>>17160
>3D FPS games ultimately changed everything
>an Amiga 2000 or 3000 and have the same upgrades
Case in point as to one of the major problems with the ASIC approach. Not only were they idle in apps that hadn't been predicted yet, sometimes limitations they imposed broke those apps, as was the case with the Amiga's weird video HW making it impossible for games like Marathon or Doom to push pixels onscreen properly even with a CPU more than powerful enough to do everything itself if given room to breath. Incidentally, the present state of vidya graphics is the result of this same problem in the opposite direction. An entire generation of vidya devs grew up in the cuckcage of fixed-function Gouraud-shaded bitmap-textured triangles, their crippled ASSumptions baked so deep into middleware, that even as reprogrammable hardware was gradually reintroduced, nobody really uses it. So ideas like fractals, voxels, NURBS, and true RTRT, are civilizational lost knowledge.
>Video Toaster
Good feather in the cap of the Amiga I'll admit, especially including LightWave. But Macs had the Spigot, which also gave us Premiere and Cinepak.
>the biggest mistake they made was not manufacturing a better ASIC to replace the old chipset
IIRC, they (like the other competitors I mentioned) tried, but in-house silicon teams couldn't keep up with Motorola/Intel/etc. cranking out faster CPUs more frequently, subsidized by way larger sales volumes.
>That kind of simple interface is sufficient for a whole lotta cases
Sure, but between a TUI and a full blown REPL, there's a ton of applications TUIs simply can't do that more powerful GUIs like WIMP add feature discoverability to. Heck, I'd argue the modern reincarnation of TUIs is in fact tablet/phone/console UIs, consume-only narrowly specific UIs for the modern dumb terminal. Nowhere in there is a legitimate use case for CLI, an abomination of the '70s that never should've existed, spawning in the shadow of better ideas. I'll admit the Lisa/Mac team's Original Sin™ was failing to carry over the REPL from Smalltalk, totally fumbling important experiments like HyperCard and MacBasic, so that when dev tools did ship they were stillborn and uninspiring slop like MPW & Think C similar to the IDEs we suffer with to this day.
>>17162
IMHO, while getting in with IBM and souring the relationship with Motorola did plant a seed of corruption in Apple, the worst thing the PPC did was to the OTHER Motorola customers.
See, before RISC hit, essentially every non-Intel platform that went 32-bit chose m68k, from cheap PCs (Apple/Commodore/Atari/Sharp/etc.), to high-end workstations and minicomputers (Sun/SGI/HP/Apollo/etc.), some of whom had even been on the opposite side of the Z80/6502 divide, forming a sort of united front against Intel. All of this was supposed to carry over to the m88k RISC. Except Apple got cold feet over a slight delay, brought in IBM to combine m88k with POWER, and created PPC for the Power Macs. Not only did this further delay the RISC transition for Macs, it (combined with IMHO true rumors of Apple execs doing this as part of an abortive scheme to offer Apple for sale to IBM) sent shockwaves of confusion and fear throughout the other m68k users, who fled to make their own new RISC ISAs if they were rich enough high-margin pro platforms, or simply fucking starved to death if they were.
In order to maintain a credible defense against Intel, all of these tried to ally with each other and Microsoft on common platforms (ARC/Jazz for MIPS, PReP/CHRP/POE for PPC, AlphaPC for Alpha, etc.) that would all run each others' OSs. But because none of them seriously targeted the consumer mass market in addition to the higher margin pro market, there were no economies of scale like the IBM WinTel clones to motivate participation.
All this gave Intel time to unfuck the Pentium, which (especially the PPro) hobbled along well enough to cannibalize those platforms, and any survivors were eventually mopped up in the Itanium scam though IMHO that ultimately wouldn't have been necessary because Linux and Beowulf clusters were already having catastrophic effects.
All in all, if the m88k hadn't been fucked over, the contemporary OS/2-NT rift between MS & IBM could well have spelled doom for the WinTel platform in the '90s.
>>17163
>before Win95 totally engulfed everything
It can't be overstated what a huge leap that was for WinTel, combined with the Pentium/K5 and PCI, plus new gens of sound/graphics/network/storage adapters. It instantly brought a platform that compared badly to many 8-bit micros, up to the level of a mediocre 32-bit micro. Just go on MobyGames and compare multiplatform screenshots between the early & late '90s, IBM was just absolute rancid diarrhea before Win95.
>everything was text-oriented, much like the BBS scene
There were also nifty platform-specific protocols with native clients/servers, like FirstClass and Hotline. Sadly, everything popular enough to be useful in the normalfag world is now centralized webshit, like Discord and Google Docs.