And even when they weren't supported you had the paddle project. BTW, by 1978 there was 3rd party memory on the market. There was even block multiplexor channel for the 360/65, from a company called CIG.
-- Shmuel (Seymour J.) Metz http://mason.gmu.edu/~smetz3 ________________________________________ From: IBM Mainframe Discussion List <IBM-MAIN@LISTSERV.UA.EDU> on behalf of Joel C. Ewing <jcew...@acm.org> Sent: Thursday, January 17, 2019 1:32 PM To: IBM-MAIN@LISTSERV.UA.EDU Subject: Re: Where's the fire? | Computerworld Shark Tank The company I went to work for in 1978 was quite successfully running their corporate workload and that of several other businesses they serviced on two 360/65's with only 1 MiB of memory each and about 3 GB of shared DASD, using IBM 3277 terminals locally and remote and many IBM 2741 terminals at hundreds of locations across the nation over communication lines managed through two IBM S/7's that were programmed to serve as a front end for the input and output queues for the remote devices. They had developed an in-house multi-tasking system that ran all the interactive workload out of a single partition under Itel DOS/VS much more efficiently than anything available from IBM. One 360/65 was used for production, the other for test and program development, and all programming-support 3277's could be physically switched between controllers on the two systems. It's amazing how much could be done with so little when you have local necessity and don't have to deal with images or even colored text. It wasn't until 1980 with a transition to IBM 4341's and DOS/VSE that we had processors with more than 1 MiB of memory, and a gradual transition to CICS and VTAM then made larger memory and virtual memory a must. Eventually corporate expansion through mergers increased the workload to the point that a transition to larger processors and MVS/XA was forced around 1985-1987, because of virtual storage constraints in DOS/VSE and constraints on the maximum number of coupled DOS/VSE systems. I learned while advocating and planning our transition to MVS in the mid-80's that several years before my time there had been an abortive attempt to migrate to MVS by migrating from S/360 to S/370 and using VM as a platform to migrate from DOS to MVS, but the only S/370 system they were able to obtain was badly under-configured for the task, with insufficient real memory to even successfully run the existing DOS workload under VM in real time, much less handle migration. Apparently at that time IBM's ability to manufacture S/370 memory couldn't keep up with the demand. While the migration could surely have been made to work, sufficient memory to upgrade the S/370 was not available from IBM in an acceptable time frame, and the only way to get the corporation DP functional again quickly was an offer through Itel of the two 360/65 systems with 1MiB each of Itel memory. That turned out to be a cost-effective solution for a number of years. I feel the understanding of "obsolete" for mainframes has changed from the old days, not to mention that a corporate for-profit environment is a very different beast than a well-funded scientific center. My recollection is that old hardware and Operating Systems were supported in some fashion for much longer periods then than they are now. The pace of change in both hardware and software had definitely accelerated by the 1990's: a revolution in cheaper,faster DASD alternatives followed the next decade by a revolution in cheaper, faster processor alternatives once CMOS processor speeds reached acceptable levels. Over the last several decades, failing to stay reasonably current with hardware and software has been more likely to put a company in a serious bind if some dramatic hardware or software announcement provides a much more cost-effective DP solution (or a drop in support demands a change), but the migration path is expensive because the DP center is too back-level. Joel C. Ewing On 1/16/19 4:36 PM, Lester, Bob wrote: > Hmmmm. I worked on a 360/75J in 1979-1980. We had 1Mib "high-speed" RAM, > and 2Mib of LCS storage. > > BobL > > -----Original Message----- > From: IBM Mainframe Discussion List [mailto:IBM-MAIN@LISTSERV.UA.EDU] On > Behalf Of Seymour J Metz > Sent: Wednesday, January 16, 2019 1:41 PM > To: IBM-MAIN@LISTSERV.UA.EDU > Subject: Re: Where's the fire? | Computerworld Shark Tank [ EXTERNAL ] > > "It's the late 1970s, and this data center has a high-performance IBM 360/75 > mainframe that sports a massive 1 MB of core memory -- one of only four in > the world," > > I might believe late 1960s, but by the late 1970s the 360/75 was well and > truly obsolete and 1 MiB was nothing to brag about. We had 2 Mib at the > Technion in 1973, and I'm sure that places like NASA Goddard had more, to say > nothing of the tri-ASP 195 complex at Suiteland and the intelligence > facilities in Maryland and Virginia. > > > -- > Shmuel (Seymour J.) Metz > ... -- Joel C. Ewing ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
