3 ms·
I believe as you do and in basically the same terms, but, having had forms of this discussion for much of my life, I've encountered a counterargument that I tho
by schoen 10y ago
I believe as you do and in basically the same terms, but, having had forms of this discussion for much of my life, I've encountered a counterargument that I thought I'd bring up as a devil's advocate:
> the way computers actually work, with functions and parameters instead of icons and settings menus
Isn't "the way computers actually work" at a fundamental level about electrical engineering, or at least about machine instructions? If you're using Unix commands, hasn't a whole lot of work already been done to create the abstractions of pipes, file descriptors, files, filesystems, and so on? Is it clear that there is some sense in which these abstractions are faithful and true to the machine and technology, where other software abstractions are not?
- schoen 10y agoI got a ton of upvotes on this, but nobody has replied to it! I wonder how many people who upvoted me noticed that I claimed to be expressing an argument against my own opinion. :-)
- exolymph 10y agoI noticed, but I'm sort of mystified about why you wouldn't agree with that opinion. Clearly the command line is an order of magnitude more abstract than practical electrical engineering. We need different people to understand different layers, and as long as there are enough people paying attention to each layer, there's no problem.
- jbermudes 10y ago> Isn't "the way computers actually work" at a fundamental level about electrical engineering, or at least about machine instructions? I agree, which is why sometimes I think perhaps a toy assembly language might be a neat experiment from a pedagogical point of view.