Can you please explain to me what does it exactly mean,
@addertooth ?
Pull up a comfortable chair, and let me tell you a personal story in which this "old saying" (mantra) came into play.
It was around 1981-1982. The job market was tough and I had relocated two-states away to find employment. I was hired by a physics lab in Tennessee. This organization did not have any Personal Computers (PC) in it. Everything was done on what people called a Mainframe computer (DEC PDP 11/8 and a DEC PDP 11/780). Every worker had a DEC VT100 text-only serial terminal as his workstation. I was their Digital and Microprocessor electronic Engineer. When anything digital needed designed or repaired, it came across my desk/bench.
This meant serial (RS-232) lines were ran hundreds of feet across the organization to connect those VT100 data terminal. They were a common victim of lightning surges. When lightning took out one of those VT100 terminals, it was my job to repair the lightning damage to the terminals. During this time we had a service contract with Digital Equipment Corporation (DEC) to provide service and repair of all the equipment, but often the delay for them to fix the equipment was intolerable to all the PHDs who worked that facility. The maintenance contract for JUST the terminals was about $700 a year for EACH for each of the Dozens of terminals. Combined, it was higher than my annual salary. A high end personal computer from IBM, fully equipped was around $5000 in this time frame.
For my section of the lab, I had built the first PC in the entire facility with third-party aftermarket "clone parts". It had cost me about $1000 at the time. In the usual Government fashion, there was a committee in charge of monitoring and managing all computer purchases. But, as I had just bought "parts", this evaded the usual clutches of this committee.
The computer committee did not track parts, just whole systems. I could order enough "parts" to build dozens of computers and it would never cause a moment of concern. Building PCs back then was a VERY new thing.
I convinced the director of the facility that it was time to get serial terminals (with their expensive maintenance contracts) off the desks of the workforce, and instead, build IBM PC clones to perform all the tasks the VT100 terminals did, PLUS run software locally at each person's desk.
It was a risky move for me. If we had ran into problems, it would be the end of my career there. And yes, the director asked: "shouldn't we buy the REAL IBM PC, instead of building "clones" from third-party vendors?". I was taking a risk by promoting this approach, but it was a significant cost savings for the organization. I felt the risk was worth it. I also felt I could address any problems/limitations which were discovered.
But, in the end it all worked out, the clone PC's not only worked well, but it re-shaped how that organization worked. The physicists and engineers were no longer as dependent upon using the mainframe to do their calculations. That, and they could do actual "graphics" (charts/images) at their desks, and not just text. After the integration was complete, a nearby University made inquiries about how we had done the switch from mainframe to desktop computers. It was explained to them, and soon they started the migration as well. It would appear the idea was infectious.
Summary: A bold idea which had real risk was taken, I could have literally specified IBM to be safe. Ultimately, we used a source which cost 1/5th as much, and it turned out great. But, if it had gone badly, I would have been fired.
Hense: "Nobody was ever fired for specifying IBM". It is about risk versus cost savings (reward).