Talk:Burroughs Corporation/Archives/2013

MCP in Tron and the Burroughs MCP

The use of the name "MCP" in Tron may indeed be very related to the Burroughs MCP. I was working in the Burrough's plant in Mission Viejo, California at the time the movie came out. At the time there were also Burroughs plants in Pasedena and Santa Barbara (all not far from Hollywood and Disney). I was told that one of the screenwriters was an ex-Burroughs employee (or a spouse of an employee?).

Don't forget the tape and disk plant in Westlake, California!

Third computer line

In addition to the two computer lines mentioned on the main page, I believe there was third. The plant in Pasedena, California produced a machine (referred to as "medium systems" in the early 1980's) that stored numbers in BCD notation, which apparently gave an advantage (at one time) in doing business calculations. This may well have originally been a company acquired Burroughs some years before.

All of this could use some verification...

-- PBannister (wikipedia login not working)

The article now mentions that (and has done so for a while, I think). Guy Harris 20:11, 23 February 2007 (UTC)

ElectroData

Could that be "ElectroData" rather than "ElectoData"?

I presume the article said "ElectoData" at one point. That appears to have been fixed at some point. Guy Harris 20:11, 23 February 2007 (UTC)

B80 series

The principles of the B1700 line (a mutable virtual machine by changing the microcode to match the high level language family currently being executed) were driven down to very small machines. The B80 (released ~1975) was a 'single operator' machine with a business role similar to what we now call a PC. The CMS operating system on this 64KByte-RAM machine was recognisably the MCP, providing multitasking and virtual memory. This was achieved by switching the microcode to run a cousin of compiled Algol/C during compiling/utilities/system calls/kernel processing, and again to run compiled COBOL/RPG for business applications. In both cases the 'machine code' run by the 'microcode' would nowadays be considered 'P-code' and 'interpreter' respectively.

[I was the CMS/B80 development manager.] Shannock9 19:36, 23 February 2007 (UTC)

You might want to add that information to the article, rather than the discussion page, or create a new Burroughs B80 page for it. Guy Harris 20:11, 23 February 2007 (UTC)

Thankyou. The problem is going to be finding any external sources to support my memories Shannock9 (talk) 20:56, 1 December 2009 (UTC)

I came here after checking out http://en.wikipedia.org/wiki/Wikipedia:Conflict_of_interest and what you just said above clarified in my mind what the problem is with the “no original research” policy.

I believe Wikipedia is squandering a golden opportunity that will never come again. There are very few individuals in their 60s/70s who have intimate knowledge of the mainframe era that preceded the PC. These individuals are numbered and their numbers are diminishing every day that goes by. I believe that the Wikipedia policy “no original research” is flawed because we as a society are missing a great opportunity to learn from people who were there in the trenches in those days and can pass this knowledge to future generations.

It is a shame to lose all this knowledge just because no one blogged about it 40 years ago. Just my $.02 Ottawahitech (talk) 01:05, 14 December 2009 (UTC)

Just in case anyone thinks this issue is reserved to outdated stuff like mainframe computers - have a look at this (way too short) wikipedia entry W. Richard Stevens. One of the most prominent writers about TCP IP and UNIX is aleady gone from this world. He died in 1999, just before Wikipedia was born, but there are still some people around who can help build Wikipedia with their own live memory. Why are we wasting precious time? Ottawahitech (talk) 14:57, 18 December 2009 (UTC)

Airline Reservations and TWA project

Burroughs, Systems Development Corporation, and TWA embarked on a project in the late 1960s to build an airline reservation system, running on a high-end Bxxxx computer. This system had two big processors, sharing a very large fixed-head hard disk drive. The application was written in ALGOL, and included development of an underlying real-time transaction processing system as well as a general-purpose database management system. By the early 70's, TWA saw the handwriting on the wall and hired Electronic Data Systems to do a study and predict the system's lifetime (performance vs. capacity). EDS concluded that the system would have a year of operation before it ran out of capacity. Burroughs had no upgrade path from that machine, and EDS recommended going with an IBM S/360 running PARS. The whole reservations operation was relocated from Rockleigh, NJ to Kansas City, and TWA's PARS system was up and running within 12 months on a S/370 Model 75 (primary) and one Model 65(backup and development). The two mainframes shared a bank of Large Core Storage and had switched access to an array of 94 disk drives. TWA filed suit against Burroughs and SDC for $40 million, and they eventually settled out of court. —Preceding unsigned comment added by 70.129.19.44 (talk) 16:42, 28 March 2008 (UTC)

Burroughs attempt at PC market early 1980s

I would say around 1981 Burroughs introduced the ET2000 line of personal computers in response to IBM’s introduction of the PC. The mind set at the time was to make a machine with better performance and capability than the competition. IBM used the Intel 8088 8-bit processor while Burroughs used the 8086 16-bit processor, IBM used a CGA graphics controller and monitor while Burroughs used the EGA graphics controller and monitor.

Fair play, you don't mean EGA you mean "much better resolution than IBM"- which wasn't very difficult.

Another innovation by Burroughs was connecting the floppy drive and later hard drive external to the main processor cabinet (housed in the monitor assemble) using a serial interface allowing storage to be concatenated or daisy chained. This is much like the USB in today’s PCs.

External serial buses were already old hat. I think HP had one for one of their calculators, and much of the early work had been done by J.D.Nicoud at ETH Zurich. Whatever management used to tell us other people were doing clever stuff as well :-)

Capability drove the ET2000 contrasting the low density 360k floppy of IBM with the 400k floppy in the Burroughs machine, likewise the 760k double density floppy of IBM was eclipsed by the 800k Burroughs floppy. These innovations proved to be the demise of the ET2000 in the PC market. Compatibility and not capability were driving the market and the proprietary architecture so prevalent in the mainframe was not the direction for the PC market. To my knowledge the only programs available for use on the ET2000 and it’s modified DOS operating system were WordStar, Visicalc and GraphWriter. ET2000’s were also used as diagnostic and boot processors for the A-Series mainframe systems.

Was that the same equipment that was sold to the banking sector as the Burroughs Modular Terminal with an extensive range of peripherals hooked together by (I think) an RS422 connection? The external connection was its downfall: the designers had specified heavy cable with multiple internal screens, but the "handicap department" had insisted on standard lightweight connectors and hoods at each end. When this was recognised as a problem instead of using decent hoods which would have supported the cable properly a lighter unscreened cable was substituted, with the result that the hardware was excessively sensitive to static electricity.

While discussing almost-PCs don't forget the B20, which was networked at the disc-controller level so that one hard disc could be shared by multiple systems (I don't know whether there was any form of locking). Also don't forget the Redactron.

Other Burroughs innovations in the computer industry were the use of LSI (large scale integrated circuits or “chips”) and VLSI (very large scale integrated circuits),

I don't believe Burroughs had a stranglehold on this. Remember that the B80 CPU was spread over multiple chips at the same time that other processors like the Z80 were in a single 40-pin package.

the concept of DLP (data link processors) separate from the main system processors that controlled peripheral devices such as disk storage, tape and printers,

I've made brief mention of the B700 as a dedicated front-end in various places. I'll try and write a decent article on it at some point.

paged memory or memory paging where memory beyond the 1M word restriction at the time could be bridged by adding additional “memory pages” of 1M words each. Another innovation was the “multiple program” processor which allowed more than one program to be run at the same time using what is called a “look ahead” prom which preconditioned machine states to allow true multi-program operation.

Perhaps you would like to expand on that. I'm having difficulty interpreting it in the context of any particular architecture.

I've started an article "Burroughs Corporation Accounting Machines" with the eventual intention of having something on the B80, L and TC series, and Sensimatic etc. I think it would also be a good place for brief mention of things like cheque sorters, in short anything that isn't one of the mainframe ranges which generally already have their own articles. I've also got a stub for "Burroughs Sensimatic"- any machine that was active for 30 years deserves an article. MarkMLl 21:11, 8 September 2007 (UTC)

[I worked for Burroughs Corp. for a little over 20 years as a systems engineer and manager. I started working with the B80, B90 and B900 before moving on to the large mainframes: B6900, A-Series A15, A17 and A19 - please address any questions or observations to benstrasser@verizon.net]

The Burroughs Sensimatic Accounting Machine

The Burroughs Sensimatic (F Series) was very popular with business customers and a mechanical wonder of it's day. From about the early 1950s until the late 1970s, it was manufactured in a number of different formats. The F50 & F100 Series had two true credit balance cross-footing accumulators. The cross-footing accumulators allowed the machine to calculate and display negative numbers. The non-credit balance down total registers could only return a reciprocal number if the total went negative. The F200 Series was a bank industry version with one credit balance cross-footer and 4 down total registers. The F300 Series had two credit balance cross-footers and 9 down total registers. The F500 Series had one credit balance cross-footer and two 9 down total registers for a total of 19 separate totals.

The 1000 Series were identical in function to the F50-F500 Series except for the addition of a Burroughs developed typewriter (writer box). The typewriter, among other things, made it possible to type the name on checks as they were posted and were very popular in payroll and accounts payable applications.

The Sensimatic 50 through F1500 Series were almost totally mechanical with an electric motor to drive the mechanism and a few internal control switches and relays. A gear driven carriage was programmed by varying length setups and stops, which controlled the direction and movement of the carriage. The non-carriage functions and totals of the Sensimatic were controlled by varying length pins that were either screwed into a magazine in the F50 to F500 Series or riveted into a similar magazine in the F1050 to F1500 Series. The superior carriage programming ability of the Sensimatic gave the machine an edge over it’s major competitor from NCR.

Toward the end of the life of the Sensimatic, a third version was introduced named the F9000 Series. These machines were basically the same as the F1000 Series except that a patent swap with IBM allowed the installation of an IBM ball style Selectric typewriter. Common lore was that the Sensimatic was so popular with business users that Burroughs had to send an engineer to Scotland to physically oversee the shutdown of the factory. Burroughs had introduced the totally electronic replacement for the Sensimatic, named the L Series, and was trying to get people to upgrade. They decided the only way to get this to happen was to stop the production of the F Series cold and dry up the supply. The market for used Sensimatics exploded afterward and many successful businesses were created to supplying this demand.

Phillip Kelley, Burroughs Field Engineer from 1965 to 1970, Founder Accounting Machines Inc. & AMI Computer Services, Miami, Florida —Preceding unsigned comment added by Flawebmaster (talk • contribs) 22:12, 26 August 2009 (UTC)

Help needed

In the section Burroughs Corporation#Developments and innovations, someone who understands this stuff needs to help out. I have made this article much more attractive to read, but this section is impervious to my efforts. Unschool 20:11, 6 September 2009 (UTC)

Beginning in the early 1960s Burroughs released the B270, B300 and B500 series computer systems. These systems were primarily used in the banking environment. Coupled with the B103 series Reader/Sorter, the systems were a dominate player in the banking check processing industry. The B300 and B500 systems incorporated 9 track tape, HPT (Head Per Track)disk, 80 column card reader and card punch peripherals. The B300/500 systems were then replaced by the 3rd generation B2000/3000 systems in the early 1970s. —Preceding unsigned comment added by Zeke1312 (talk • contribs) 19:29, 11 August 2010 (UTC)

Inaccurate Lede

Burroughs bought Univac, it wasn't assimilated. Pretty sure Burroughs had larger installed base at that point, and anyway, except for its association with Remington Rand/Sperry, Univac was never in the same class as IBM, NCR, or the like, in spite of its historical role in general purpose electronic computing. 72.228.189.184 (talk) 23:35, 27 July 2012 (UTC)

In legal and accounting terms it was a friendly merger. Burroughs bought Sperry Univac (IIRC by swapping Sperry shares for Burroughs shares). Then they held a competition to name the new entity resulting in the enlarged Burroughs being renamed as Unisys (resonant of United Systems, which name was not legally available). For both employees and customers it was culturally very much a marriage (by size Burroughs ~ 60%, Sperry Univac ~ 40%). Both companies were at a similar stage of evolution, moving from 'soup to nuts' vertical providers towards consulting companies and system integrators. Both were groping towards commercial use of a 4GL (LINC and MAPPER respectively) and these were each ported to some of the "other" Unisys architectures. Apocryphally, to emphasise the value of both heritages there was a rule that post merger your boss had to be from the other company. Shannock9 (talk) 02:28, 28 July 2012 (UTC)

I've had a go at fixing the Lead Paragraph. See what y'all think. Shannock9 (talk) 02:53, 28 July 2012 (UTC)

Restored my original title for the thread and indented your comment. Thanks for making the correction to the lede (haven't actually looked at, trust it's OK), I disagree that they were at the same level but it's picayune to contest the matter at this point, and I did work with both architectures but relatively little with univac, that being an 1108 in college. 72.228.189.184 (talk) 21:52, 28 July 2012 (UTC)