Cha-ching

Cha-ching
Dall-e imagines an early computerized register at Christmastime.

Whether the clink in the Salvation Army kettle, the clatter of coin at Scrooge and Marley, or the thunk of the cash drawer in the till reaching the end of its travel, Christmas is the season of coin. Sweet silver bells.

Here on the Eve, many of us have reached the end of our own travels just in time for the annual Christmas Day blog marathon, or maybe just Star Wars if you’re still discombobulated from the loss of Google Reader. Why not both? Reliable sources tell me that the wupiupi was coin of the (Huttese) realm on Tattoine. Pull on that thread and I’m sure there is a marathon’s worth of blog material on leveraged finance the Han Solo way. My own view is hold, don’t dump your portfolio at the first sign of an Imperial cruiser.

Now that our stockings are hung by the chimney with care, a moment for the humble cash register, through which so many holiday coins pass.

This is the part where I dash off to Wikipedia to learn enough about cash registers to weave a story together for you but it turns out that the early history of cash registers is boring. Paranoid employers willing to gamble that technology, basically adding machines in this case, will solve problems real and imagined with employee performance. Blah, blah, blah from about 1880 to World War Two. It turns out that the machines seem pretty boring but National Cash Register, NCR, is maybe more interesting the sum total of their products.

Charles Kettering, an automotive pioneer, executive, and inventor, left NCR in 1906. Kettering is remembered most for his invention of the ubiquitous electric automobile starter and distributor-based automobile ignition system. He is known less for his earlier work to add an electric motor to the manual works of the NCR register. T.J. Watson is probably the single person most closely associated with today’s IBM (only nerds know about Hollerith). He also got his start at NCR before getting fired and setting about to recreate NCR’s methods at IBM.

Maybe NCR machines seem boring because they became ubiquitous. They had sold their millionth machine by 1911 — the year IBM was founded. By 1922, they were selling sophisticated accounting machines into the business market and starting to build electromechanical machines long before IBM. They just kept executing.

So who did the Navy choose to build their version of the Enigma-smashing Bombe? NCR, of course. NCR’s version ran over thirty times faster than its contemporary British cousins and they were available in quantity.

Things change after the war as the race to own computing, the future, is on. NCR starts with an enormous advantage over IBM. NCR already has more installed terminals, in registers, than anyone else and the ability to build a vertical enterprise that moves business data as paper tape from registers at the point of sale directly into their own adding, accounting, and computing machines.

Everybody knows IBM. What happened to NCR? Oh, the same set of mid-century business school mindworms that invaded Hughes, Rockwell, GE, and dozens of other manufacturers-turned-holding companies. Divisions within NCR kept quietly cranking out innovative, well made, and almost totally forgettable machines. NCR started actually selling Unix machines in 1982, the year Sun Microsystems was founded. NCR sold over 100,000 0f those early machines. Sun sold 200 Sun-1 systems. Sun systems could accept SCSI disks. NCR helped invent SCSI and sold the first SCSI interface chip. Yawn.

One of the reasons NCR retains such a low profile in the minds of today’s UNIX user is that their machines were built to run their own software. If you worked at one of the stores or banks built up around NCR systems, the boring box in the back may have been thought of simply as the backend, not even as a computer. They were certainly computers, though. The GNU project supported GNU Emacs on the NCR UNIX machines from almost the very beginning of the GNU — before GCC even existed.

As for the NCR registers — they became computers too. First, they became essentially electronic versions of their previous mechanical selves. Then they became computers. Today, NCR register computers run everything from Android to CentOS to Windows.

NCR did something else to shore up their cash position. They got into the ATM business and are still major players. The death of cash could have proved an existential crisis for a company built on accepting it.

IBM, on the other hand, eventually entered the networked electronic register business when it became clear that a register hegemony, NCR or otherwise, was a threat to their business data processing market. Who won? Pick a timescale. On almost any timescale longer than a Sears register tape, neither won. Retailers are terrible customers.

I promised someone a thousand words on Point of Sale technology but my shift is almost over and I still have to go count out my drawer. I’ll leave it here and give you back the change. Would you like a receipt?

If you would like to learn more about vintage POS systems, try here.

What’s that? You received the newsletter version of this post and it’s a couple of hundred words short? You‘d like a refund? Would you consider exchanging it for this version which is exactly a thousand words? You’re very kind.

NCR was gobbled up by AT&T in 1991 and spat back out five years later as that large conglomerate broke up. By every account, AT&T had failed to leverage synergies from NCR’s experience building computers and AT&T’s experience with communications. Those synergies were largely imagined. By the time of the merger AT&T had already entered and exited the UNIX workstation business and was poised to unveil their own sleek Safari laptop. I think the merger acquisition had been defensive. The threat? The IBM phone.

Subscribe to Paper Tiger

Sign up now to get access to the library of members-only issues.
Jamie Larson
Subscribe