I needed a little mental relaxation this weekend, so I spent a while playing mainframe model trains by bringing up Hercules.

I initially tried the MVS Turnkey system, but ran into some issues — mainly that there are no working (free) 3270 emulators for Big Sur. Since I couldn’t set up any consoles, and Homebrew x3270 didn’t seem to work under Xquartz, and I had no intention of spending $29 just to fool around with MVS for a bit, or multiple hours trying to get X11 builds working, I dropped back to Jay Maynard’s MVT installation instructions.

They’re a bit out of date at the moment, and the Right Thing would probably be to make the fixes in both the instructions and the files, and move the corrected instructions over to a wiki somewhere. For now I’m leaving my notes here so I don’t forget what I did, and so I can do that later if I get the time. More fooling with the console and running jobs, less file twiddling.

• You can get the OS/360 “CD-ROM” at http://www.jaymoseley.com/hercules/downloads/archives/os360mvt.tar.gz — this really ought to be on archive.org. It works with Maynard’s instructions and has these fixes:
  • dlibs/DN554.XMI was recovered. The srclibs/dn554 and related files in srclibs/TAPEFILE.ZIP have not been modified (recovered) to match, and should be at sone point.
  • srclib/fo520/IEYUNF.txt has been renamed to original.IEYUNF.txt and a version recovered from the MTS distribution has been added as IEYUNF.txt. The related file in srclibs/TAPEFILE.ZIP needs to be fixed as well.
  • These files are fine as they are to build MVT.
• You need the JCL and HASP II tapes; they’re at http://www.conmicro.com/hercos360/os360ctl.tar.gz. Yes, I wish it was HASP IV, but I’m playing, so it’s not that big of a deal.
• Maynard’s instructions are definitely of their time, when cutting and pasting commands was not a thing. The relevant commands are mentioned once, and one is expected to remember them. If I update these, I’ll inline the relevant command and note which console they get typed into. Switching back and forth between the Hercules “hardware” and the MVT console was a tad confusing at times. A significant omission: the devinit 00c foo.jcl command works okay for the MFT starter system and MVT without HASP, but once HASP is running the devinit must include eof at the end of the command or the reader hangs and the job never starts. Also, one of the HASP job filenames is called out by its right name in the section header, but by the wrong one in the text. Looks like a cut and paste error.
• You should comment out or remove the 3270 definition at 0C0 in the mvt.cnf Hercules config file; if it’s there when you boot MVT, but no 3270 emulator is connected to it, the machine will hang with wait state 21. Took some googling to find that. TCAM will grumble about it when it starts up, but it doesn’t hurt anything.
• You will need to install telnet on your Mac, since Big Sur removes it. telnet localhost 3270 to connect the virtual 3215 console. It might be worth trying to configure some 3215s for TSO and see if that works. We don’t really need to emulate real serial terminals.
• Doing the mn jobnames, t and mn status commands will save you a lot of wondering whether anything is going on or not, even after HASP is up.
• Be sure to copy the prt00e.txt virtual printer text file when:
  • You’ve finished doing both stages of the sysgen on the MFT system.
  • You’ve finished installing HASP.
  • You’ve finished installing TCAM and TSO

Otherwise you lose all those useful assembler outputs of the HASP hooks and the TSO interfaces to it. The sysgen and TCAM build output isn’t critical, but it’s nice to have.

Other things — http://hercstudio.sourceforge.net/ is supposed to be a hardware console emulator (lights and dials and stuff) interface for OS X and Linux; it’s written in Qt, which is fine, but the Makefile that qmake builds works to build it, but wants to install things in /usr/bin and Big Sur will not let it go there even with sudo. I might consider just writing a native iOS one instead.

I’m up a couple hours later than I intended, but I have my notes, and as Adam Savage says, “the difference between science and screwing around is writing it down!”