Learning From the Old

My dad was big into computer graphics in the late 80s and early 90s. He worked at an IBM research lab - drawing cubes and spheres and manipulating matrixes and generally being pretty happy, I think.

Out of this, I got an appreciation for good animation, and my family got a whole bunch of graphics textbooks, manuals, and reference material from 1992. In conversations with my dad about graphics, he swears that reading these books would be the best way of learning any/all computer graphics material. His argument: these books are very old and so very simple, and new work is just a derivative of the old work anyways.

The material in these old textbooks usually is much simpler. But I’m not sure that really leads to more learning; being really challenged by material can be pretty cool sometimes - although this goes too far very easily. But the material is probably not always simpler. Progress can also lead to simpler techniques that accomplish the same goal - e.g. it’s a lot simpler to add numbers with algebra than with a compass and a straight-edge. And, of course, progress can be made in terms of how to explain + communicate the material.

With History and Philosophy and Literature, I think that reading the old material almost always make sense. The historical context is literally “everything”, in some ways. This seems less true with math or computers or something that is more about results than dialogue.

So, my takeaways: reading old material can be cool, but it’s not always the best thing to do. For some subjects (any one that is all about the dialogue), the old material is great.

Also, make sure to try to exploit the nice selection bias that comes along with old things. The famous books from 100 years ago are probably some of the best books that were written then (efficient markets, etc). But the same might not be true of the old textbooks that your dad just happened to keep.

P.S. I missed posting yesterday, so I’ll be writing a second post today.