I had been interested in how slow signals looked better on a meter while fast signals looked better on a scope. On an early visit to Tektronix I suggested to Cal Diller that a slow scrolling display would make a scope a good meter. This was judged a good idea but too expensive given the cost of memory at the time.
Paul Barina was exploring 6800 microprocessor programming as an engineer at CTS microelectronics. I wrote a scrolling acquisition application that would display on Paul's oscilloscope.
The effect was disappointing even for band-limited signals because without line drawing it became hard to see subsequent dots as connected.
I modified my program to acquire and display four dots in every column of Paul's display. It looked better.
I remembered this experience when I eventually came to work for Tektronix and found myself practically discussing digital storage and scrolling displays. I'm thinking, what would Paul's display look like if I put 400 dots in every column?
Tektronix 4015 Direct View Storage Tube Terminal.
I suggested that a more general solution would be to acquire both the maximal and minimal sample in any low-speed sampling interval. This would preserve the acquisition units bandwidth even when acquiring at slow speeds expected of a scrolling display.
I called this MIN/MAX. I wrote simulators for what displays might look like in APL and rendered them using APL's Tek Graphic plotting libraries. APL was useful for modeling all kinds of non-band-limited signals that one might want to view in classic oscilloscope applications.
An AM modulated carrier was easily rendered on an analog oscilloscope where the full bandwidth of the instrument was available even when the sweep was at audio frequencies.
One classic example was an AM modulated carrier. With an acquisition sampling rate appropriate for the modulation signal I would capture carrier extremes as mins and maxes which I would then render as straight vertical lines joining them.
This always looked good when the carrier frequency was many times the highest modulation frequency. However, I manufactured many borderline signals and showed that the instrument's results would still be interpretable (and useful) by someone who understood the nature of the acquisition
Another classic signal was narrow pulses riding on a low frequency signal. Digital acquisition people were starting to capture these signals and calling the process "glitch" detection.
One could not program freely in APL without a keyboard with function and operator symbols on the key caps. wikipedia 
A technician who had worked in our terminal division noticed that I spent considerable time translating APL keystrokes to the 4014 keyboard layout. He volunteered to replace the keyboard and character ROM with those made for the 4015 APL terminal. This was awesome. APL was a functional language with many special functions and operators originally typed on custom made IBM Selectric terminals. Tektronix was the only other manufacturer to offer this functional programming affordance.
It is hard to describe what it felt like programming in APL. Here, for example, is Conway's game of Life coded in a dialect of APL now representable through the wonders of unicode. wikipedia 
life←{↑1 ⍵∨.∧3 4=+/,¯1 0 1∘.⊖¯1 0 1∘.⌽⊂⍵}
One would stream array and matrix object through sequences of functions much like one streams text through program pipelines in the unix shell. Of course one had to learn the functions. Without knowing them one could only marvel at the apparent complexity.
.
History of APL including the decade before Tektronix announced its product in 1973. Tektronix then devoted all 16 pages of its marketing communication publication to explaining why the company was all in on APL.
pdf