@provuejim Didn't the Apple II start shipping with 48 KB standard around then? Perhaps he was conflating baseline with maximum. (I think @danb is active here, he might know the story behind the quote)
@elkmovie @danb The Apple II had a 6502 processor, most other machines used an 8080 or Z80. All of these machines had 16 bit address bus for a theoretical maximum of 64k, but usually you couldn’t get all the way to 64k because of i/o cards or other obstructions in the memory space. But there certainly wasn’t ever a time when the Apple II had triple the capacity of the other microcomputers of the era. According to @danb ’s web site (link below), “I created that first PC prototype over a weekend on an Apple ][ I borrowed for the purpose from Dan Fylstra”. So I think they used the Apple II because they could get their hands on one. It was certainly possible to sell commercial software for CP/M machines at that time, I know because that’s when I published my first program, a visual text editor called PolyVUE. (As you can see, I’m still using the “vue” moniker to this day.)
@provuejim @danb Thanks - hadn't seen that piece, but that definitely makes it seem like it was a stroke of luck on Apple's part that they had that year or so of exclusivity.
@elkmovie @provuejim As I recall, our publisher, Personal Software, recommended the Apple first, because floppy drives were more available for it, he told us. We already had access to good tools for the 6502 chip, which Bob knew well, so that was OK. So Jobs was right to point out disks in his interview (see info about 0:55 in http://bricklin.com/tedxtalk.htm ).
@provuejim @elkmovie As I recall, there were 8" and 5 1/4" floppy drives. We are talking about 5 1/4". Supporting them was one thing, being able to ship lots of them was another. I remember hearing (but I can't confirm) that Apple had locked up excess production from Shugart Assoc, the main manufacturer, so TRS-80, etc., users couldn't find them in stock for many months once they became popular. In any case, they were more prevalent to Apple II's.
CP/M would be much worse than Apple/TRS80/Pet because no standard memory-mapped screen, which VisiCalc needed for the multi-window, scrolling display. I had lots of experience with that on the DEC video displays (I had influence on).
@danb @elkmovie Yes, many CP/M computers were attached to external CRT terminals. Though by 1978, the external terminals were XY addressible, so it was possible to write visual applications like VisiCalc. I know that for sure because I wrote a visual editor for CP/M that I sold thru Lifeboat Associates. In fact I had a booth at the 4th West Coast Computer Faire in May 1979, according to Wikipedia you demonstrated VisiCalc there. So I guess we were in the same room at that time! I did sell PolyVUE on 8” floppies, not 5 1/4”. But it definitely was easier to write visual software for a memory mapped display. In fact my IMSAI computer had a memory mapped VDM-1 card that I originally used to write PolyVUE, then I adapted it to work with external CRTs so it could be sold. (The VDM card was designed by Lee Felsenstein, later famous for the Osbourne.) There was a pretty decent market for CP/M software, with many other apps like WordStar, dBase, etc.