It's good we're having discussions like these. People tend to oversimplify history, so young people may end up with different ideas about how things were then t

It's good we're having discussions like these. People tend to oversimplify history, so young people may end up with different ideas about how things were then than how they actually were. It's not only important to bring lessons from the Unix workstation era of computing to the present and to the future, it's also important to realize which lessons we've already learned. Does your video game make incredibly heavy use of your GPU? Who do you think was deeply integrating GPUs in to their OSes and software in the '90s? TCP/IP networking? Where do people think that originated and matured? One lesson most people now don't know much about is the quality of well-built hardware. Most people are aware that there's mainframe hardware, which is just worlds different than what we commonly see, then there's everything else. But in the Unix workstation era, much of the hardware was build with a different set of priorities - cost-cutting really wasn't a thing for most of the hardware, which is why VAXen and Alpha machines are often still running today. (A lot of Sun and some SGI hardware, though, was built like personal computers of the time, and have a higher failure / repair rate.) I wish it was possible to pay more for hardware and get components that are just made measurably better than common hardware. If you've ever seen the traces on a VAXStation or an AlphaServer, you'd know what I mean. We can, sorta - we can buy faster memory and run it slower, we can manually overprovision our SSDs, et cetera, but it's just not part of the market (and x86 servers have a very high failure rate, so I'm not talking about them). I still have my AlphaServer DS25 colocated, running public services, compiling NetBSD pkgsrc packages, et cetera. The hardware is incredibly high quality.