DevReads.sandordargo.com

Coders at work: Bernie Cosell

It was inspiring to read about some his thoughts which appear in many modern programming books.

It is partly related to this, how he was considered a great bugfixer. He tried to understand where the bug was, what that component was doing, but if it was not evident he didn’t spend too much time on it. After identifying the problematic part, he rewrote it.

He speaks a lot about cleaning up the code. When you fix a bug, you shouldn’t just fix the bug, but you should reorganize the component, you should make things the right way. He thinks that you should never ask approval for a new design in your software. Just sneak it in little by little. In fact he speaks about refactoring, a term he didn’t know.

He always overestimates compared to the easiest solution, so in the end he always has some time to improve the code.

He also speaks about premature optiomization in two contexts. He thinks that programmers are aweful in optimization. In fact they are the worst optimizers in the world. We always optimize the part of oue code that’s the most interesting to optimize, but we almost never optimize that part that actually needs optimization. We end up having unreadable, unmaintainable code without any reason.

He returns to similar thoughts when speaks about students. We already did the same as students. We see opportunities to implement “an AB unbalanced 2-3 double reverse backward pointer cube thing”, we don’t see opportunities to make the code better.

He emphasizes the importance of practicing outside of work. Writing a lot of programs is the most important things, and if we are lucky we can do it at work, but it’s not enough. The ultimate goals of practicing and delivering production are differnt.

He read from cover to cover the Big Book of D. Knuth, but he wouldn’t recommend it as tutorial to programming.

This LISPer who sonsiders himself a combination of an artist and craftsman also thinks that C is actually a problem. It’s not a problem itself, but it’s too complex and used at too many places. Few people understands it well enough. Perl is more secure and lets him to do the same stuff in many ways. He can decide how he does it, that’s why it’s artistic.

He thinks that Java is autoriatian because it takes away a lot of things. But maybe that’s a good thing. I found this particularly interesting because Uncle Bob in his Clean Architecture also writes about what different paradigms take away from you.

In the header the author mentions that Cosell moved to a farm and barely programs. I was a bit said that this aspect of his life was not elaboreated in the book because he was not at all so negative about IT as L Peter Deutsch was.

Sandor Dargo

C++ Developer, writer, speaker, traveller, happy father of two

Your subscription could not be saved. Please try again.
Thanks for subscribing. You'll receive our next monthly newsletter.

Subscribe

Get a monthly review on the best books for developers!