RS-232...
I have not-so-fond memories of RS-232 from those days, debugging both hardware and software could be a right pain in the proverbial.
I ended up making my own test lead set, two cables with 9 and 25 pin D-connectors at one end and a 25 pin at the other, plus a single long extension and a bought in test/breakout box. It served me well for a few years when serial printers and comms were common.
Hardware/software issues included twice developing suitable hardware and drivers to run a plotter over the same serial line as the terminal. For reasons lost in the mists of time we couldn't use the AUX (or was it PRT?) port on the DEC VT-220 for the first so had to construct a wire tap, for the second we could just set the VT to pass-through but had to tweak the drivers as one of the Calcomp plotter commands would cause the terminal to lock up.
In a later job we had a MicroVAX 2000 with 4 separate RS-485 connected serial port hubs. The hardware was second-hand (it had been the test/demo kit for the MRP system provider) so the hubs came with a single backplane. It made far more sense to run a single RS-485 connection to the more distant offices instead of large bundles of serial cables but DEC wanted a ridiculous sum for separate PSUs for each hub. I handed it over to our in-house electronics engineer who built me 3 wall boxes to act as RS-485 terminations and power supplies so the hubs could just plug in.