@enkiv2 I do feel like this piece completely forgot about the consumer computing landscape of the 90s in its attempt to set up a dichotomy between old and modern computers.
70s and to an extent 80s computers *WERE* extremely reliable and responsive.
The 90s, though (and even the late 80s to an extent)? No.
Increasing processing demands of multitasking GUIs and applications dragged performance down to horrific levels, and stability of 90s multitasking OSes was dreadful.
machines like the PDP-8 and -11 were designed at the hardware level to be easy to use and understand, same with the 6502 and the Z80.
i'm not sure where i'm going with this, but if we had the same focus on providing hardware that was made to be programmed well, we wouldn't have such terrible software
maybe. i dunno. whatever. hahahah
@enkiv2 I mean, for things that Western consumers ran in the 90s, you basically had:
* Windows: 3.x was cooperative multitasking, 9x was legendarily unstable due to bugs and shit HW
* Mac OS: cooperative multitasking, optional and AFAIK weak memory protection
* Amiga OS: no memory protection, culture of hackish software
* RISC OS (in UK): cooperative multitasking, weak memory protection (kernel unprotected!), culture of hackish software
(Consumers didn't use OS/2 or BeOS, so...)
@enkiv2 my favorite writing on this subject is Quinn Norton's "Software is a Long Con"