From: shair@uiuc.edu (Bob Shair) Newsgroups: alt.folklore.computers Subject: Re: Oldest Living Computer languages: The Date: 25 Aug 1994 01:08:39 GMT mje@pookie.pass.wayne.edu (Michael J. Edelman) writes: > >Here's a few more for the IBM old timers. Please post responses: > >1. What was the *last* vacuum tube machine shipped by IBM? Don't know for sure, but it might be the 024/026 Keypunches, which were still shipping in the mid- to late-60's. >2. What was the *first* IBM machine with virtual memory? (extra credit > for first installed site) Stick my neck out on this one, and say that it was 2040-10002, System/360 Model 40 serial number 2, installed at IBM/MIT lab around 1966. If so, that system may have the most illustrious history of any System/360. When I was running on it in late 69, it was the principal development machine at the Utilities Industry development center in Des Plaines, Illinois. The product they developed on it, CICS, has found substantial usage outside of utility companies. > >2.5 What was the device that enabled virtual memory on this machine called? DAT Box?? That's what it was called on later systems. Acronym for Dynamic Address Translation. (At IBM, everything became a TLA or FLA.) > >3. What was the name of the famous virtual memory parody circulated in the early 70s? OS/VU > >4. What is ROSS, and what's it made of? Well, ROS is Read-Only Storage, another TLA. What it was made of was model- dependent. For example, on 360/30 it was made of punched cards. (Standard shape, but not standard material)... they COULD be punched in a keypunch, though. > >5. What's the difference between a green card and a yellow card? Green: S/360 Refsum Yellow: S/370 Refsum > >6. What's the quickest way to add 1 to a register in System 360? Model-dependent. LA Rn,1(Rn) was popular, as was doing subtracts with BCTR Rn,0 ... On small systems (models 30 and 40, for example) these were the fastest ways. > >7. What was the only member of the 360 line that couldn't run OS/360, and why? The 360/20 couldn't... It's instruction set lacked a few little things such as BALR, and its 8 GPR's were only 16-bits wide. Don't think the 360/44 could, either. You could either run a special Model 44 OS, or DOS, but not OS. It, too, was missing a few instructions, I think. > >8. What was the name of the special high-speed computer IBM built for a few >government sites in the 60s? Models 91 and 95. > >9. Who made the first solid-state System/360 memory, and where was the first >installed site? What size chips were used in it? No idea... Wasn't IBM. We were loyal to core storage for too long, partly because the users of IMS/360 NEEDED static memory so that, after a power failure, they could boot the utility which would write the memory buffers to the IMS log tape. > >10. Sometime in the 370 era a 'mode bit' disappeared from the 360/370 >architecture. What did this mode bit do? Ah, yes. Bit 12 of the PSW. In the 360 this bit determines whether the system will use ASCII or EBCDIC code internally. Since, as far as I know, no one ever wrote an operating system which would work in ASCII mode, it was available when the 370 came out. I believe it now indicates whether the system is in BC or EC mode. -- Bob Shair shair@uiuc.edu Open Systems Specialist SHAIR@UIUCVMD (bitnet) Champaign, Illinois 217/356-2684 Article: 43008 of alt.folklore.computers Path: tridom!emory!swrinde!cs.utexas.edu!uunet!utcsri!eecg.toronto.edu!bernecky Newsgroups: alt.folklore.computers From: bernecky@eecg.toronto.edu (Robert Bernecky) Subject: Re: Oldest Living Computer languages: The Message-ID: <1994Aug24.232337.24128@jarvis.cs.toronto.edu> Nntp-Posting-Host: algonquin.eecg.toronto.edu Organization: University of Toronto, Computer Engineering References: <33fr17$7gt@hobbes.cc.uga.edu> <33g0i3$5jh@oak.oakland.edu> Date: 25 Aug 94 03:23:37 GMT Lines: 65 In article <33g0i3$5jh@oak.oakland.edu> mje@pookie.pass.wayne.edu writes: >In article 7gt@hobbes.cc.uga.edu, mcovingt@ai.uga.edu (Michael Covington) writes: >Here's a few more for the IBM old timers. Please post responses: > >1. What was the *last* vacuum tube machine shipped by IBM? Cheap shot. I'd guess they're still shipping various sorts of CRT-based monitors. Oh, You Meant tube logic. Ah. I'd guess 650 but dunno. > >2. What was the *first* IBM machine with virtual memory? (extra credit > for first installed site) The /165 with [see below] stunt box? >2.5 What was the device that enabled virtual memory on this machine called? > >3. What was the name of the famous virtual memory parody circulated in the early 70s? I always thought it was MVS. >4. What is ROSS, and what's it made of? Do you mean "ROS" - Read-only-Storage? If so, there's a buncha kinds: CROS/ccros: capacitor ROS, TROS: transformer ROS, Funnymylarcard ROS for the 360/30 [the only machine I ever knew to fail because its vacuum cleaner motor broke]. >5. What's the difference between a green card and a yellow card? 360 vs 370 > >6. What's the quickest way to add 1 to a register in System 360? Load address: la rx,1(,rx) but rx can't be zero. >7. What was the only member of the 360 line that couldn't run OS/360, and why? I'll guess 360/44, due to lack of decimal and ss instructions. >8. What was the name of the special high-speed computer IBM built for a few >government sites in the 60s? Stretch. >9. Who made the first solid-state System/360 memory, and where was the first >installed site? What size chips were used in it? We used "Intermem" memory, although Itel[?] also made it. Intermem was run by a couple of Very Bright enginers who left IBM. Somewhere down Pookeepsie way, I believe. They also built us a multiprocessor shared high-resolution timer which was precise to a few milliseconds/year, and which sat on the 360/75's memory bus. We needed it for an MP project we did back in the dark ages, for a BIG APL system. I think it was 1k dips. The first IBM Solid State memory was, I think in the 370/145. It was ECL, and cranked out a LOT of heat for size. >10. Sometime in the 370 era a 'mode bit' disappeared from the 360/370 >architecture. What did this mode bit do? Enabled the never-did-work-with-any-IBM-operating-system ascii bit. Here's one for you: Why doesn't anyone except IBM use ICM, STCM, CLCL, and MVCL, except in very rare occasions, preferring to use sr/ic, shift/stc, clc loops, mvc loops? The last one has two answers. Bob From: elf@ee.ryerson.ca (luis fernandes) Newsgroups: alt.folklore.computers Subject: "The Third Age" Date: 22 Oct 1994 14:57:55 GMT The Sept. 17th issue of _The Economist_ has a series of articles dedicated to the state of the computer industry. The first article, titled "The Third Age" has an interesting opening paragraph: If anything ushered in computing's first age, it was the Type 6550 Magnetic Drum Calculator, a mountain of a machine unveiled in 1953 by International Business Machines. The Type 650 did not have th obvious makings of a winner. For a start, the "numerical, stored-program, data-processing machine" had about as much computing power as a modern video cassette recorder. And it cost $3,250 a month to rent-- equivalent to $18,000 in today's money. A small wonder that IBM, was cautious. After all, a few years earlier, it had thought the global market for computers to be five machines at best. Nonetheless, quietly confident of the Type 650's appeal, IBM designed it to make a profit if 50 were sold. Should the firms famously aggressive salesmen make their target of 250, IBM would make a fortune. It did. By the time the Type 650 was withdrawn in 1962, several thousand had been sold, making the primitive machine the worlds first mass produced computer. Two decades later, IBM set the computers on the road from mass production to a second age, one of ubiquity. Yet the machine with which IBM managed this upheaval-- a personal computer, first known as Project Chess, then as Project Acorn and finally, when it hit the market in 1981, as the IBM PC-- was deemed inconsequential by its maker. The third age, is a reference to the new age of computing that will be "heaven for consumers, but hell for the industry". It's nice to read an article about computers and actually learn something new; e.g. I wasn't aware of the 2 code-names for the PC. So, if you come across the issue, I think you'll find it's a good read. I usually approach any article on computing, the Internet, etc. (especially in mainstream periodicals) with skepticism, because I automatically expect the articles to be badly researched and rife with factual errors; _The Economist_ *seems* to be the exception. I should also mention that another issue (Sept. 10) dealt with the WWW and how digital-cash transactions were to be securely performed (essentially a third party would act as a broker between the buyer and seller). This issue has 2 camels having sex (or perhaps "humping" would be a better term) on the cover; the issue is about corporate mergers :-). -lf From: jones@pyrite.cs.uiowa.edu (Douglas W. Jones,201H MLH,3193350740,3193382879) Newsgroups: alt.folklore.computers Subject: Re: Operating Systems beginning Date: 17 Oct 1994 14:10:11 GMT From article , by prener@watson.ibm.com (Dan Prener): > In article <3294287.74190.6920@kcbbs.gen.nz> Richard_Plinston@kcbbs.gen.nz > (Richard Plinston) writes: > >> Operating systems started when it was noticed that many programs had >> the same set of routines included by the compiler and, to save load >> time, devised a mechanism so that these routines could be loaded once >> and then each program could be shorter by relying on the copy still >> in memory. > > That is an interesting explanation, but it does not agree with my > admittedly hazy memories, which say that operating systems started as > a way to automate the work of machine operators. In short, both theories are correct. If you look at the work done at Cambridge under Wilkes in the early 1950's, for example, it's clear that their first motivation was avoiding having each user responsible for their own I/O code -- monsterous stuff to debug, and a monsterous waste of resources when it goes awry. So they invented a crude library and crude linkage conventions, and their operating system work grew from that. If you look at the early work on the IBM 701 and 704, it looks like the first code that resembled an operating system was the automated transfer of control from loader to loaded program, and from the program back to the loader to get the next job. Both of these approaches to an embryonic operating system were followed in parallel. Doug Jones jones@cs.uiowa.edu