I was lucky in that I went through my nostalgia phase when old computers were simply obsolete. That is to say, I was comparing spinning rust to spinning rust. T

I was lucky in that I went through my nostalgia phase when old computers were simply obsolete. That is to say, I was comparing spinning rust to spinning rust. Then modern spinning rust may have been significantly faster than obsolete spinning rust, but other factors partially compensated for the difference (such as leaner software that was written much closer to the hardware). While I still maintain an interest in computer history, i.e. in a literary sense rather than maintaining hardware, it is from an entirely different perspective. I like thinking about how things would have been different if we, say, focused upon energy efficiency or if we minimized the number of layers of abstraction of if main memory was non-volatile. Of course these all involve trade-offs. The interesting thing is how these trade-offs would change the path of technological development. Of course, it is entirely possible that there was only one viable path. It is nearly impossible to sell operating systems based upon technical improvements, so the emphasis is upon a bunch of end-user visible ephemeral trash. Similarly, it would have been nearly impossible to sell more energy efficient computers so the emphasis was on performance. (Even today, a laptop battery that lasts a day and a mobile phone battery that lasts two is considered good enough.) And, of course, the sale of those high performance processors financed the development of improved fabrication processes that enabled the development of even higher performance processors - and more energy efficient ones. It is unlikely that pursuing energy efficiency would have financed the development of processes to perpetuate itself.