff Sarginson types: > Long ago there was the 6th Edition of UNIX, the first available > outside of Bell Labs. It was cute, minute and very fast. It had > a handful of system calls, the original Bourne Shell (which had > a "goto" in it I recall, implemented as an external program). > Then there was the 7th Edition. Everything was made "bigger", > data structure sizes, system calls, tools (like make!) etc.. > It ran like a dead pig on the PDP11's of the time. v6 was very small enough that one person could understand the kernel, and the source code fit in a book small enough to be xeroxed. I don't know that anyone ever tried to do that with v7. But "you are not expected to understand this." > At UCB they took this version and BSD started to grow... > > The 7th Edition didn't last long. BSD Unix was starting to make a > noise, particularly on VAXes..later came AT&T System 3, there was > never a System 4, and then System 5. System IV was never released. Under the consent decree, AT&T couldn't compete in the computer software market, which meant they only sold Unix that was no longer being used internally. So while they were selling System III, they were running System IV and V internally. When they were broken up, they could compete, and started selling System V. > System V integrated many of the things from BSD. So did System III; just not as many. > BSD became the basis for certain versions of UNIX, most notably > Sunos..aka these days Solaris, and Ultrix (DEC Unix of it's day, > the most awful Unix system ever let loose). Sun wins the award for the most confusing naming, hands down. SunOS through and including 4.x is BSD-based (and fed stuff back to BSD). SunOS 5.0 and beyond is SysV-based, but has a good mix of SunOS 4.x features. The *windowing* system that is bundled with SunOS is called Solaris. Solaris 1 was what went with SunOS 4. With SunOS 5, you can't get SunOS by itself (at least not to my knowledge). So the SunOS 5/Solaris 2 combination is what people mean when they say "Solaris". SunOS 5 and Solaris 2 are lockstepped - SunOS 5.x is bundled with Solarix 2.x. Of course, that's not confusing enough. With SunOS 5.7/Solaris 2.7, Sun declared they weren't going to do another major release, so they were going to drop the first digit. So if you run Solaris X (X >= 7), you get SunOS 5.X and the Solaris 2.X windowing system. People *hated* Solaris (the SunOS 5.x/Solaris 2.x bundle) when it first came out. There were "Death to Solaris" BOFs at Usenix. I still remember Solaris 2.3 panicing regularly when fed certain flags to ps. Not only that, they flagged it as "unfixable; don't do that until you get 2.4". That kind of thing drove many people to Ultrix and HP-UX. Finally, Ultrix can only be considered the most awful Unix system ever let loose if you don't consider AIX to be Unix. Having done my time in h ... maintaining AIX systems, I won't argue with that position. > Most others got based loosely or otherwise on licensed code > from AT&T, and there was always obeissance to the Regents of > the University of California" in the copyright notices. > This is the basis for HP-UX, Dynix/PTX, SCO etc etc... Yup - everybody but AT&T ships some mixture of AT&T and BSD. And even AT&T uses BSD code. > So ... without the BSD kernel stuff and other things UNIX would > not be the system it is today. The TCP/IP stack was probably the most important bit (that migrated to other OS's), but csh and vi, and rogue rank right up there. > Since RISC has been imho the biggest single barrier to the > development of a common operating platform, this may not > be a loss. I suspect proprietary version of UNIX as such will > eventually die out. Possibly not Solaris for a long while, but > Sun have always been en enfant terrible, refusing to accept standards > except their own for example. We have a common operating system. It's just not good enough for some people; that's why we run FreeBSD. More seriously, I think you've confused cause and effect. The reason we don't have a common OS is because most vendors felt they had to extend the standards to provide "extra value". Vendors who were trying to make money selling software for CISC hardware played the same game. That's why RISC became popular - because the vendors felt they could provide more value than they could with CISC. > BSD I gather contains no code from the orginal BSD Unix, but > my recent limited exposure to it has definitely given me the (old) > BSD impressions. I have no exposure to the other systems with > BSD in their name, so I will not comment. Did you mean to say "BSD Lite has no code from the original AT&T Unix"? That's true. FreeBSD has code that shipped with BSD in it, and BSD had AT&T code in it, *except* for the BSD Lite distribution. > Oh I forgot to mention AIX, IBM's Unix. There I've mentioned it. "It'll remind you of Unix."