"Lost" ideas [was Role of OS in filesystem (was "File types")]

Harvard architectures are not dead. If you look at just the CPU core and cache of a modern machine, you have a basically Harvard setup; it doesn't become von Neumann until you lump the CPU and cache together as the processor and go to main memory. Harvard architectures are also used in specialized applications, such as DSPs and microcontrollers. Non-binary radixes...I'm not sure I think those are better lost. Base -2 has some advantages over base 2, for example, and ternary has its uses too. Probably not for general-purpose computing, at least not now, given how far down the binary pathway we've gone - though I note that ternary hardware does sometimes get used when building binary machines (the third state being "undriven", aka "high-Z"). Part of this may be the plethora of gates; there are only 16 different two-input binary gates, and of those, two are constant (and thus not really "gates"), two simply copy one input, two more are inverters (ignoring the other input), and four more are asymmetric (inputs not interchangeable), leaving only six "real" gates, of which five are common (the sixth being NXOR). But in ternary, there are 19683 basic two-input gates (versus 16 for binary), of which there are still 729 after eliminating gates which aren't symmetric in their inputs - too many to do useful design work with; even if 90% of those don't find use (including constants, ignore-one-input, etc) - could you imagine having over seventy varieties of IC just for basic two-input gates, never mind more complicated devices? (It does make an interesting thought experiment to try to design the semantics for devices such as a flip-flap-flop (ternary analog to a flip-flop), but that's not to say such a thing would be useful. I suspect trying to analogize binary devices would be a losing game, versus letting useful devices emerge /~\ The ASCII der Mouse \ / Ribbon Campaign X Against HTML mouse at rodents.montreal.qc.ca / \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B