big announcement for Friday: I have acquired a datacenter that has not been touched since approximately 2002, and is FULL of old 80s/90s/2000s IBM mainframe equipment.
-
@wec OMG the first programming I ever did was on a 370 and all that JCL!
-
@wec You need to acquire a Bond Villain chair and a fluffy white cat for your mainframe lair. Just saying.
-
@wec 3745!!! i am so excited!!!
-
@wec I'm starting to think 80s computer hardware is our generation's answer to "basement full of model trains"
@WizardOfDocs @wec Something like this?
-
@wec I'm starting to think 80s computer hardware is our generation's answer to "basement full of model trains"
-
@wec OMG the first programming I ever did was on a 370 and all that JCL!
@SamanthaJaneSmith @wec Way back in the depths of time one of the folks at Interactive Systems (the first commercial Unix company, circa 1980) wrote a JCL interpreter and job processor for us by those who wanted to treat Unix as a batch processing machine. We copied the job pages and other stuff from UCLA's data center. It was quite impressive - an entire JCL deck just to copy a file, including several pages of resulting printout.
-
Just bear in mind that you can't get spare parts or maintenance for 3745s any more. (The $Dayjob quartet (split over two datacentres for resilience) were finally shut down after this moose retired; when the last remaining user "You can't shut us down, it's Safety Critical!" was advised of the running cost for their _eight_ sessions, and they would be charged for licenses, power, aircon & maintenance...) I spent 4 years trying to persuade them to switch to TCP/IP. 3:O#>
🤬 🤬 🤬 🤬
-
@SamanthaJaneSmith @wec Way back in the depths of time one of the folks at Interactive Systems (the first commercial Unix company, circa 1980) wrote a JCL interpreter and job processor for us by those who wanted to treat Unix as a batch processing machine. We copied the job pages and other stuff from UCLA's data center. It was quite impressive - an entire JCL deck just to copy a file, including several pages of resulting printout.
@karlauerbach @wec Oh yes! I can imagine although I don't know why anyone would want to do that. Although I do remember some really complex JCL for when the 370 acted as a front end for a Cray 1s... That was horrendous and generated a few forests of LP paper with stack overflow errors... I am going to have nightmares tonight!
-
buncha S/390 G3s, G5s, and original Zs. craploads of 3270 and 3174 too lol (also two 3745s and two 3172s)
@wec Wait, are you gonna outdo me on 3174s now? I thought I had cornered the market! (Had 20+ at one point, down to like 12)
-
@karlauerbach @wec Oh yes! I can imagine although I don't know why anyone would want to do that. Although I do remember some really complex JCL for when the 370 acted as a front end for a Cray 1s... That was horrendous and generated a few forests of LP paper with stack overflow errors... I am going to have nightmares tonight!
@SamanthaJaneSmith @wec We did the JCL-on-Unix thing for the fun of it - and to show people who did not comprehend time sharing.
I got really good at JCL - I was doing satellite stuff and we had compilation jobs that could take 18 to 24 hours. So I rebuilt the JCL to optimize things and got it down to roughly 6 to 8 hours. Most people at that time did not realize that for sequential files - like compiler intermediary files - tape was much faster than disk. So I had umpteen tape drives spinning away. (Each run produced a mountain of printout - about eight feet high!!)
Another yuck-thing we did was to put the Unix swap onto a DecTape. Poor tape drive, but it did work.
BTW, out at the Livermore labs we used an old CDC 7600 and an obsolete Cray 1 to manage our tape library for the newer Crays. (I ported Unix onto those machines.)
-
@SamanthaJaneSmith @wec We did the JCL-on-Unix thing for the fun of it - and to show people who did not comprehend time sharing.
I got really good at JCL - I was doing satellite stuff and we had compilation jobs that could take 18 to 24 hours. So I rebuilt the JCL to optimize things and got it down to roughly 6 to 8 hours. Most people at that time did not realize that for sequential files - like compiler intermediary files - tape was much faster than disk. So I had umpteen tape drives spinning away. (Each run produced a mountain of printout - about eight feet high!!)
Another yuck-thing we did was to put the Unix swap onto a DecTape. Poor tape drive, but it did work.
BTW, out at the Livermore labs we used an old CDC 7600 and an obsolete Cray 1 to manage our tape library for the newer Crays. (I ported Unix onto those machines.)
@karlauerbach @wec I love the idea of doing it for fun! I am now nostalgic for tape drives and i hear you about the paper piles
Ah yes I remember putting a job to run in on a Monday and getting the results in a Friday. We did one run every two weeks. So it meant you really had to check the code thoroughly!
I also had the pleasure of using a CDC6700 and 7600 at the University of London computer centre. I really liked using them despite the large card decks due to the use of update and insert cards.
You said you worked in "satellite stuff" can I ask who for. I spent my whole career in the space industry hence the interest.
-
@wec I'm starting to think 80s computer hardware is our generation's answer to "basement full of model trains"
And I would be confident that none of that amazing hardware will ask you for your age or date of birth
-
@kkarhan @WizardOfDocs @wec Connor brought some ancient stuff from the archives to TechXchange last October. He had a working S/36 that was running a banner print program. It was super popular especially with the young uns.
-
@karlauerbach @wec I love the idea of doing it for fun! I am now nostalgic for tape drives and i hear you about the paper piles
Ah yes I remember putting a job to run in on a Monday and getting the results in a Friday. We did one run every two weeks. So it meant you really had to check the code thoroughly!
I also had the pleasure of using a CDC6700 and 7600 at the University of London computer centre. I really liked using them despite the large card decks due to the use of update and insert cards.
You said you worked in "satellite stuff" can I ask who for. I spent my whole career in the space industry hence the interest.
@SamanthaJaneSmith @wec My initial satellite stuff was back circa 1971 when we built a satellite system for a US military group to monitor and track all kinds of things around the world - we essentially built the system shown in the movie War Games - all the way from the satellites, to ground stations around the world, to big war rooms. (And I can say this: watching, and participating, in a simulated nuclear attack and response played out in a Dr. Strangelove like war room is seriously spooky.)
Later (early 1990's) I worked on satellite stuff at Sun where we were trying to re-purpose a Soviet, then Russian LEO constellation to get data coverage (much like today's Starlink) over the continental US to support highly mobile computing. (We even had a bicycle on the net.) One of our team members even had the "fun" of sitting atop a flaming flying Saturn rocket.
More recently our social circle has had included several astrophysicists (with leanings towards astrobiology. One of whom helped repair the Hubble.)
