Re: IBM 7074 and then some: "Systems we love" conference
On 1/26/2017 8:37 PM, Fred Cisin wrote: Even the phrase "shipped" need not be when the customer starts printing out the nine billion names of god. That was a printout ... not card I/O. (1953, so it predates the 7074) Great Story! -- Grumpy Ol' Fred ci...@xenosoft.com
Re: recursive emulation (was: IBM 7074 and then some: "Systems we love" conference)
On 24 January 2017 at 18:46, Jecel Assumpcao Jr. wrote: > This allowed us to figure out that an 8 MHz ARM2 would be able to > run PC programs at nearly the speed of a 4.77 MHz 8088. Though there > were much faster PCs at the time, the original configuration was still > being sold and was popular enough that this would have been considered > usable. It was usable enough that Acorn sold a PC Emulator for the Acorn A305/A310 -- the original, first-series Archimedes with 8MHz ARM2 chips. E.g. http://web.archive.org/web/20021213050437/www.maxandanna.pwp.blueyonder.co.uk/pcem.htm I used it on my A310, in 1988/1989, to do real work in QuickBasic 3.0, brought home from the office. It ran at something around 2-3MHz as equivalent to an original 8088 PC, but with a much faster screen and hard disk, so overall performance was good and quite usable for MS-DOS stuff. -- Liam Proven • Profile: https://about.me/liamproven Email: lpro...@cix.co.uk • Google Mail/Talk/Plus: lpro...@gmail.com Twitter/Facebook/Flickr: lproven • Skype/LinkedIn/AIM/Yahoo: liamproven UK: +44 7939-087884 • ČR/WhatsApp/Telegram/Signal: +420 702 829 053
Re: IBM 7074 and then some: "Systems we love" conference
In the story, Chuck (no last name, so not confirmed to be OUR Chuck), had the last word: "Look" http://downlode.org/Etext/nine_billion_names_of_god.html On Thu, 26 Jan 2017, Chuck Guzis wrote: I remember reading that one many years ago. The bit about the stars quietly winking out at the end stuck in my mind.-- Nine Giganames doesn't seem so much in our days of tera- and peta- and yottawhatsis. Unless they increased commensurate with Moore's law, . . . How large is the NSA Utah Data Center?That is a LottaBytes. They should have enough computrons to get the job done in minutes.
Re: IBM 7074 and then some: "Systems we love" conference
On 01/26/2017 07:50 PM, Fred Cisin wrote: >> Even the phrase "shipped" need not be when the customer starts >> printing out the nine billion names of god. (1953, so it >> predates the 7074) > > OK, that was a "Mark V, Automatic Sequence Computer". > > In the story, Chuck (no last name, so not confirmed to be OUR > Chuck), had the last word: "Look" > > http://downlode.org/Etext/nine_billion_names_of_god.html I remember reading that one many years ago. The bit about the stars quietly winking out at the end stuck in my mind.-- Nine Giganames doesn't seem so much in our days of tera- and peta- and yottawhatsis. --Chuck
Re: IBM 7074 and then some: "Systems we love" conference
Even the phrase "shipped" need not be when the customer starts printing out the nine billion names of god. (1953, so it predates the 7074) OK, that was a "Mark V, Automatic Sequence Computer". In the story, Chuck (no last name, so not confirmed to be OUR Chuck), had the last word: "Look" http://downlode.org/Etext/nine_billion_names_of_god.html
Re: IBM 7074 and then some: "Systems we love" conference
On Thu, 26 Jan 2017, Eric Smith wrote: On Wed, Jan 25, 2017 at 8:07 PM, Chuck Guzis wrote: "Transactions of Society of Actuaries" 1959, Volume 11, Number 31 On a system like that, how much time could elapse between, "We (customer) are switching to..."? decision planning negotiating contract begin manufacturer of first bespoke component finish manufacture of last bespoke component begin shipping delivery of last awaited part begin installation complete assembly and turn on power initial diagnostic IPL diagnostics redo whatever needs redoing first IPL to actually begin customer's data processing 'twould seem that there could be a moderately substantial amount of time between the customer saying, "we are getting" V "we are now using" Even the phrase "shipped" need not be when the customer starts printing out the nine billion names of god. (1953, so it predates the 7074) imagine them getting a date wrong, but it seems pretty surprising that they would specifically claim that the 7070 was announced far before the 7090 but shipped six months after, if that wasn't true. IBM generally didn't consider a data processing system to have "shipped" until it passed the field acceptance criteria, e.g., assembled on-site and passed diagnostics. Perhaps the 7070 didn't pass acceptance testing until April 1960? Yep, That is implied, but not explicitly confirmed by the customer saying, "we are switching to..." ANY "FIRST" will run into overlaps between the ways that the "first" date can be chosen and marked for multiple products in development. -- Grumpy Ol' Fred ci...@xenosoft.com
Re: IBM 7074 and then some: "Systems we love" conference
On 01/26/2017 06:49 PM, Eric Smith wrote: > IBM generally didn't consider a data processing system to have > "shipped" until it passed the field acceptance criteria, e.g., > assembled on-site and passed diagnostics. Perhaps the 7070 didn't > pass acceptance testing until April 1960? Does IBM retain a corporate archivist on staff? That might be a resource. I do get the idea that the 7070, designed as a replacement for the 650, was a machine that IBM probably would like to forget. --Chuck
Re: IBM 7074 and then some: "Systems we love" conference
On Wed, Jan 25, 2017 at 8:07 PM, Chuck Guzis wrote: > "Transactions of Society of Actuaries" > 1959, Volume 11, Number 31 > That's certainly more convincing, but it still seems to be a bit of a mystery. The four authors of IBM's Early Computers were all engineers and engineering managers intimately involved with the development of the computers in that time frame, were researchers at IBM when the book was written, and based much of the book on material from IBM archives. I could imagine them getting a date wrong, but it seems pretty surprising that they would specifically claim that the 7070 was announced far before the 7090 but shipped six months after, if that wasn't true. IBM generally didn't consider a data processing system to have "shipped" until it passed the field acceptance criteria, e.g., assembled on-site and passed diagnostics. Perhaps the 7070 didn't pass acceptance testing until April 1960?
Re: IBM 7074 and then some: "Systems we love" conference
On 01/25/2017 05:47 PM, Eric Smith wrote: > Based on the bios of the authors, I'm actually more inclined to > believe the book than annotation of a photo on the IBM web site. Computerhistory.org mentions that the 1959 date was part of other information written on the back of the photo. Nevertheless... Okay, how about this? https://www.soa.org/Library/Research/Transactions-Of-Society-Of-Actuaries/1959/January/tsa59v11n3182.aspx "Transactions of Society of Actuaries" 1959, Volume 11, Number 31 "MR. LLOYD G. ROLLERSON stated for the past 7 years the Crown Life has been using a seriatim validation method. They are now in the process of installing an IBM, 7070 system, but will retain the seriatim system." So, how could IBM be installing a 7070 at a life insurance company in 1959 if the machine hadn't been released? I can dig up more references, if your confidence in the authors of the 1986 book is still unshaken. --Chuck
Re: IBM 7074 and then some: "Systems we love" conference
On Wed, Jan 25, 2017 at 4:00 PM, Chuck Guzis wrote: > So who you gonna believe--a photo taken in 1959 or some guy writing 27 > years later saying it didn't exist? Based on the bios of the authors, I'm actually more inclined to believe the book than annotation of a photo on the IBM web site. > > Perhaps they've got the 7070 > confused with the 7074. > Doubtful. They do mention the 7074 but don't give much detail.
Re: IBM 7074 and then some: "Systems we love" conference
On 1/25/2017 9:54 AM, Ray Arachelian wrote: On 01/23/2017 02:00 PM, Steven Maresca wrote: Just wanted to share an excerpted story just sent to me by a colleague, regarding an IBM 7074 supplying data to Java middleware, ultimately feeding a modern webapp stack: http://thenewstack.io/happens-use-java-1960-ibm-mainframe/ This comes from a conference which occurred last month titled "Systems we love" : http://thenewstack.io/systems-we-love/ http://systemswe.love/ Cheers, Steve Someone at work recently shared this: http://youtu.be/TPe6UXMDMGM - it's almost the whole thing ~ 8h long. You'll have to skip around the breaks. There is a blog which exposes the video publicly, but the video itself is marked to do limited sharing. I think it is perfect for here, but maybe share the referenced blog pages, which has an expanded index in case they want to take it down or change it. Looks like it was an interesting meeting. thanks jim
RE: IBM 7074 and then some: "Systems we love" conference
I dunno – there’s something about the sheep welcoming the 7070 that struck me funny. Sent from my Windows 10 phone From: Chuck Guzis Sent: Wednesday, January 25, 2017 4:01 PM To: General Discussion: On-Topic and Off-Topic Posts Subject: Re: IBM 7074 and then some: "Systems we love" conference On 01/25/2017 02:07 PM, Eric Smith wrote: > The 7070 was announced in Sept. 1958, but did not ship until April > 1960. According to IBM's DPD Chronology for 1959: "On August 3, DPD introduces the IBM Datacenter -- facilities in which customers rent the use of IBM 7070 systems by the hour and supply their own programmers and operators. DPD foresees a nationwide network of 25 to 30 Datacenters in major cities, with the first three located in New York City, Chicago and Los Angeles. " > The first IBM computing device to use transistors and no vacuum tubes > was the 608 calculator, shipped in December 1957. IBM's first > transistorized computers were the 7090 (36-bit scientific, > transistorized version of 709) shipped in November 1959, and the > 1401, shipped in early 1960, before the 7070. As a matter of fact, here's a 7070 on its way to an installation in Naples in 1959: https://www-03.ibm.com/ibm/history/exhibits/vintage/vintage_4506VV2070.html And I *did* specify computer, not calculator with regard to transistors. So who you gonna believe--a photo taken in 1959 or some guy writing 27 years later saying it didn't exist? Perhaps they've got the 7070 confused with the 7074. --Chuck
Re: IBM 7074 and then some: "Systems we love" conference
On 01/25/2017 01:37 PM, ben wrote: > What ever happened to computrons? Were computrons ever deployed in a real product? I was under the impression that they were stillborn. --Chuck
Re: IBM 7074 and then some: "Systems we love" conference
On 01/25/2017 02:07 PM, Eric Smith wrote: > The 7070 was announced in Sept. 1958, but did not ship until April > 1960. According to IBM's DPD Chronology for 1959: "On August 3, DPD introduces the IBM Datacenter -- facilities in which customers rent the use of IBM 7070 systems by the hour and supply their own programmers and operators. DPD foresees a nationwide network of 25 to 30 Datacenters in major cities, with the first three located in New York City, Chicago and Los Angeles. " > The first IBM computing device to use transistors and no vacuum tubes > was the 608 calculator, shipped in December 1957. IBM's first > transistorized computers were the 7090 (36-bit scientific, > transistorized version of 709) shipped in November 1959, and the > 1401, shipped in early 1960, before the 7070. As a matter of fact, here's a 7070 on its way to an installation in Naples in 1959: https://www-03.ibm.com/ibm/history/exhibits/vintage/vintage_4506VV2070.html And I *did* specify computer, not calculator with regard to transistors. So who you gonna believe--a photo taken in 1959 or some guy writing 27 years later saying it didn't exist? Perhaps they've got the 7070 confused with the 7074. --Chuck
Re: IBM 7074 and then some: "Systems we love" conference
On Jan 24, 2017 7:57 PM, "Chuck Guzis" wrote: > The 7070 was very early (1958) and is probably the first (or close to > the first) IBM transistorized computer. The 7070 was announced in Sept. 1958, but did not ship until April 1960. The first IBM computing device to use transistors and no vacuum tubes was the 608 calculator, shipped in December 1957. IBM's first transistorized computers were the 7090 (36-bit scientific, transistorized version of 709) shipped in November 1959, and the 1401, shipped in early 1960, before the 7070. Source: _IBM's Early Computers_, Bashe, Johnson, Palmer, and Pugh, 1986, MIT Press
Re: IBM 7074 and then some: "Systems we love" conference
On 1/25/2017 11:55 AM, Chuck Guzis wrote: On 01/25/2017 09:39 AM, Jon Elson wrote: Well, of course. If you look at the design of some of the last gasps of the tube generation like the Bendix G15, you will see what incredible hoops they had to jump through to make a viable product. Or, look at SAGE, which filled an enormous building with walls of tube-encrusted cabinets. Transistors and core memory really changed the landscape completely. On the other hand, computers like the LINC were quite useful with a really modest number of transistors, certainly no more than 1000 or so, while the 7070 used 30,000! Try, say, the PB-250, with 400 transistors; a 22 bit machine. There was a healthy mistrust of early transistors. Witness the one-transistor DTMF encoder used on early Touch-Tone phones or the very low transistor count in the early Dataphones. Nowadays, of course, we think of a million transistors as being modest. What ever happened to computrons? --Chuck Ben.
Re: IBM 7074 and then some: "Systems we love" conference
On 01/25/2017 09:39 AM, Jon Elson wrote: > Well, of course. If you look at the design of some of the last gasps > of the tube generation like the Bendix G15, you will see what > incredible hoops they had to jump through to make a viable product. > Or, look at SAGE, which filled an enormous building with walls of > tube-encrusted cabinets. Transistors and core memory really changed > the landscape completely. On the other hand, computers like the LINC > were quite useful with a really modest number of transistors, > certainly no more than 1000 or so, while the 7070 used 30,000! Try, say, the PB-250, with 400 transistors; a 22 bit machine. There was a healthy mistrust of early transistors. Witness the one-transistor DTMF encoder used on early Touch-Tone phones or the very low transistor count in the early Dataphones. Nowadays, of course, we think of a million transistors as being modest. --Chuck
Re: IBM 7074 and then some: "Systems we love" conference
On 01/23/2017 02:00 PM, Steven Maresca wrote: > Just wanted to share an excerpted story just sent to me by a colleague, > regarding an IBM 7074 supplying data to Java middleware, ultimately feeding > a modern webapp stack: > http://thenewstack.io/happens-use-java-1960-ibm-mainframe/ > > This comes from a conference which occurred last month titled "Systems we > love" : > http://thenewstack.io/systems-we-love/ > http://systemswe.love/ > > Cheers, > Steve > Someone at work recently shared this: http://youtu.be/TPe6UXMDMGM - it's almost the whole thing ~ 8h long. You'll have to skip around the breaks.
Re: IBM 7074 and then some: "Systems we love" conference
On 01/25/2017 12:19 AM, Chuck Guzis wrote: On 01/24/2017 10:01 PM, Jon Elson wrote: And, the 7074 was a serious computer, given the vintage. Either 4 or 6 us core cycle time was QUITE good in 1961 or so. 10 us instruction execution time was pretty decent. I find the whole period amazing. Consider that the 709 was introduced in mid 1958. It's only 6 years to the S/360. It seems that when transistor manufacture matured enough, the floodgates really opened. After 1958, nobody but developing or Warsaw pact countries even considered vacuum tube computing. Well, of course. If you look at the design of some of the last gasps of the tube generation like the Bendix G15, you will see what incredible hoops they had to jump through to make a viable product. Or, look at SAGE, which filled an enormous building with walls of tube-encrusted cabinets. Transistors and core memory really changed the landscape completely. On the other hand, computers like the LINC were quite useful with a really modest number of transistors, certainly no more than 1000 or so, while the 7070 used 30,000! Jon
Re: IBM 7074 and then some: "Systems we love" conference
On 01/24/2017 10:01 PM, Jon Elson wrote: > And, the 7074 was a serious computer, given the vintage. Either 4 or > 6 us core cycle time was QUITE good in 1961 or so. 10 us > instruction execution time was pretty decent. I find the whole period amazing. Consider that the 709 was introduced in mid 1958. It's only 6 years to the S/360. It seems that when transistor manufacture matured enough, the floodgates really opened. After 1958, nobody but developing or Warsaw pact countries even considered vacuum tube computing. --Chuck
Re: IBM 7074 and then some: "Systems we love" conference
On 01/24/2017 11:14 PM, Chuck Guzis wrote: Given that the whole SMS thing was the common denominator among the 7000 and 1400 and 1620 lines, it's hard to say who was first. But the 7070 was first delivered. --Chuck Which is damned amazing, as the 7070 was a VERY ambitious machine. 30,000 transistors, 22,000 diodes on 14,000 circuit cards. TOTALLY mind boggling. And, the 7074 was a serious computer, given the vintage. Either 4 or 6 us core cycle time was QUITE good in 1961 or so. 10 us instruction execution time was pretty decent. Jon
Re: IBM 7074 and then some: "Systems we love" conference
On 01/24/2017 07:55 PM, Jon Elson wrote: > Well, they were all built using SMS card technology, and a few > pictures seem to show very similar backplane setup. I didn't know the > 7070 was the first of that generation. While the first 7030 was > delivered after the 7070, development of the Stretch probably started > first. Yup, I was going by delivery dates. "If you can't deliver it, it doesn't exist". IIRC, Gene Amdahl and friends originally proposed a design using point-contact transistors (trying to get a uniform batch of those was "interesting"). The proposal died a couple of times, until IBM pitched what was then vaporware to LANL (then LASL). Development was started in 1956, according to WikiP. Given that the whole SMS thing was the common denominator among the 7000 and 1400 and 1620 lines, it's hard to say who was first. But the 7070 was first delivered. --Chuck
Re: IBM 7074 and then some: "Systems we love" conference
On 01/24/2017 08:57 PM, Chuck Guzis wrote: On 01/24/2017 06:37 PM, Jon Elson wrote: On 01/24/2017 12:38 PM, Chuck Guzis wrote: Was the 7070 IBM's first machine with a wire-wrapped backplane? --Chuck No, all SMS machines used similar wire wrapping. So, I think that goes back to the 7030 (Stretch) and probably 1620, 1401 and others of that time. I'm pretty sure those preceded the 70xx series. The 7070 was very early (1958) and is probably the first (or close to the first) IBM transistorized computer. The 1620 and 1401 were both 1959 introductions. First delivery of STRETCH to LANL was 1961. Well, they were all built using SMS card technology, and a few pictures seem to show very similar backplane setup. I didn't know the 7070 was the first of that generation. While the first 7030 was delivered after the 7070, development of the Stretch probably started first. Jon
Re: IBM 7074 and then some: "Systems we love" conference
On 01/24/2017 06:37 PM, Jon Elson wrote: > On 01/24/2017 12:38 PM, Chuck Guzis wrote: >> Was the 7070 IBM's first machine with a wire-wrapped backplane? >> --Chuck > No, all SMS machines used similar wire wrapping. So, I think that > goes back to the 7030 (Stretch) and probably 1620, 1401 and others of > that time. I'm pretty sure those preceded the 70xx series. The 7070 was very early (1958) and is probably the first (or close to the first) IBM transistorized computer. The 1620 and 1401 were both 1959 introductions. First delivery of STRETCH to LANL was 1961. --Chuck
Re: IBM 7074 and then some: "Systems we love" conference
On 01/24/2017 12:38 PM, Chuck Guzis wrote: Was the 7070 IBM's first machine with a wire-wrapped backplane? --Chuck No, all SMS machines used similar wire wrapping. So, I think that goes back to the 7030 (Stretch) and probably 1620, 1401 and others of that time. I'm pretty sure those preceded the 70xx series. Jon
Re: IBM 7074 and then some: "Systems we love" conference
On 01/24/2017 09:25 AM, Jon Elson wrote: > I wonder how late IBM still supported the 7074 microcode emulation? > And, of course, anybody could write a software-level emulation for > the 7074, in IBM or other hardware. One reason maybe to not run simh > on a PC is if the data comes in on old mag tapes (gasp, maybe even > 556 BPI NRZI half-inch tapes)? That would seem to me to be an insane decision also. How many 2400' 556 bpi 7-track tapes can you fit on a 1TB PC drive? Why fool with maintaining a bank of drives in that light? The biquinary coding used on the 7070 different from that of the 650. A two-out of 5 bit scheme was used (01236, with 0 being represented as 12). IIRC, the 650 used 7 bits. A word was 10 digits plus sign; a reference to 55 bit length is made, which would seem to devote a whole digit to the sign--it's not clear why this was done, as the sign appears to have only 2 values. https://www.computer.org/csdl/proceedings/afips/1959/5054/00/50540222.pdf Has a good detailed run-down. There are some interesting details; for example, although it employed a digit-sequential ALU, circuitry apparently did leading nonsignificant zero detection, so that only the number of significant digits was operated upon. Another interesting aspect was the scatter/gather tape I/O facility. One normally thinks of this as a feature on later gear; to see it in 1959 is a bit surprising. Also interesting is hardware prioritizing of I/O operations. Was the 7070 IBM's first machine with a wire-wrapped backplane? --Chuck
Re: IBM 7074 and then some: "Systems we love" conference
On 01/24/2017 12:00 AM, Chuck Guzis wrote: On 01/23/2017 09:42 PM, Jon Elson wrote: On 01/23/2017 07:45 PM, Jon Elson wrote: This blog seems to indicate that there is NO 7074, but an emulator running on 370 hardware. http://nikhilism.com/post/2016/systems-we-love/ This makes a lot more sense, some of these microcode emulations were still available of fairly late machines. Also, it probably would not be real hard to write a decent emulator for the 7074 and run it on modern hardware. Once you are using emulation, why not keep the host hardware current? Too bad the excerpts from the original talk are totally scrambled. It wouldn't surprise me if the IBM S/370 isn't being emulated as well. It wouldn't be the first time for "nested" emulation. Is there a "recursive" emulator setup wherein one machine emulates another one...where the final emulation is for the original hardware? I have no idea, although VM/370 systems tended to have a LOT of instances of OS'es and virtual machines running. I wonder how late IBM still supported the 7074 microcode emulation? And, of course, anybody could write a software-level emulation for the 7074, in IBM or other hardware. One reason maybe to not run simh on a PC is if the data comes in on old mag tapes (gasp, maybe even 556 BPI NRZI half-inch tapes)? Jon
Re: IBM 7074 and then some: "Systems we love" conference
On 01/23/2017 11:46 PM, Chuck Guzis wrote: Bob Bener has written a short squib about how the 7070 came into being: http://www.bobbemer.com/BIRTH.HTM Funny, in a tragic way. WOW! But, at the end, he says the 707x is a 6-bit machine. It seems, in fact, that the 707x was a WORD machine, not a character machine. There is a fair amount of info on it that seems to confirm this. Also, the stated speeds seem to be invariant for short vs. long (10-digit) operations, pretty much proving it was a word machine. Jon
Re: recursive emulation (was: IBM 7074 and then some: "Systems we love" conference)
> Chuck Guzis asked on Mon, 23 Jan 2017 22:00:15 -0800 >> Is there a "recursive" emulator setup wherein one machine emulates >> another one...where the final emulation is for the original hardware? An example where that could be useful is in validating an emulation. I did that recently: running an EL-X8 emulation implemented in ALGOL (by Dick Gruene) on an EL-X8 emulation in SIMH. Run the test program that comes with it both on the nested emulation, and "standalone" on the outer emulation. Compare the outputs from the two runs, and fix or justify any differences. It helped significantly, especially because the inner emulation came with a bunch of important details about strange machine quirks. paul
recursive emulation (was: IBM 7074 and then some: "Systems we love" conference)
Chuck Guzis asked on Mon, 23 Jan 2017 22:00:15 -0800 > Is there a "recursive" emulator setup wherein one machine emulates > another one...where the final emulation is for the original hardware? In 1988 I designed an ARM2 based computer (my Merlin 4, which was only built in 1992 when the ARM2 was already obsolete) and wondered if it could emulate a PC fast enough to be usable. I had written an ARM assembler and a friend did an ARM emulator running in QNX on the PC, so we wrote an 8088 emulator in ARM assembly. We then ran some simple 8088 programs on the PC emulator running on the ARM emulator running on the PC. This allowed us to figure out that an 8 MHz ARM2 would be able to run PC programs at nearly the speed of a 4.77 MHz 8088. Though there were much faster PCs at the time, the original configuration was still being sold and was popular enough that this would have been considered usable. It was not enough to convince my partner, however, which is why the machine was not finished in 1988. -- Jecel
Re: IBM 7074 and then some: "Systems we love" conference
On 01/23/2017 09:42 PM, Jon Elson wrote: > On 01/23/2017 07:45 PM, Jon Elson wrote: > > This blog seems to indicate that there is NO 7074, but an emulator > running on 370 hardware. >> http://nikhilism.com/post/2016/systems-we-love/ > > This makes a lot more sense, some of these microcode emulations were > still available of fairly late machines. Also, it probably would not > be real hard to write a decent emulator for the 7074 and run it on > modern hardware. Once you are using emulation, why not keep the host > hardware current? > > Too bad the excerpts from the original talk are totally scrambled. It wouldn't surprise me if the IBM S/370 isn't being emulated as well. It wouldn't be the first time for "nested" emulation. Is there a "recursive" emulator setup wherein one machine emulates another one...where the final emulation is for the original hardware? --Chuck
Re: IBM 7074 and then some: "Systems we love" conference
Bob Bener has written a short squib about how the 7070 came into being: http://www.bobbemer.com/BIRTH.HTM Funny, in a tragic way. --Chuck
Re: IBM 7074 and then some: "Systems we love" conference
On 01/23/2017 07:45 PM, Jon Elson wrote: This blog seems to indicate that there is NO 7074, but an emulator running on 370 hardware. http://nikhilism.com/post/2016/systems-we-love/ This makes a lot more sense, some of these microcode emulations were still available of fairly late machines. Also, it probably would not be real hard to write a decent emulator for the 7074 and run it on modern hardware. Once you are using emulation, why not keep the host hardware current? Too bad the excerpts from the original talk are totally scrambled. Jon
Re: IBM 7074 and then some: "Systems we love" conference
On 01/23/2017 09:21 PM, Jon Elson wrote: > Oh, and the picture in the article is CLEARLY a posed IBM sales > brochure photo, and not from the recent operation at the unnamed > government agency. The photo's from Wikipedia, and is a photo of the system at the Deutsches Museum in Munich, which is a place in which to spend several days wandering about. It's a marvelous collection of everything: https://www.linkedin.com/pulse/20140820132312-15475825-ibm-vintage-machinery-in-munich-deutsches-museum --Chuck
Re: IBM 7074 and then some: "Systems we love" conference
On 01/23/2017 09:04 PM, Chuck Guzis wrote: On 01/23/2017 05:45 PM, Jon Elson wrote: WOW That is QUITE amazing! And, I can't possibly imagine why anyone in their right mind would do this! Seems an emulator on a PC would be faster, and way more reliable, not to mention taking up MUCH less space, power and cooling. How reliable can a 60 year old machine possibly be? Where do you get parts? There have to be a whole lot of special parts that are deteriorating, like the plastic parts on the console. Even the PC boards (IBM SMS cards) are pretty fragile, easily damaged during rework, and some of them dissipate a lot of power, causing slow thermal degradation. Are we SURE this isn't a preview of the April 1st edition? I wondered about this too. Even the USAF eventually replaced the 7080s with S/370 running emulation. Keeping a 7074 running (if my memories of keeping a 7094 going are accureate) would be quite some task. Yes! There's all sorts of little things that would drive you nuts. Things like cooling fans, power supply capacitors, cable routing hardware (clamps, ties, etc.) that they no longer make. it would be a constant job of finding suitable replacements for unavailable parts. The 7000 series had "pages" that made up a "book" with tons of old wiring that flexed every time you opened up the pages to access the circuit cards. I just CAN'T believe somebody is actually keeping such a machine in daily service. (On the other hand, CHM does have a working 1401, that also requires folding out racks of boards to access the cards, flexing similar cables.) Wikipedia says the 7070 had 14,000 SMS circuit cards, with 30,000 transistors and 22,000 diodes. Having worked on some much more recent gear with Germanium transistors, I saw about 10% of them were bad. I didn't run that gear long once I fixed it, I sold it on eBay before any more went out. But, I can't imagine that a machine with that many components could keep running awfully long between failures. As for the 1-6 ms response time, that is totally bogus. The article is complete gibberish, talking about a vast library of mag tape and ms response time in the same sentence. Maybe the 7074 prepares data weekly for some other (newer) system that is actually connected online. And, of course, to connect anything to the 7074, you'd have to build custom hardware. RS-232 had not even been invented when the 7000-series came out. They did have a 1414 unit that apparently was some kind of comm adapter, but I'll bet it took milliseconds to send one character. Oh, and the picture in the article is CLEARLY a posed IBM sales brochure photo, and not from the recent operation at the unnamed government agency. Jon
Re: IBM 7074 and then some: "Systems we love" conference
On 01/23/2017 05:45 PM, Jon Elson wrote: > WOW That is QUITE amazing! And, I can't possibly imagine why > anyone in their right mind would do this! Seems an emulator on a PC > would be faster, and way more reliable, not to mention taking up MUCH > less space, power and cooling. How reliable can a 60 year old > machine possibly be? Where do you get parts? There have to be a > whole lot of special parts that are deteriorating, like the plastic > parts on the console. Even the PC boards (IBM SMS cards) are pretty > fragile, easily damaged during rework, and some of them dissipate a > lot of power, causing slow thermal degradation. > > Are we SURE this isn't a preview of the April 1st edition? I wondered about this too. Even the USAF eventually replaced the 7080s with S/370 running emulation. Keeping a 7074 running (if my memories of keeping a 7094 going are accureate) would be quite some task. --Chuck
Re: IBM 7074 and then some: "Systems we love" conference
On 01/23/2017 01:00 PM, Steven Maresca wrote: Just wanted to share an excerpted story just sent to me by a colleague, regarding an IBM 7074 supplying data to Java middleware, ultimately feeding a modern webapp stack: http://thenewstack.io/happens-use-java-1960-ibm-mainframe/ This comes from a conference which occurred last month titled "Systems we love" : http://thenewstack.io/systems-we-love/ http://systemswe.love/ Cheers, Steve WOW That is QUITE amazing! And, I can't possibly imagine why anyone in their right mind would do this! Seems an emulator on a PC would be faster, and way more reliable, not to mention taking up MUCH less space, power and cooling. How reliable can a 60 year old machine possibly be? Where do you get parts? There have to be a whole lot of special parts that are deteriorating, like the plastic parts on the console. Even the PC boards (IBM SMS cards) are pretty fragile, easily damaged during rework, and some of them dissipate a lot of power, causing slow thermal degradation. Are we SURE this isn't a preview of the April 1st edition? Jon
David Gelernter - was Re: IBM 7074 and then some: "Systems we love" conference
On 2017-01-23 6:52 PM, Chuck Guzis wrote: On 01/23/2017 12:25 PM, Paul Koning wrote: On Jan 23, 2017, at 2:16 PM, Chuck Guzis wrote: On 01/23/2017 11:00 AM, Steven Maresca wrote: Just wanted to share an excerpted story just sent to me by a colleague, regarding an IBM 7074 supplying data to Java middleware, ultimately feeding a modern webapp stack: http://thenewstack.io/happens-use-java-1960-ibm-mainframe/ The 7074 was referred to as a "supercomputer". Can any decimal machine really bear that title? I suppose it could. ... The 7070/74 was just a member of the 7000 line. The 7030 STRETCH and even the 7090/94 were both binary and far faster. It's just that I bridle a bit when hearing the young 'uns refer to any physically large machine as a "supercomputer". It's the same feeling that I get when I see press releases today that relate that David Gelernter single-handedly developed the parallel computation. He's not old enough; at 61, he was still in high school during the ILLIAC IV era. Gelernter's profile may have been boosted by a Scientific American special issue on Advanced Computing (1987). It featured his experimental Linda system. It's certainly I first heard of him -- I think it was this article: https://www.scientificamerican.com/article/programming-for-advanced-computing/ And stuff like this wouldn't have hurt either (found just now while googling the above citation): http://www.nytimes.com/1992/01/19/business/david-gelernter-s-romance-with-linda.html?pagewanted=all And, well... in more recent news... https://www.washingtonpost.com/news/speaking-of-science/wp/2017/01/18/david-gelernter-fiercely-anti-intellectual-computer-scientist-is-being-eyed-for-trumps-science-adviser/ --Toby Now, get off of my lawn! --Chuck
RE: IBM 7074 and then some: "Systems we love" conference
Oh, no. Of course not. Perish the thought. Sent from my Windows 10 phone From: Toby Thain Sent: Monday, January 23, 2017 12:45 PM To: cctalk@classiccmp.org Subject: Re: IBM 7074 and then some: "Systems we love" conference On 2017-01-23 5:16 PM, Chuck Guzis wrote: > On 01/23/2017 11:00 AM, Steven Maresca wrote: >> Just wanted to share an excerpted story just sent to me by a >> colleague, regarding an IBM 7074 supplying data to Java middleware, >> ultimately feeding a modern webapp stack: >> http://thenewstack.io/happens-use-java-1960-ibm-mainframe/ > > The 7074 was referred to as a "supercomputer". Can any decimal machine > really bear that title? > > The USAF used 7080s well into the 1980s--another decimal system. One of > the the reasons for doing so was a system implemented in 7080 COBOL with > miles and miles of undocumented Autocoder patches.No one person had > a full grasp of the resulting system and its nuances. Thank God this could not happen today! --T > > --Chuck > >
Re: IBM 7074 and then some: "Systems we love" conference
On 01/23/2017 12:25 PM, Paul Koning wrote: > >> On Jan 23, 2017, at 2:16 PM, Chuck Guzis wrote: >> >> On 01/23/2017 11:00 AM, Steven Maresca wrote: >>> Just wanted to share an excerpted story just sent to me by a >>> colleague, regarding an IBM 7074 supplying data to Java >>> middleware, ultimately feeding a modern webapp stack: >>> http://thenewstack.io/happens-use-java-1960-ibm-mainframe/ >> >> The 7074 was referred to as a "supercomputer". Can any decimal >> machine really bear that title? > > I suppose it could. I would apply the term to a computer that's the > fastest out there by a fair margin, and uses innovative or > distinctive bits of architecture to make it so. A CDC 6600 clearly > qualifies on that basis, as do the Cray 1 and the ILLIAC IV. I've > heard the IBM Stretch mentioned as well, I don't know it enough to > comment. It seems hard to imagine that a decimal machine could > overcome the inherent disadvantages of being decimal so successfully > that it can reach supercomputer status, but in theory I suppose it > might be possible. The 7070/74 was just a member of the 7000 line. The 7030 STRETCH and even the 7090/94 were both binary and far faster. It's just that I bridle a bit when hearing the young 'uns refer to any physically large machine as a "supercomputer". It's the same feeling that I get when I see press releases today that relate that David Gelernter single-handedly developed the parallel computation. He's not old enough; at 61, he was still in high school during the ILLIAC IV era. Now, get off of my lawn! --Chuck
Re: IBM 7074 and then some: "Systems we love" conference
On Mon, Jan 23, 2017 at 12:33 PM, Mouse wrote: > >> The 7074 was referred to as a "supercomputer". Can any decimal > >> machine really bear that title? > > I suppose it could. I would apply the term to a computer that's the > fastest$ > > Consider Babbage's Analytical Engine. It was decimal and it was, not > so much by intrinsic merit as by lack of competition, the fastest > machine of its day. > > Admittedly, it ended up being little but vapourware until modern times. > But applying "supercomputer" to vapourware machines is a longstanding > tradition (though perhaps not one dating back quite to Babbage's day). > > "The Holy Roman Empire, which was neither Holy, nor Roman, nor an Empire" -- Charles
Re: IBM 7074 and then some: "Systems we love" conference
>> The 7074 was referred to as a "supercomputer". Can any decimal >> machine really bear that title? > I suppose it could. I would apply the term to a computer that's the fastest$ Consider Babbage's Analytical Engine. It was decimal and it was, not so much by intrinsic merit as by lack of competition, the fastest machine of its day. Admittedly, it ended up being little but vapourware until modern times. But applying "supercomputer" to vapourware machines is a longstanding tradition (though perhaps not one dating back quite to Babbage's day). /~\ The ASCII Mouse \ / Ribbon Campaign X Against HTMLmo...@rodents-montreal.org / \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
Re: IBM 7074 and then some: "Systems we love" conference
> On Jan 23, 2017, at 2:16 PM, Chuck Guzis wrote: > > On 01/23/2017 11:00 AM, Steven Maresca wrote: >> Just wanted to share an excerpted story just sent to me by a >> colleague, regarding an IBM 7074 supplying data to Java middleware, >> ultimately feeding a modern webapp stack: >> http://thenewstack.io/happens-use-java-1960-ibm-mainframe/ > > The 7074 was referred to as a "supercomputer". Can any decimal machine > really bear that title? I suppose it could. I would apply the term to a computer that's the fastest out there by a fair margin, and uses innovative or distinctive bits of architecture to make it so. A CDC 6600 clearly qualifies on that basis, as do the Cray 1 and the ILLIAC IV. I've heard the IBM Stretch mentioned as well, I don't know it enough to comment. It seems hard to imagine that a decimal machine could overcome the inherent disadvantages of being decimal so successfully that it can reach supercomputer status, but in theory I suppose it might be possible. paul
Re: IBM 7074 and then some: "Systems we love" conference
On 2017-01-23 5:16 PM, Chuck Guzis wrote: On 01/23/2017 11:00 AM, Steven Maresca wrote: Just wanted to share an excerpted story just sent to me by a colleague, regarding an IBM 7074 supplying data to Java middleware, ultimately feeding a modern webapp stack: http://thenewstack.io/happens-use-java-1960-ibm-mainframe/ The 7074 was referred to as a "supercomputer". Can any decimal machine really bear that title? The USAF used 7080s well into the 1980s--another decimal system. One of the the reasons for doing so was a system implemented in 7080 COBOL with miles and miles of undocumented Autocoder patches.No one person had a full grasp of the resulting system and its nuances. Thank God this could not happen today! --T --Chuck
Re: IBM 7074 and then some: "Systems we love" conference
On 01/23/2017 11:00 AM, Steven Maresca wrote: > Just wanted to share an excerpted story just sent to me by a > colleague, regarding an IBM 7074 supplying data to Java middleware, > ultimately feeding a modern webapp stack: > http://thenewstack.io/happens-use-java-1960-ibm-mainframe/ The 7074 was referred to as a "supercomputer". Can any decimal machine really bear that title? The USAF used 7080s well into the 1980s--another decimal system. One of the the reasons for doing so was a system implemented in 7080 COBOL with miles and miles of undocumented Autocoder patches.No one person had a full grasp of the resulting system and its nuances. --Chuck
IBM 7074 and then some: "Systems we love" conference
Just wanted to share an excerpted story just sent to me by a colleague, regarding an IBM 7074 supplying data to Java middleware, ultimately feeding a modern webapp stack: http://thenewstack.io/happens-use-java-1960-ibm-mainframe/ This comes from a conference which occurred last month titled "Systems we love" : http://thenewstack.io/systems-we-love/ http://systemswe.love/ Cheers, Steve