Talk:Xenix
This is the talk page for discussing improvements to the Xenix article. This is not a forum for general discussion of the article's subject. |
Article policies
|
Find sources: Google (books · news · scholar · free images · WP refs) · FENS · JSTOR · TWL |
![]() | This article is of interest to the following WikiProjects: | |||||||||||||||||||||||||||||
Please add the quality rating to the {{WikiProject banner shell}} template instead of this project banner. See WP:PIQA for details.
Please add the quality rating to the {{WikiProject banner shell}} template instead of this project banner. See WP:PIQA for details.
|
![]() | A fact from this article was featured on Wikipedia's Main Page in the On this day section on August 25, 2004. |
Microsoft and UNIX
I remember a presentation in 1999 of the editor of the Linux Journal playing up a number of Microsoft ads from the early to mid 1980s where Bill Gates states that he is convinced that UNIX was the best Operating System around. Not entirely unbelieveable: DOS not only survived, but thrived despite both MS's & IBM's best attempts otherwise. And I suspect MS would like to forget their involvement in this 30-year-old technology; best reason for the changes I made to the relevant articles. -- llywrch 03:02 Mar 14, 2003 (UTC)
- Truth be told, this is not a simple tale and I doubt there is much shame at MS over their long UNIX history (only perhaps a bit of marketing spin to keep consumers focused on Windows). There have long been rumours Gates recommended Xenix to IBM for the PC but was flatly turned down since MS was but a licensee of AT&T, IBM's biggest potential competitor at the time (the AT&T breakup was looming and with it, AT&T's entry into the computer market). Gates then offered CP/M, but the IBM guys didn't negotiate well with DR and came back to Gates still looking for an OS so at last he scrounged and came up with Quick 'n Dirty, which was about his last easy option. MS/PC-DOS, even cleaned up, was utterly primitive compared to Xenix, but its fast, unprotected calls and wide open structure worked well for games too and licensing costs were nill. When he realized how much money could be made on an open hardware platform with a light OS (program launcher, really) that MS owned outright and which met a cheap, marketable standard (which included "hackability" for lucrative games), he went with the market (never mind the pressures of MS having gone public after the IBM deal) and the public ultimately bought the cheapest OS available- MS-DOS. There are also rumours one of the many reasons Paul Allen left MS was his unhappiness with the shift in focus away from UNIX-like products (MS had been a small but respected language vendor edging into UNIX for the low end mini-computer market before its meeting with destiny and IBM). Meanwhile by 1987 Microsoft's Xenix was still the most widely installed form of UNIX in the world: MS-DOS, OS/2 and NT were all fundamentally influenced by it. Early Windows was built on Xenix boxes. In effect, internally MS was a UNIX shop until 1993. Gates was still preaching the virtues of UNIX as late as 1996 and even claiming that NT was (in a "weak" sense, his word) a form of UNIX. Microsoft's gradual withdrawl from the UNIX market likely made it many more billions than it would have made otherwise but also left a wide open opportunity for Linux and later the BSDs to attract the markets MS left untended. Meanwhile MS bought/licensed/
stoleimitated the best, cheaply implemented the spoils and along with some cunning business practices shoved them into the market pipeline which by 1995 or so it utterly dominated. Gwen Gale 19:01, 15 January 2007 (UTC)
--- Much of this story is true. Paul left because he got sick, and I think he felt left out when he came back over a year later. He never really got back up to speed again.
Initially, Bill Gates thought games were a good reason someone might want a computer, but by the time the IBM PC shipped in 1981, he'd come to hate them as a time-wasting distraction. the perception of MS/DOS as a great game platform is at best misguided. IBM specifically did NOT do sound in the initial PC because they felt that sound would make it seem like a toy and scare business people away. to put it on the motherboard as the Mac did 2 years later would have cost pennies. Game support was the last thing IBM wanted, and it was with great resistance that MS ported their 8 bit games to it. User:Hans42 8 Dec 2011 —Preceding undated comment added 08:50, 8 December 2011 (UTC).
SCO & 8086 MMUs
To me at least this article doesn't make clear the the SCO here is "old SCO" —The preceding unsigned comment was added by 68.45.4.118 (talk • contribs) 10:09, April 16, 2006 (UTC)
<daniel@nelxe.dk>: I think that Microsoft actually used to own Santa Cruz Operation....so they didn't really abandon the UNIX efforts...SCO released SCO Unix..
---
Microsoft wound up getting partial ownership of (old) SCO for what I'd understood to be non-payment of bills. I don't think it was ever more than 1/3rd.
Microsoft owned 20% of The Santa Cruz Operation as of 1988. They invested in SCO and this allowed them to, by 1989, effectively exit the UNIX business while still generating revenue through their investment in SCO and royalties. — Preceding unsigned comment added by Ldimershein (talk • contribs) 04:08, 11 June 2018 (UTC)
The first port of Xenix 8086 was to Altos. Altos had several machines which implemented this hardware configuration. There were a couple of compatible versions called, I think, 8600 and 586. this had an 8086 CPU connected to a Memory Management Unit (MMU) which allowed a level of demand paging. it didn't support pagefaults or any such niceties but it did allow a reasonable level of swapping. Real mode 8086 doesn't support any of this. I don't think it had a general bus but it did support an 8087 and extra memory. The port was done in late 1980 and early 1981 by Gordon Letwin and the first version shipped in the Autumn of 1981. The second port was also done by Gordon in late 1981 for Intel. The first prototype we got from Intel had a square front about a foot on a side and was about 20 inches deep. I can't remember what it was called, but it had an 8 inch floppy drive oriented vertically, in addition to the internal hard disk. it was built around a Multibus backplane. by early '82 it had been replaced by a similar box using 5" floppies and I think a 20MB hard disk in the same cubic format, and a couple of years later a flatter white box more typical of the desktop configurations of the era.
The first SCO real mode port of xenix didn't come along until 1983--two years after Gordon's original--and it didn't work well for more than a year after (or ever, really--it's unmapped 8086...). By this point the 286 was available, which had a real memory map on the chip, so there was little point in continuing with 8088 or 8086 real mode. (a memory map helps several things: it allows more physical memory to be available than the address space of the machine (1MB in the case of the 8086, and only 640K on the PC). It allows each app to have its own address space, which makes it more practical to run multiple of them, and it keeps errant pointers from being used to trash the running kernel or other apps). the macintosh and real mode windows imposed fairly onerous constraints on applications to allow very limited (and flakey) multitasking, and the SCO real mode port wound up doing something similar--namely "small" and "medium" model apps.
Most of the system and support was done by the Microsoft Xenix group through this period--1980-87 or so. SCO, Altos, Tandy and the others maintained or updated ports which were /started/ at microsoft. /NOT/ the other way around.
-Hans, Xenix Compiler guy — Preceding unsigned comment added by Hans42 (talk • contribs) 17:10, 3 June 2012 (UTC)
- Yeah, I've been working on the Altos page and articles about the 586/8600 (e.g. [1]) do talk of something that sounds like virtual memory (using 4K pages), with swapping to disk supported as well, but without demand paging. They almost certainly had an external MMU (as the whole Altos 8086 arch wasn't PC compatible, e.g. it used an 8089 for I/O). But details about the MMU are very scarce; I'm guessing it was all discrete. Someone not using his real name (talk) 15:49, 20 February 2014 (UTC)
- I was able to find out that another 8086 Xenix-dedicated machine, called Nabu 1600 had an external MMU (board) [2], so most likely Altos had something similar. Unfortunately, while I can find block diagrams for the Altos 8000 on the net (in its manuals), there aren't any for the 8600 (on the net.) Someone not using his real name (talk) 15:59, 20 February 2014 (UTC)
- Seattle Computer for example offered a MMU board to go with their 8086 CPU board [3], which they even say were used (in combination) in their own Xenix computer. Someone not using his real name (talk) 16:04, 20 February 2014 (UTC)
- Here are the schematics of the Nabu. The last 4 page or so are the MMU (made of discrete ICs). Someone not using his real name (talk) 16:51, 20 February 2014 (UTC)
SCO Xenix for PC ship date?
The 1983 date for the port is probably a "it worked then" kind of date. By 29 May 1984 they only had a beta to give out for PC Magazine to review. Since IBM did not distribute Xenix for the XT (they distributed PC/IX instead), SCO would have had to sell Xenix for the XT as shrink-wrapped (or at least mail order direct). So it must have had some clear shipping date sometime in 1984, but I'm unable to find it. It's also possible that the PC Magazine review was delayed a fair bit because it appeared in a themed number, with QNX and Idris also reviewed in it. (The June 12 issue of PC Mag then reviewed PC/IX, uNETix, Venix and Coherent.) Someone not using his real name (talk) 03:16, 22 February 2014 (UTC)
By the way, Pate (p. 10) says this: "In 1983 the PC emerged. SCO started porting to the 8088 but concentrated on the 8086, producing a release of SCO XENIX in 1984 which ran in 640 Kbytes with a 10 Mbyte hard disk. The release could support three or more users simultaneously, had multiscreen (virtual console) facilities, Micnet local area networking and enhancements added from 4.2BSD." By "In 1983 the PC emerged" he probably means the XT (the first hard-disk version of the PC), which was released in March that year. Someone not using his real name (talk) 03:37, 22 February 2014 (UTC)
I bought an IBM XT/286 (IBM's "official AT clone") in about 1989 and I bought IBM XENIX for it. I have never heard of Microsoft selling XENIX. I know I purchased extra memory to support XENIX. Yes, the XT was the first PC with a hard disk but the XT/286 had a 80286 processor, not an 8088. Sam Tomato (talk) 01:56, 9 July 2017 (UTC)
- When you say I have never heard of Microsoft selling XENIX, I think it depends on what you mean by selling. I believe that in most (all?) cases, Microsoft didn't sell Xenix directly to end-users, they would sell it via hardware manufacturers (such as IBM). So most Xenix end-users didn't directly purchase Xenix from Microsoft, they bought it from the vendor of their hardware (OEM). (This wasn't unique to Xenix, but was common to many Microsoft products; MS-DOS was only sold to OEMs for the first (almost) ten years of its existence, and couldn't be bought retail until 1991; most home computer users who used Microsoft Basic never bought it from Microsoft, it came with their computer; etc.) SJK (talk) 09:54, 10 July 2017 (UTC)
Xenix descendants
I've added a little information on the later descendants of Xenix (SCO UNIX and OpenServer) and a redirect from OpenServer, but the information is pretty sparse. Anyone who knows more about this should add to the article. (I know, I know, SCO is "evil" these days - I run Linux myself - but that doesn't mean Wikipedia shouldn't list information about them.) Beinsane 00:02, 3 Mar 2005 (UTC)
Available Microsoft XENIX Software
I removed this section, which previously said, in full: "Microsoft Pascal Compiler for the 286 XENIX Operating System (didn't support 386 extentions". I considered cleaning it to: "Microsoft Pascal Compiler for the 286 Xenix Operating System (did not support 386 extensions)" but then considered the section was too sparse to stand alone. If anyone can flesh out the section with other software available then feel free to add it back in. Pelago 22:25, 31 May 2006 (UTC)
None of the XENIX compilers for 8086, Microsoft C, supported either optimization or the iAPX-286. None of the XENIX compilers for XENIX 3.0 supported the iAPX-386, and you could not run XENIX 3.0/IBM XENIX on a 386. The compilers for XENIX 5.0 supported the 386, and its features.
Internally XENIX Compilers for the VAX supported 7 or 8 architectures, 8086, 88000, 68000, 80286 ISA, 80286 MC, 80386 ISA, you had the Bus Wars, the UNIX wars and the compilers all going on.
"Microsoft Pascal Compiler for the 286 XENIX Operating System User's Guide, part number 8511I-330-05, document number 020-092-013, from 1985. We used this compiler on XENIX 386, but not for long, because it could not take advantage of 386 instructions and, as with all things Microsoft, it was a little buggy."[ https://williambader.com/museum/at/pascal.html ]
Microsoft Pascal 3.3, for DOS supported XENIX 2.2.3, and 8086 real mode. https://books.google.com.au/books?id=jBnPebByBGkC&lpg=PA162&ots=mHxSt-6wYP&dq=%22Microsoft%20Pascal%22&pg=PA162#v=onepage&q=%22Microsoft%20Pascal%22&f=false
In the middle of this, you have Microsoft XENIX 3.2.3, which supports three Architectures, 286 MC/PS/2, 286 ISA/AT, and 386/ISA, It is unknown if it worked on 386 MC, or what Pascal 3.3 supported for Targets, in a few months, Microsoft would sell XENIX to SCO, invest in OS/2, have a fight with IBM over OS/2 and secretly start on Windows NT, which would mean that all their XENIX work from the DEC VAX compilers and cross compilers would stop.
Pascal in particular, feeling pressure from Turbo Pascal would be transformed into Quick Pascal. Microsoft was feeling pressure on the OS front, and the development front, retiring their program development tools slow and buggy. They would in the following years, get a working OS NT 3.1, get their mail servers running on it, and get all their development tools running on it, MS C++, MS Pascal, MASM 5.
The glimpse from Sept 12, 1988 is a market in Transition, IBM drops the prices on The Model 50/60 286, Borland announces Turbo Pascal 5, and Turbo C. https://books.google.com.au/books?id=fToEAAAAMBAJ&lpg=PA59&ots=1d6bhlXyvB&dq=%22Microsoft%20Pascal%22&pg=PA51-IA19#v=onepage&q=%22Microsoft%20Pascal%22&f=false
Microsoft Pascal 4.0, for DOS supported OS/2, and 286 features. ( and pretty quickly Windows NT ) It was an entirely different approach, emitting C++ code which would use the C++ compiler, and allow access to mixed language on the same projects, going from a 3 pass compiler to a 4 pass compiler, and becoming even slower. https://books.google.com.au/books?id=fToEAAAAMBAJ&lpg=PA59&ots=1d6bhlXyvB&dq=%22Microsoft%20Pascal%22&pg=PA51-IA19#v=onepage&q=%22Microsoft%20Pascal%22&f=false
https://williambader.com/museum/dell/xenixhistory.html
I am kind of getting tired of how easy it was to find this out with ONE google search. ONE. — Preceding unsigned comment added by 170.75.140.124 (talk) 20:27, 13 November 2020 (UTC)
A Question
How could microsoft get the licensed from AT&T in the late 1970s while it was founded in 1975? (see microsoft page) —The preceding unsigned comment was added by 75.80.152.28 (talk • contribs) 04:22, September 27, 2006 (UTC)
- Why not? Gwen Gale 19:02, 15 January 2007 (UTC)
- The story here is complicated. SCO, HCR and a few others had been doing pdp-11 unix support for a while. Bob Greenburg at Microsoft had learned unix internals while at Harvard, and when he saw that 16 bit microprocessors would soon be available (early '78 I think), he started negotiating with AT&T for license. AT&T resisted for a long time, but eventually came around some time in 1980, with the caveat that it not be called "unix". at the time, AT&T was still under the consent decree which prevented them from playing in any markets other than telephony, which complicated negotiations considerably. The deal that was structured required payments to AT&T inverse relation to the number of binaries sold: the first few were very expensive, gradually getting cheaper as more were sold, until it got down to pretty cheap when a really lot were sold. It was not very profitable for AT&T, which fit with the consent decree constraints. AT&T made this deal with a bunch of companies: Microsoft, Onyx, Unisoft, and several others. the deal also allowed third parties to sign on as part of the various second parties groups, for purposes of source distribution and count of binaries sold. HCR, SCO, Altos, Intel and about a dozen more, were part of Microsoft's group. Sun, and a number other 68000 vendors, mostly based on Andy Bechtolshein's design, were under Unisoft's, etc. (we also did one to Bechtoshein's design, for Forward Tech, and our port ran with very little work on Sun's hardware). Most of the others either disappeared or became third parties to either Unisoft or our deal with AT&T. this largely ended in 1987 with the merge of Microsoft, Sun (which had absorbed Unisoft), and AT&T's source code. Microsoft decided that three operating systems was one too many (Xenix, OS/2 and MS-DOS) so we offloaded xenix to SCO at this time. (windows ran /on top/ of MSDOS until win386 in 1989) --Hans — Preceding unsigned comment added by Hans42 (talk • contribs) 17:46, 3 June 2012 (UTC)
- The thing I don't understand is how can AT&T sell a license for Unix when they were legally prevented from selling Unix? It is my understanding that large portions of Unix were developed by others, such as universities, and were in the public domain. So if any licensing was necessary then I don't understand how that is possible, legally speaking. Sam Tomato (talk) 02:15, 9 July 2017 (UTC)
- @Sam Tomato: I think this is the point–they were never "legally prevented from selling Unix" per se. What they were prohibited from doing was leveraging their near-monopoly in telecommunications to gain market share in other markets, such as computer hardware or software. So, they could sell licenses to computer software, so long as they structured the deal in such a way that they did not gain substantial market power in the computer software industry as a result. So, by licensing Unix to multiple value-added resellers (VARs)–who would compete among themselves–instead of directly to end-users (academic institutions excluded), and by ensuring that AT&T would only make a modest profit (with most of the profits going to the VARs instead), they could comply with the consent decree–they could argue since most of the profit went to the multiple competing VARs, it was their market share not AT&Ts, and hence AT&T wasn't gaining market share in the computer industry. (Obviously they asked their own lawyers to interpret what exactly the consent decree would and wouldn't allow, and I think it is plausible–but I don't know for sure–that they may have even asked the US Department of Justice for its opinion on what the consent decree allowed too–and the lawyers believed this sort of deal would comply.) But then, something quite legally complex gets simplified and distilled down to "AT&T wasn't allowed to sell Unix" (which isn't strictly true–they were allowed to sell it, and they did, but subject to very substantial restrictions), and becomes part of the folklore the everyone remembers, and I'm sure you'll even find printed books containing that claim, even though it is clearly an oversimplification. SJK (talk) 09:42, 10 July 2017 (UTC)
- @SJK: Whenever people talk about AT&T licensing Unix there is never recognition of the fact that much of Unix was developed by others and were in the public domain. So it is likely the truth is somewhere between. Whatever the truth is, I hope it is accurately documented. I am not trying to re-write history and I don't want others too either. (Did I do the ping correctly?) Sam Tomato (talk) 05:47, 11 July 2017 (UTC)
- A correction - almost none of Unix was in the public domain. The parts developed by others generally had their copyrights and were subject to their license restrictions. For example, BSD Unix was copyrighted by the Regents of the University of California (the standard copyright for anything developed there), and available under the original form of the BSD license - freely available (modulo the whole USL copyright brouhaha), but decidedly not public domain. Mahousu (talk) 15:33, 11 January 2019 (UTC)
First Unix/Xenix
"I had the first Xenix distribution (developed by HCR in Toronto) in the US, ahead of Microsoft."
"We ended up selling a few of the boxes. The company was called MSD, located in Grand Rapids, Michigan. The only record of such is in a 1981 (Jan?) issue of Byte with our little ad in the back."
http://slashdot.org/comments.pl?sid=29920&cid=3213453
Maybe the Xenix article should mention HCR & MSD?
HCR & Logica
"The actual delivery of Xenix was not done by MS, it was built by Human Computing Resources (HCR) in Toronto. "
http://groups.google.com/group/comp.sys.next.advocacy/msg/1c4b56e1f9099c1b?rnum=8&filter=0
- I've added a mention of HCR to the article, with this reference. Letdorf (talk) 11:39, 2 April 2008 (UTC).
- Their role isn't very clear. They (HCR and Logica) surely acted as distributors outside the US. Pate's book (p. 10) says this about them "In 1982, a joint development and technology exchange agreement was reached between SCO and Microsoft bringing together engineers from SCO and Microsoft to further enhance the XENIX operating system which was increasing in popularity. Microsoft and SCO worked together with Logica in the UK and HCR in Canada, producing enhancements to XENIX and porting XENIX to other platforms." The 1983 OEM directory says nothing about HCR (SCO might have bought it already by then), but says this about Logica:
“ | Product Description
|
” |
- It also shows that Logica had "First customer ship: April 1981" and also supported the fairly obscure 16032 processor besides the MS-supported ones. Someone not using his real name (talk) 03:34, 22 February 2014 (UTC)
Ronald Michael Baecker - University of Toronto
Founder and CEO, Human Computing Resources Corporation (HCR), 1976-78 (part time), 1978-81 (half time), 1981-84 (full time).
"Founded, led, and built a successful, world class, multi-million dollar software company, HCR Corporation, 1976-84. I founded HCR (formerly, Human Computing Resources Corporation) in 1976, investing $11,000. Between 1978 and 1982, although no other money had been invested, I led the company through growth of 100% per annum to annual revenues of $1.3 million, and to a position as one of the world's premiere companies specializing in UNIX-based software. With the help of venture capital, HCR grew to 1984 annual revenues of $4 million. I hired a new President to run the company in 1984. We sold HCR to the Santa Cruz Operation (SCO) in 1990, and it continued as SCO Canada until early 1996. "
http://kmdi.utoronto.ca/rmb/CV_RMB_2006Nov15.pdf
Byte article from May 1994
"There were 33 institutions on Ferentz's 1975 list of users; there were 138 in September 1976, 37 of them outside the U.S. And, in 1977, Interactive Systems (Santa Monica, CA) became the first company to support Unix commercially. It was soon followed by Human Computing Resources in Toronto."
http://www.byte.com/art/9410/sec8/art3.htm —The preceding unsigned comment was added by NevilleDNZ (talk • contribs) 13:27, 30 April 2007 (UTC).
NevilleDNZ 12:55, 30 April 2007 (UTC)
- Well, ISC did PC/IX (and other things) for IBM (totally independent of Xenix from all accounts), but it's not clear what HCR did. Also Venix was available for the XT before either PC/IX or Xenix were. Someone not using his real name (talk) 04:03, 22 February 2014 (UTC)
- ISC did IS/1 before that: a value-added V6 Unix for the PDP-11. QVVERTYVS (hm?) 16:13, 22 February 2014 (UTC)
Early Xenix Versions not mentioned in article
I worked in a Tandy Computer Centre around 1985. I took to Xenix in a big way and ended up collecting a few of The Tandy's. I also ended up with a lot of manuals. (All gone now..) The TRS-80 Model 16a - 16b were 8" floppy. The first Xenix I could find (I ended up working in Tandy Australia's Head Office - Doing Computer Support) Was Version 7. Not System III. Version 7 shipped with TRS-80 Model 16a's. 16b's were sold with the Version 7 and needed to be upgraded to System III when it became available. I would say that all Tandy Model 6000's (the last in the range) were sold with System III. Althought the earlier System 7 would operate. There were subtle hardware differences between the 2 models. Therefore System III seemed the better choice. Also the system 7 implementation plain and simple had more issues. The System III version had more bells and whistles. Of course developement systems were available to both flavours of Xenix. The source of both versions contained SCO and Microsoft ownership messages. So SCO and Microsoft were involved in Xenix pre System III. I also had an early Tandy Xenix manual that described the proceedure for bootstrapping Xenix on a PDP11 from a tape backup unit. It was supplied as a guide with one of the early tandy units I picked up. That unit was running 8" winchester 8MB hard disks from memory. But the manual wasn't relevant to use to boot with. It was more as a early system admin guide. I think some of this needs to be mentioned. The earlier versions of Xenix were supplied on 8" floppy. No one mentioned Cromemco? When we were selling the TRS-80 Model 16a (for AU $10k basic system) the Cromemco was our nearest competitor in cost terms? —Preceding unsigned comment added by 203.173.51.90 (talk) 05:58, 21 October 2007 (UTC)
Intel?
Intel made computers? Intel is a chipmaker, I don't think they ever made a complete computer therefore why would they OEM Xenix??? —Preceding unsigned comment added by 24.185.28.136 (talk) 02:58, 14 April 2008 (UTC)
Intel made several computers, mainly using multibus. They did a series of 8080 based development machines in the 70s, that included the infamous ICE (in circuit emulator), which they also sold separately. The first 8086 one we had in the xenix team at microsoft in 1981 had an 8 inch floppy, a purpose made memory map using the "buddy system", and a smallish hard drive--probably 20MB or so. I can't remember its name. Later on, they made a more compact system that had 5" floppies and fewer card slots. it resembled the 386 based "320" in the reference above. I think it was called the "586" but my memory isn't clear. we worked with intel extensively on the design of the 80286, and integrating the memory mapping system with the CPU was largely due to pressure from us. — Preceding unsigned comment added by Hans42 (talk • contribs) 08:40, 8 December 2011 (UTC)
From the '83 Xenix OEM catalog:
“ | Product Description Microsoft XENIX Operating System from Intel Intel offers the Microsoft XENIX operating system to volume end-users and OEMs using Intel boards and systems. The Intel product is the complete XENIX product derived from Bell Laboratories' UNIX Version 7. This includes the operating system, program development environment, text processing, information handling and miscellaneous utilities. The current Intel product is configured for operation in the Intel System 86/330 and System 86/730 families. This includes the System 86/330X, System 86/380X and System 86/730. The System 86/330X includes an iSBC 86/30 processor board, an iSB[E]C 309 memory management unit, 384KB RAM, a 35MB Winchester disk, and a 1.1MB floppy disk in a pre-assembled, pre-configured system. The System 86/730 is an 86/330X with 768KB RAM and integrated software (not available separately) that includes spreadsheet (Multiplan), word processing, relational database management, and a menu/forms package. Both these systems offer high performance multi-user support in a systems environment. The current Intel product will also operate with an equivalent user-assembled system of Intel boards. These include the iSBC 86/12 with an iSBC 308 or an iSBC 86/30 or 86/14 with an iSBC 309 for the processor board and iSBC 215 and iSBX 218 disk controller. A minimum of 20MB of hard disk is also required. Future releases of the product will include additional device drivers, Microsoft XENIX enhancements (including a System Ill Upgrade), and configurations tor the Intel iAPX 186 and 286 microprocessors. Intel offers the complete XENIX solution including hardware, software, and support. |
” |
I think the one appearance of "iSBEC" was a typo (but iSBX isn't--Intel was very fond of coining brands back then), but otherwise the ad speaks for itself. Also given is "First customer ship: July 1982". Someone not using his real name (talk) 04:09, 21 February 2014 (UTC)
Commercial Information
Does anyone care about non-technical information about Xenix, such as that from a SCO price list? I'm trying to throw away some old literature from the Xenix days, and don't want to discard anything that historians might consider to be useful some day. The pricelist includes, for example, the list of hardware manufacturers and models for which Xenix distributions were available in November 1, 1987, the products available for each, and the prices, which differed for the various platforms. Also, I have a technical training schedule and pricelist, a SCO background paperon "Multiuser vs. Networking", a "Xenix 2.2 Technical Backgrounder" dated January 1987, and "The SVID as Today's Best Non-Proprietary Specification for an Interface to an Operating System" background paper dated September 23, 1987. R68000 (talk) 23:11, 21 December 2008 (UTC)
the Xenix file system
- See file:///usr/src/linux/Documentation/filesystems/sysv-fs.txt (if you're on a Linux system)
- Very little at http://tldp.org/HOWTO/Filesystems-HOWTO-9.html#ss9.26 , except that it is classified as a derivative of System V FS (in agreement with the previous "ref").
- On SCO EFS/AFS/EAFS : http://osdir.com/ml/linux.file-systems/2004-05/msg00027.html
--Jerome Potts (talk) 11:47, 24 March 2010 (UTC)
link errors
link to http://www.unicom.com/pw/faq/sco-xenix.faq is dead —Preceding unsigned comment added by 76.105.243.211 (talk) 08:15, 24 September 2010 (UTC)
File:Xenix Screensnap.PNG Nominated for speedy Deletion
![]() |
An image used in this article, File:Xenix Screensnap.PNG, has been nominated for speedy deletion for the following reason: Wikipedia files with no non-free use rationale as of 19 May 2012
Don't panic; you should have time to contest the deletion (although please review deletion guidelines before doing so). The best way to contest this form of deletion is by posting on the image talk page.
To take part in any discussion, or to review a more detailed deletion rationale please visit the relevant image page (File:Xenix Screensnap.PNG) This is Bot placed notification, another user has nominated/tagged the image --CommonsNotificationBot (talk) 10:14, 23 May 2012 (UTC) |
Only Z8001?
According to the 1983 OEM directory, there was no such limitation. There are only two Z8000-based systems hardware providers there (w/Xenix, that is). Central Data Corporation (with a Z8000 product, first customer ship: January 1981) and Paradyne Corporation (with a Z8001 product, so I'm guessing this was the fabled first port, although given that their first customer ship was March 1981, this issue might be wrong as well.) As a side note, I think the external MMU issue applied to them as well. Someone not using his real name (talk) 17:08, 20 February 2014 (UTC)
- The CDC Z8000 stuff has some archival docs here (possibly incomplete). But there's nothing about Xenix in there so maybe they never shipped a product themselves (even though MS shipped Xenix to them.) Someone not using his real name (talk) 18:36, 20 February 2014 (UTC)
- Well, if the marketing department and the XENIX Product manager are to be believed, then Microsoft/SCO was planning on porting XENIX to both the Zilog Z8001 and the Z8002. ( Oh, like I have ever seen/heard of these ever... and I was there. ) Looks like the Z8001 was the chip, and the Z8000 was the processor. 209.129.161.251 (talk) 01:13, 12 March 2019 (UTC)
- Both the Wikipedia article on the Z8000, and the BYTE magazine article by Bob Greenberg, Jun 1991, mention that XENIX was made for the Z8000, and The article refrences on the Z8000 mentions it too. http://www.softpanorama.org/People/Torvalds/Finland_period/xenix_microsoft_shortlived_love_affair_with_unix.shtml 209.129.161.251 (talk) 01:13, 12 March 2019 (UTC)
"believing that it could not compete"
Indeed, the AT&T 6300 Plus shipped Unix System V for $99. The lowest I have SysV from anyone else around this time frame was $595. I'll see if I can find some secondary source to comment on this huge price discrepancy between AT&T's own distro and what OEM vendors could get (from AT&T). Someone not using his real name (talk) 18:00, 20 February 2014 (UTC)
- It was indeed $99 for the OS for the AT&T 6300. You just had to have an AT&T 6300 UNIX PC. 209.129.161.251 (talk) 01:06, 12 March 2019 (UTC)
Funny thing
In that '83 OEM directory, MS put themselves as Xenix customer/OEM for... Apple Lisa (with a note that the hardware was supplied by Apple.) Someone not using his real name (talk) 18:45, 20 February 2014 (UTC)
SCO sold a XENIX distribution for the Apple LISA back then. Microsoft was involved with its development. — Preceding unsigned comment added by Ldimershein (talk • contribs) 04:02, 11 June 2018 (UTC)
IBM floppy photo
It looks like the floppy is of the IBM PC AT Xenix 1.0 mentioned here, which is actually based on System III, so probably based on MS Xenix 2.x or even 3.0. IBM had this issue that they didn't want to release products for their machines with version numbers based on somebody else's product versioning scheme. Someone not using his real name (talk) 11:41, 21 February 2014 (UTC)
Quote from [4]:
“ | In April 1984 IBM announced the XENIX OS for it’s m68k-based System 9000. This one was based on MS XENIX 2.3 and UNIX V7, for more details see Announcement 284-159.
The IBM PC XENIX OS Version 1.0 was announced in August 1984 (284-279) for IBM PC AT 5170 Model 99 and should had become available at Q1 1985. So it’s basically an OEM version of XENIX 286 implementing a subset of UNIX System III (called XENIX System III) which has been ported by the Santa Cruz Operation. IBM PC XENIX OS Version 2.0 announced in March 1986 (286-115) was based on a subset of UNIX System V. |
” |
HTH, Someone not using his real name (talk) 11:51, 21 February 2014 (UTC)
Versioning contradiction
We have a contradiction in the article. On the one hand, we have primary sources for the existence of 3.0 and secondary for 5.0; on the other, we state that the very last version was 2.3.4.
Now my interpretation, based on the screenshot, is that the confusion stems from the following:
- Xenix was at some point upgraded to System III (aka Unix 3.0, as it's called in the System III source code), then System V (aka Unix 5.0).
- After System V, AT&T froze the numbering so the next major release of AT&T Unix was System V Release 2. The screenshot reflects this numbering by identifying the running OS as "SCO Xenix System V", more specifically "SysV release 2.3.2".
I'm going to make changes to the article based on this interpretation, which does not contradict any of the cited sources. If a source can be found that contradicts my edits, feel free to add it and edit accordingly. QVVERTYVS (hm?) 10:31, 28 February 2014 (UTC)
Your changes are accurate, but simplified. XENIX 1.0, was System 7, XEXIX 2.0 was System III, XENIX 5.0 was supposed to be System V, but was release before that happened, but became System V before its release. It was started as system III+BSD, morphed into System 7, and was late because of the pending System V. There is a chart, but its only slightly accurate as product changes were made between announcement and shipping. "Fries! Now with Katsup!" This was at the middle ages of the UNIX Wars. ( there were at least 10 UNIXs for PCs at this time. )
Virtual terminals
Although it is not discussed in the article, I suspect that virtual terminal capability turned out to be one of the most influential and historically important features of XENIX. Apple's melding of virtual terminals with the Xerox/PARC GUI idea resulted in the windowing GUI - an incredibly powerful synergy. XENIX is a historical sleeper that way.
75.111.20.66 (talk) 22:13, 13 December 2014 (UTC)
- Can you prove it? QVVERTYVS (hm?) 15:15, 14 December 2014 (UTC)
- "I suspect...." Draw your own conclusions. Interesting comment about XENIX and the Apple LISA above. While XENIX-LISA was offered by Microsoft, hardware requirements for XENIX were very specific, i.e. the compatibility of the LISA with XENIX was not likely to be accidental. It is virtually certain that Apple had its hands on XENIX while developing the LISA GUI. Under the Features section, it is noted that XENIX included a visual shell for menu driven operation (also a feature of Apple's GUIs), although it is arguable that it was an adaptation of earlier menu driven interfaces.
- The multiple desktop feature of LINUX is another melding of virtual terminals and GUI. Virtual console might be a less ambiguous way to refer to the feature that uses one display and keyboard (or GUI pointer system) for logically independent interfaces (windows, virtual desktops).75.111.20.66 (talk) 21:32, 16 January 2015 (UTC)
Secure/Trusted Xenix
Trusted Xenix, under the name Secure Xenix, was first developed at IBM Federal Systems Division, beginning in 1985. There were (as I recall) two releases, in 1987 and 1988. We also released a text-windowed version, based on Viewnix. The B2 evaluation process began while it was an IBM product, but everything was given over to TIS starting in 1989. This was just the 286-based version; while we considered Xenix 386, at that point, IBM's focus had shifted to AIX. (Note: I'm familiar with the history as I was part of that group, starting in late 1986.) Mahousu (talk) 01:18, 10 August 2015 (UTC)
External links modified
Hello fellow Wikipedians,
I have just added archive links to one external link on Xenix. Please take a moment to review my edit. If necessary, add {{cbignore}}
after the link to keep me from modifying it. Alternatively, you can add {{nobots|deny=InternetArchiveBot}}
to keep me off the page altogether. I made the following changes:
- Added archive https://web.archive.org/20131219145551/http://www.computerhope.com:80/unix/xenix.htm to http://www.computerhope.com/unix/xenix.htm
When you have finished reviewing my changes, please set the checked parameter below to true to let others know.
An editor has reviewed this edit and fixed any errors that were found.
- If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
- If you found an error with any archives or the URLs themselves, you can fix them with this tool.
Cheers. —cyberbot IITalk to my owner:Online 04:04, 29 August 2015 (UTC)
AT&T selling Unix license to Microsoft
@Schily:, here you reverted my change where I had inserted the bolded clause into the below sentence:
- Bell Labs, Unix's developer, was part of the regulated Bell System and could not sell Unix directly to end customers.
with the comment "rv false claim: AT&T was not allowed to sell or market anything that telephone services". But surely they sold Microsoft (and many other companies) a Unix license (I assume by that you meant to say other than), which contradicts your claim that the only thing AT&T was allowed to sell was telephone services, since surely the Unix license they sold to Microsoft was not a telephone service. Also, the next sentence "It instead licensed the software to others" doesn't make any sense, since it sets up a false opposition between sell and license – AT&T sold a Unix license to Microsoft, and Microsoft sold Xenix licenses to end customers, which were sub-licenses of the AT&T Unix license. Legally, what AT&T did to Microsoft isn't that different from what Microsoft did to its end-customers. Microsoft could only sell customers Xenix licenses because that was legally permitted under the terms of the Unix license AT&T sold to it. Microsoft was effectively a Value-added reseller (VAR) of AT&T (whether or not AT&T ever actually used that term). Given that, I think my wording is more accurate than yours. SJK (talk) 10:15, 25 April 2016 (UTC)
- Please do not confuse the rules that have been valid before 1982 and what happened after that time. Schily (talk) 10:17, 25 April 2016 (UTC)
- @Schily: And how am I doing that? I'm not. SJK (talk) 12:09, 25 April 2016 (UTC)
- Of course, you are doing that because AT&T was not allowed to sell anything besides telecommunications before 1982 and this was nullified in 1982, so that AT&T could sell UNIX. Your changes completely miss these facts. Schily (talk) 12:19, 25 April 2016 (UTC)
- But they sold Microsoft a V7 Unix license in 1978, as the article itself points out ("Microsoft... purchased a license for Version 7 Unix from AT&T in 1978"). Your reversion completely misses that fact. AT&T was allowed to sell a Version 7 Unix license to Microsoft in 1978, which wasn't any kind of telecommunications, and was before 1982. SJK (talk) 12:29, 25 April 2016 (UTC)
- Wikipedia is not a reliable source and I cannot verify the given citation. The information I have is that AT&T did give away UNIX sources for free to various sites before the telecommunication monopoly ended in January 1982. As you claim something that is in conflict with everything I read about AT&T and UNIX before, you would need to give a reliable source to support your new claim. Schily (talk) 12:50, 25 April 2016 (UTC)
- You are claiming that "AT&T was not allowed to sell anything besides telecommunications before 1982" – where's your reliable source for that claim? Where is your reliable source for your claim that AT&T "did give away UNIX sources for free" prior to 1982 as opposed to charging fees for them? You can demand reliable sources for my claims, I'll do the same for yours. SJK (talk) 13:05, 25 April 2016 (UTC)
- According to this Byte magazine article, AT&T sold Unix V6 licenses for $200 for educational institutions, and $20,000 for commercial entities. Unix V6 was before 1982 (V6 came out in 1975, V7 came out in 1979). This directly contradicts your claim that AT&T couldn't legally sell Unix prior to 1982. SJK (talk) 13:10, 25 April 2016 (UTC)
- That Byte Magazine article (Fiedler, Ryan (October 1983). "The Unix Tutorial / Part 3: Unix in the Microcomputer Marketplace". BYTE. p. 132. Retrieved 30 January 2015.) says:
- Because, under a government-ordered consent decree, the Bell System was forbidden to compete in the commercial marketplace, Unix was offered to all users under very restrictive terms. A source license only would be granted, and the software would be offered as is, with no support, no refunds, no warranty and no maintenance.
- These terms were decidely uncompetitive because a Unix license would cost a company $20,000 (educational institutions were charged $200)...
- What this makes clear, is that AT&T could legally sell Unix, and did. What it could not do, was sell it in such a way which would be considered active competition in the computing market. By selling Unix without any of the ancilliary services usually expected with a commercial computer operating system (such as warranties, support, and maintenance programming), and at a commercially non-competitive price, they could avoid falling foul of the consent decree – but they still sold it. SJK (talk) 13:21, 25 April 2016 (UTC)
- Wikipedia is not a reliable source and I cannot verify the given citation. The information I have is that AT&T did give away UNIX sources for free to various sites before the telecommunication monopoly ended in January 1982. As you claim something that is in conflict with everything I read about AT&T and UNIX before, you would need to give a reliable source to support your new claim. Schily (talk) 12:50, 25 April 2016 (UTC)
- But they sold Microsoft a V7 Unix license in 1978, as the article itself points out ("Microsoft... purchased a license for Version 7 Unix from AT&T in 1978"). Your reversion completely misses that fact. AT&T was allowed to sell a Version 7 Unix license to Microsoft in 1978, which wasn't any kind of telecommunications, and was before 1982. SJK (talk) 12:29, 25 April 2016 (UTC)
- Of course, you are doing that because AT&T was not allowed to sell anything besides telecommunications before 1982 and this was nullified in 1982, so that AT&T could sell UNIX. Your changes completely miss these facts. Schily (talk) 12:19, 25 April 2016 (UTC)
- @Schily: And how am I doing that? I'm not. SJK (talk) 12:09, 25 April 2016 (UTC)
External links modified
Hello fellow Wikipedians,
I have just modified 3 external links on Xenix. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
- Added archive http://web.archive.org/web/20140107072346/http://www.tenox.net/docs/microsoft_xenix_30_286_press_release.pdf to http://www.tenox.net/docs/microsoft_xenix_30_286_press_release.pdf
- Corrected formatting/usage for http://www.robert.to/reports/scoxenix.html
- Added archive http://web.archive.org/web/20081202010444/http://www.unicom.com:80/pw/faq/sco-xenix.faq to http://www.unicom.com/pw/faq/sco-xenix.faq
When you have finished reviewing my changes, please set the checked parameter below to true or failed to let others know (documentation at {{Sourcecheck}}
).
This message was posted before February 2018. After February 2018, "External links modified" talk page sections are no longer generated or monitored by InternetArchiveBot. No special action is required regarding these talk page notices, other than regular verification using the archive tool instructions below. Editors have permission to delete these "External links modified" talk page sections if they want to de-clutter talk pages, but see the RfC before doing mass systematic removals. This message is updated dynamically through the template {{source check}}
(last update: 5 June 2024).
- If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
- If you found an error with any archives or the URLs themselves, you can fix them with this tool.
Cheers.—InternetArchiveBot (Report bug) 16:35, 16 July 2016 (UTC)
External links modified
Hello fellow Wikipedians,
I have just modified one external link on Xenix. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
- Added archive https://web.archive.org/web/20070114203948/http://kamidake.other-space.com/ to http://www.levenez.com/unix/
When you have finished reviewing my changes, please set the checked parameter below to true or failed to let others know (documentation at {{Sourcecheck}}
).
An editor has determined that the edit contains an error somewhere. Please follow the instructions below and mark the
|checked=
to true
- If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
- If you found an error with any archives or the URLs themselves, you can fix them with this tool.
Cheers.—InternetArchiveBot (Report bug) 21:05, 20 July 2016 (UTC)
succeeded by unrelated system
A "successor" should have more in common than just being later on the timeline. A reliable source is needed to support the editor's opinion TEDickey (talk) 22:09, 5 November 2017 (UTC)
- OS/2 is partly (~80%, give or take) compatible with Xenix, though not fully. It's more of a spiritual successor, though MicroSoft did obviously intend OS/2 to be it's replacement. As put on OS/2's page, "Because of this heritage, OS/2 shares similarities with Unix, Xenix, and Windows NT in many ways" and on Xenix's own page, "It agreed with IBM to develop OS/2,[4] and the Xenix team (together with the best MS-DOS developers) was assigned to that project". Both systems have a similar design concept. --2600:8801:2180:2E17:68B7:B7C:7E09:7EE8 (talk) 01:53, 6 November 2017 (UTC)
- That's interesting. You should provide a reliable source of information sometime, as a reality check. TEDickey (talk) 00:27, 7 November 2017 (UTC)
- "Similar design concept" does not a "predecessor" nor "successor" make. Nor does a high degree of "compatibility", whatever that may mean. Nor does chronological sequence, though that is generally essential, it is not sufficient. Jeh (talk) 10:24, 8 November 2017 (UTC)
- That is a fair point, they're not directly related. Frankly OS/2 has more in common with NT than it does with Xenix, but it is helpful to point out the OS/2 did succeed Xenix at least goal wise. --2600:8801:2180:2E17:5834:A62A:6A1F:A4A3 (talk) 04:36, 9 November 2017 (UTC)
Not related to contents. I am retracting everything I said.
|
---|
Hello. You might want to know that the IP address of this person matches a very well-known vandal, sock and block evader. He (or maybe she) has no rights to edit Wikipedia. Best regards, Codename Lisa (talk) 13:46, 6 November 2017 (UTC) Well, if I ever wanted an actual proof that this is the same sock, this revert proves it. If it was an innocent person, he or she would have replied with "you are wrong", "you are a liar", or "you are an idiot". But this one postpones the proclamation of "I am not every IPv6 anon" to the very end". His priority, as always, is villification, by saying "bad faith comment by Codename Lisa". And most importantly, calls me "she". I have not announced a gender and the pronoun "she" is only used by a certain group of people, against whom the sock has grudges. People who meet me first hand have never called me "she". — Codename Lisa (talk) 17:45, 6 November 2017 (UTC)
|
I think it is true that OS/2 was technically influenced by Xenix, the influence mainly goes through MS-DOS/PC-DOS. If you read the DOS 2.0 source code and the associated documentation, it is very clear that DOS 2.0 was heavily influenced by Xenix, and Microsoft planned to evolve DOS to become more and more like Xenix, although they never got that far (I guess "European DOS 4.0" was the closest DOS ever became to becoming Xenix-like). OS/2 was obviously heavily influenced by DOS, although I'm not sure how much independent influence Xenix (or other Unix variants) had on it. Microsoft and IBM knew that DOS was a very limited operating system, and they needed something more advanced to succeed it – initially Microsoft saw Xenix as having that role, and then OS/2, and then finally Windows NT. So I think it is true that OS/2 succeeded to the role of Xenix in Microsoft's business strategy. But, it is always troublesome trying to distill the complexities of reality to a single link in an infobox, and so in the end I think I agree with TEDickey that the link to "OS/2" is better omitted from the infobox, since it could cause confusion (e.g. the logical inference that Xenix had significant direct technical influence on OS/2, when it is unclear how true that actually is.) SJK (talk) 08:27, 7 December 2017 (UTC)
irrelevant edits
A see-also to a system lacking any influence other than a 30-year-separated involvement by one of the developers is irrelevant. TEDickey (talk) 21:57, 18 April 2018 (UTC)
- I agree. The link is to a Linux distribution by Microsoft. That has very little to do with Xenix. Also, if one is going to mention that, why not mention the NT POSIX subsystem, SUA/SFU, Windows Subsystem for Linux (WSL), SQL Server for Linux, Linux running in Azure, etc. This is all very tangential. SJK (talk) 20:01, 19 April 2018 (UTC)
Capitalization of Xenix to XENIX
It appears that period publication used all-caps for the name. If this was the standard name during its time then it seems that the article should replace Xenix with XENIX for greater accuracy. See an example: https://i.postimg.cc/VLbyqzt9/xenix.png — Preceding unsigned comment added by 107.77.193.114 (talk) 05:36, 20 August 2019 (UTC)
- Also, if that seems frivolous, imagine if people on Wikipedia were to start writing Dos instead of DOS. I doubt that would last long, for good reason. — Preceding unsigned comment added by 107.77.193.114 (talk) 05:44, 20 August 2019 (UTC)
- It looks like Wikipedia is a real mess when it comes to this. UNIX sometimes, Unix others. Venix sometimes, rather than VENIX. Someone in authority needs to sort this out by capitalizing names that need to be capitalized. Again, imagine if people were to randomly write Dos instead of DOS. It would be silly, like the situation currently is with UNIX and UNIX-like systems here. — Preceding unsigned comment added by 107.77.193.114 (talk) 06:10, 20 August 2019 (UTC)
What does :9 [note 2] mean?
What is this on the Article page? [12]:9[note 2] right after: The first Intel XENIX systems shipped in July 1982. Jimj wpg (talk) 7:48 pm, Today (UTC−5)
Most of the stuff in double-curly-braces refers to a Template; you can read about this one by searching Wikipedia for "Template:Rp" TEDickey (talk) 01:12, 7 September 2019 (UTC)
- C-Class software articles
- Mid-importance software articles
- C-Class software articles of Mid-importance
- C-Class Computing articles
- Unknown-importance Computing articles
- All Computing articles
- All Software articles
- C-Class Microsoft articles
- Mid-importance Microsoft articles
- WikiProject Microsoft articles
- Selected anniversaries (August 2004)