> especially the comments about how unscientific SWE is After trying for months to get into a CSE course at my university (they had a disagreeable policy where

> especially the comments about how unscientific SWE is After trying for months to get into a CSE course at my university (they had a disagreeable policy wherein CSE courses were almost entirely open only to CSE majors), I was surprised by this aspect. We were taught methods for proving correctness for recursion-based code, re-expressing recursion as iteration, and things like loop invariants - the latter in particular presented a more methodical way to approach programming than I had seen before. But the general reaction was almost uniformly "yeah, sure, like I'm ever going to use this". And sure, I don't think about loop invariants when I write most loops. And KISS tends to make most of my loops pretty trivial to understand. But for more algorithm-intensive work, the ideas behind these approaches have proven to be really useful tools! It's disappointing for me to see scientific approaches to SWE so quickly shrugged off. I seem to see the "cobble something together, and then patch it until nobody complains about bugs" approach to coding more often than "assemble this code in a way where it's demonstrably correct and easy to comprehend", and I'm often in the minority by strongly preferring the latter.