It's good of him to remind us not to live in the clouds, I guess, but I think he's hitting a bit of a straw man here. I just finished teaching a CS2 course whe

It's good of him to remind us not to live in the clouds, I guess, but I think he's hitting a bit of a straw man here. I just finished teaching a CS2 course where a recurring theme is that big-O complexity isn't and can't be the only thing you consider when choosing an algorithm. My understanding is that this is typical---that this idea is usually covered somewhere in the first couple terms of a standard CS major. Admittedly, I didn't talk specifically about memory paging (since they won't really hit that until their OS/Networking class next year), but it's a fairly natural extension of the tradeoffs I did train them to consider. Put another way, if presented with this article, I'd expect my students not to say "ZOMG what have I been doing??" but rather "Ah, another thing to look out for!" And I suspect this is typical of modern CS2 courses, not to mention more advanced algorithms courses.