And SCCS was way easier than anything you could do with cards.
My first experience with PWB was maintaining the source code and QA for a project that was
targeted at RSX-11M.
Amusingly, it as my job twice at different facilities (BRL and Rutgers) to get rid of all
the card processing equipment.
I finally offered to move the Cyber RJE station and its accompanying keypunch machine into
the office of the last person using it. He finally decided he'd make the jump to
timesharing.
Amusingly, nobody figured out how to make CyberRJE work from a PDP-11. The labs had
purchased a half dozen PDP-11/34s with a optical card reader, line printer, and graphical
display (Vector General) and a DQ-11 and 56K (them was fast in those days) short haul
modem.
Of course, in Mike Muuss's typical style, he said we could use them. They all got
recycled into graphics workstations. Mike wrote a driver for the Vector General and the
first ersatz "BRL CAD" package. We still used the card reader to read in old
"COMGEO" graphic model decks which CAD could edit.
UNIX for the 11/34 (a hybrid V6 kernel), we installed overlays. When TCP/IP came around
which needed another segment register for mbufs, we just couldn't make it work
anymore on a non-split-I/D machine. The VGs got moved over to the 11/70's and I
recycled the 34's into internet routers. We actually used the DQ's and modems
for internal communciations for a while until we subsequently bought IMPs and later
Proteon rings.