Talk:Burroughs MCP

Copies of substantially the same article exist in several locations including Fact Archive; many of those sites cite Wikipedia as the source, so it is difficult to tell now if this is a copyvio. Someone better equipped to research the history may want to have a closer look. --Kgf0 19:08, 3 October 2005 (UTC)Reply

How is this page not wikified? Just wondering since a cursory look reveals no non-wiki markup. But then again, WTF do I know? Thirdgen 09:45, 5 January 2006 (UTC)Reply

I added a link to Category:Computing by operating system and only later noticed that it was already linked to Category:Operating system. I'm not quite sure what is supposed to be in the former Category so I'm going to leave it as it is and let someone correct me if I'm wrong. --JeffW 22:51, 13 February 2006 (UTC)Reply

I added an OS Infobox (is this a template?). I invented the term priced source for the source model because unlike closed source the source can be had for a price. I also used menu-based interface for the default UI because the default interface for the user, MARC, consists of a menu, unlike a command line interface, but it is requires the menu choices be entered as text, unlike a GUI. --JeffW 19:39, 16 February 2006 (UTC)Reply

I'm proposing to move this page to the name "MCP (Operating System)" because no one calls it the Master Control Program. I'm not sure how many people know that that's what it stands for. As I understand it, the only way to make it show up on the Category: Operating systems page as MCP is to change the name of the page. I won't move it for four days to give anyone who wants to a chance to comment (and because as a new user I can't use the move command for four days anyway).--JeffW 04:36, 14 February 2006 (UTC)Reply

Since I proposed this move I've discovered that Burroughs actually had three OSs called MCP, one each for their large, medium, and small systems. So would the names of MCP (Burroughs Large Systems), etc. be ok. Would it be better to have MCP (Burroughs Operating Systems) in the Operating System category with that page linking to MCP (Large Systems)? --JeffW 12:38, 20 February 2006 (UTC)Reply

MCP was the generic term Burroughs used for their operating systems. They usually shared the same general philiosphy but implementation and features differed considerably. The large systems MCP was the first and probably the model for the others. There were MCP's for the medium systesm B2x00, B3x00 and B4x00, another for the B1x000, one for B700 systesm followed by the CMS MCP which had implementations on the B[8|9]0, B[8|9]00 and B1[8|9]00 computers in the mid 1970's. 18:09, 24 July 2006 (UTC)

The Name is "Master Control Programm" as the manual says ;) greetz alex http://www.cs.virginia.edu/brochure/images/manuals/b5000/brochure/b5000_broch.html

I worked as a Burroughs systems programmer in the 80s and had a B6700 and B6800. At least then, and I imagine still, all of the source code for all the systems programs was available and in fact that was how you did MCP and other component patches and upgrades, i.e. recompilation. There was/is no reason to protect the sources because they only run on the B machines. -juan@acm.org

In fact, because I was quite young as Systems Programmer I printed out all of the MCP and most of the other base OS elements

Signing and correcting,just received some correspondence on it. Dunno when I posted this, but I did confirm later that Unisys still ships source, I think in the '08/09 timeframe . Lycurgus (talk) 11:01, 26 August 2015 (UTC)Reply

I'm not sure I buy the claim that MCP was the first open-source OS. In particular, IBSYS was available in open source on the IBM 7090, introduced in 1959, and it descended in part from an OS developed by SHARE members on the 709. Further, it was not common for operating systems code to be shipped in binary only form in that era; that practice did not start until the late 1970s in the IBM world, and it was hardly unique in that regard. Jay Maynard 00:59, 17 December 2006 (UTC)Reply

"Unlike IBM, which faced hardware competition after the departure of Gene Amdahl, Burroughs software was designed to run only on proprietary hardware."

I suspect, but offer no evidence, that IBM software was designed to run only on proprietary systems, i.e., IBM's S/360 architecture. Unfortunately for them, Mr. Amdahl designed compatible hardware.

If someone had bothered to build Burroughs-compatible hardware, presumably Burroughs software would be equally 'portable'.

That is, this is not a matter of software design; it is a matter of whether there were functional clones of the original hardware. — Preceding unsigned comment added by 108.20.148.80 (talk) 02:23, 4 February 2016 (UTC)Reply

Addressed this. Vt320 (talk) 18:09, 20 March 2022 (UTC)Reply

Subtopic, no need for a standalone article. Enterprisey (talk!) 09:04, 19 November 2018 (UTC)Reply

    Y Merger complete. Klbrain (talk) 20:36, 25 January 2020 (UTC)Reply

The name should be change to Burroughs/Unisys Large system MCPs.

There were also MCPs for the small system CMS (Computer Management System) range which ran on the B80/CP+, B700+, B1700+ architectures. There was another version for the medium system range B2000-B4000. This also came with editable source code, though most used the standard builds i.e 20S. All three MCPs used the same basic SPO operator commands i.e MX WY IL UL DS... — Preceding unsigned comment added by Adsjaldjjfgwetwlla (talk • contribs) 19:47, 17 July 2024 (UTC)Reply

"Unisys stopped producing the hardware in the early 2010s, and the operating system is now run under emulation." The 'Mainframe Six' reference to this seems based on an opinion on an OS2200 site. It is not a direct credible reference and perhaps a misunderstanding of emode that has existed since B5900. So the line is misleading. I propose the line is removed or clarified since it conveys the wrong message that the line might be discontinued. Ian.joyner (talk) 11:09, 25 September 2024 (UTC)Reply

The reference says "Both exited custom CPUs in the early 2010s but have fast emulators.", where "both" stands for both the line of machines running MCP and the line of machines running OS 2200. As far as I know, that claim is true, and has nothing to do with emode, at least as I understandand it from this page from Jack Allweiss' site. The "emulation" refers to binary-to-binary translation of either the B6500-descended instruction set or the Univac 1108-descended instruction set to x86-64 code; it's not like, for example, the microcode in the microengine of the B5900 implementing the emode instruction set (which I wouldn't consider "emulation"; I only consider microcode implementing some or all of an instruction set different from the native one of the microcoded machine to be emulation). I don't know how much, if any, of the MCP or OS2200 control program code is run in that fashion, or whether calls to the OS turn into calls to code running atop either Windows or Linux that provide equivalent services.

That sentence should probably be clarified to indicate that B6500-descended hardware is no longer being made, but systems - Xeon-based servers running software that emulates the instruction set and OS services - that run software for that hardware are still being built.

Additional references should probably provided; some possibilities include this article in The Register from 2012 and this article in The Register from 2013, as well as this Computerworld article.

(Speaking of emode or E-mode or however it's spelled, a discussion of the evolution of the instruction set, include emode, might be useful in the Burroughs B6x00-7x00 instruction set page.) Guy Harris (talk) 19:05, 25 September 2024 (UTC)Reply