> You definitely don't want a 30 years old C++ book except for some sort of historical research purpose.
I think even here it depends on what you're trying to get out of it. I wouldn't read K&R to get the latest information about how to write modern C, but I read that book once every 5 years or so because there are timeless aspects at its core. This is even more true of a book like SICP.
I think even here it depends on what you're trying to get out of it. I wouldn't read K&R to get the latest information about how to write modern C, but I read that book once every 5 years or so because there are timeless aspects at its core. This is even more true of a book like SICP.