Steve, Being old school (assembler, cobol, rexx, vm exec/exec2 DB2, hardware and others) and currently retired, I am wondering how much our skills are still worth. Documentation of our old code was primarily done by commenting the code as you went along no matter what the language was. As long as those reading the comments were competent we never gave it a second thought. Today, that lack of competence could be a real issue, so I totally agree with your thinking. Is it worth throwing my hat in ring? 💍 Just curious about what our legacy knowledge is or might still be worth. Best Regards, Doug .
On Aug 15, 2025, at 19:59, Steve Thompson <ste...@wkyr.net> wrote: I whole heartedly agree. But I think, somehow, IBM is still gun shy because of DOJ anti-trust actions they've had to deal with. Anyhow, having done capacity planning with z/VM and linux images.... I look at this going off to cloud differently. It is a big cost that only gets larger. So we all retire and the new kids get to figure out all this code we wrote where companies allowed documentation to get trashed, not realizing that they needed to capture it... I say this because I'm being contacted now to help companies document all their obsolete COBOL code, and missing source... You would have thought they would have learned from the Y2K problems. Hey Steve Beaver, what do you think someone is worth that can read all that code and run book doc and explain all the interfaces that DON'T exist outside of z/OS?-- Regards, Steve Thompson Make Mainframes Great Again They use far less Electricity than Clouds and can do more work On 8/15/2025 6:38 PM, Mike Schwab wrote: > This is a marketing opportunity. IBM should contact Google. IBM should > suggest a parallel test with returned hardware (z16, z15, z14). Convert > their stack to run in a Linux on z image, then duplicate a portion of > workload to the z?15? with limited processors and see how they run. > Gradually increase the workload until the z has delays the up the number of > z processors. Do workload balancing, shifting between images when workload > changes, etc, all while documenting workload, power use, costs on newer > models, etc. > > On Fri, Aug 15, 2025 at 4:49 PM Steve Thompson <ste...@wkyr.net> wrote: > >> Well, I got connected to AI, and got into a bit of a chat. >> >> Bottom line, all things considered, cloud will cost more over >> time than using mainframes and merging linux workloads under z/VM. >> >> But sales people revenue off of modernization off to a cloud and >> not getting all the speed and processing a z/Arch machine can do >> on the MVS side and the Unix System Services side. >> >> Things go in cycles. Wonder if I will live long enough to migrate >> back to z/Arch or whatever it will be called with the next >> generation... >> >> -- >> >> Regards, >> Steve Thompson >> Make Mainframes Great Again >> They use far less Electricity than Clouds and can do more work >> >> >> On 8/15/2025 2:17 PM, Steve Thompson wrote: >>> I don't have pilot.... What answer would one get to a question >>> of which is faster (second question more efficient) a 4 CORE >>> (pick the CPU CHIP) or a z/1x with 4 CPs? >>> >>> I just wonder what answer one would get.... >>> >>> >>> -- Regards, >>> Steve Thompson >>> >>> Make Mainframes Great Again >>> >>> >>> On 8/15/2025 1:15 PM, Jon Perryman wrote: >>>> On Fri, 15 Aug 2025 13:55:35 +0000, Seymour J Metz >>>> <sme...@gmu.edu> wrote: >>>> >>>>> So apparently an incorrect answer is better than "I'm sorry >>>>> Dave, I don't know that." >>>> If AI correct answers was a standard, "I'm sorry Dave" would >>>> be the most common answer. Better an incorrect answer than no >>>> answer at all. >>>> >>>> I asked copilot "which cpu has the fastest single core >>>> speed?". It did not answer IBM TELUM where the instruction set >>>> and IBM boasts each performs 10X better than any other CPU >>>> core while running Linux. >>>> >>>> I asked copilot "is el capitan supercomputer a server farm?" >>>> and it replied, "not exactly". Copiliot ignores that el >>>> capitan is built using 5,000 blade servers, each running Linux >>>> that boasts a network speed of 12tb/sec. "server" is literally >>>> in the name. >>>> >>>> ---------------------------------------------------------------------- >>>> >>>> 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 >> > ---------------------------------------------------------------------- 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
