Just watched that. I had no idea Prince of Persia was not a hit when it was first released. Did poorly on Apple II because the market was dying, did poorly on PC because it was a conversion and retailers looked at the poor Apple II sales and assumed it would have poor sales, and then finally on the Mac it came out just as there was enough Mac sales to have customers wanting good software.
Once again it seems that releasing on the right platform at the right time is super important for success.
What does everyone think is the right platform right now (or emerging within the next couple of years), especially for an Indie title? People don't seem to play web games much anymore, Steam is flooded with junk, Mobile seems to be totally saturated, Nintendo Switch seems promising right now but it's starting to go down the same road as Steam. VR seems to be slow to get moving (and looks like it might remain a niche for several more years, at least). Are there any good options to get in at the right time like he did?
As you say, the real issue is discoverability, but I don't think the problem is the "junk", as much the large number of perfectly average and acceptably fun games released every day.
I read Mechner's journals a few years ago, and they were illuminating in how basically as an Indie developer at the time you needed not only the willpower and creativeness to make a game, but also have the connections/luck of getting it published (no Steam, no Apple Store... no Internet!) and you also had to dominate the technical aspect and know the underlying hardware, with very little information around (again, no Internet).
If you could manage all this (!) your game was going to be noticed, since there were not so many "good" games being released in general, much less masterpieces like PoP.
Nowadays, the barriers of entry are minimal, there are multi-platform engines with highly refined tools, with plenty of tutorials about pretty much everything, publishing is pretty much an automated process, that gives millions of people instant access to your game.
There was little chance a few years ago of a "good" indie title going unnoticed, nowadays it must be truly excellent to get the spotlight, and even then... it's easy to spot junk, much less to spot a great game in the middle of a sea of perfecty good ones.
I had a "76 in 1" cassette and my friend a "63 in 1" with exactly same games with some repeats. One day I used a pencil (graphite) and randomly drew a connection between some circuit line on the cassette and mine got converted in "63 in 1".
Hah, this is how it all started for me. I went a bit further and tried to reverse engineer the purpose of different pins in a NES cartridge, by grounding them one by one in a running system with a small incandescent lamp from a flashlight to avoid short circuits. I've got all kinds of crashes and artifacts which gave a hint to their purpose. Thankfully, the NES was resilient enough for me not to kill it or burn my house down.
Still, I assumed a cartridge to be a simple ROM with a plain addressing scheme, which it wasn't. Only a year after that I've read an EE book with occasional NES architecture details which all made sense in the context of my experiments.
That's why I've taught my son to program with a simple self-designed embedded system first. You can almost physically see its parts moving, no black magic involved.
Oh I'm not a professional, so this wasn't a super serious project. It's a 1801-series CPU (an old soviet single-chip LSI-11 clone, chosen because of the beautiful PDP-11 ISA) with 64KB of SRAM, an LCD from Aliexpress, an RS-232 controller, some LEDs in key points to illustrate the inner workings, a pulse-modulated speaker, and some discrete logic to glue all this together. It has no flash/ROM so you have to load the program from a PC each time. It took me ~three months to build in my spare time, including the PCB manufacturer round trip time. The biggest issue was making a modern C compiler work for this CPU, but in the end I've even got a legacy Forth implementation working (more or less). Forth is great on this ISA because one word translates precisely to one machine instruction.
We used it to drive our Christmas lights the last time, then to make some simple platform games. I was trying to make MicroPython run on this board without much success, so we'll probably have to drop it in favor of full-featured Python.
There was a lot of bugs in this thing, which made it even more fun to use.
It was awesome on my Atari ST. That feeling when you got to the end, thought you'd saved the princess and she kicked your arse because you were in a fighting stance. Utter genius.
Apple joysticks were these terrible analog things that were completely useless for anything other than flight simulator. Anything that needed gamepad-like input just used the keyboard.
The game supports both keyboard and joystick input. It's really a keyboard game, though, it's not Karate Champ or Way of Exploding Fist or games in that vein.
Huh. I beat Karateka on the C64 with a joystick. I don't think it would have occurred to me to play it with a keyboard, though, even if I was told it was better. At the time I was always baffled when action games had a keyboard option at all, when surely everyone had a perfectly good and superior joystick.
(After a little googling, it seems like the Apple II joystick was analog and kinda shitty in the way that early analog joysticks always were for some reason, so that may explain it as well. The C64's Atari-type sticks were not nearly as sturdy as one would wish, but they performed very well while they lasted.)
They were pricey, compared to your typical C64 Quickshot but also functionally different. It's just not a good fit for a fighting game. Few Apple ][ games absolutely required a joystick. Rescue Raiders is one that comes to mind, off the top of my head.
The Making of Karateka https://www.amazon.com/dp/1480297232/
The Making of Prince of Persia https://www.amazon.com/dp/1468093657/
I recommend reading them in order.