Wow! It's frightening how little people learn from history. Scheme/Lisp is much better for that use now and has been for, oh, decades. [But, then again, that's coming from a guy at that little community college on the peninsula. :-)]
It's truly sad to me that Berkeley and MIT seem to have dumped Scheme as the introductory language for the "job skills" approach of using less flexible but more popular languages.
Being able to code fast one-offs to solve problems or automate tasks is one of the strongest reasons to get into programming in the first place. This is hugely useful in engineering for example, not just in minimum wage grunt jobs. Scheme/Lisp facilitate this VERY poorly.
Python is maps far better to how problems are commonly presented, very rarely has gotchas that would make you drag out some advanced programming construct, and has batteries included at a level which scheme/lisp may NEVER catch up with:)
All of the world is not calculus! Spend all your time playing in that area and you won't become a very well rounded or capable programmer/engineer.
Indeed. That's why I'm totally an advocate for a requirement for a concurrent track that's all about installing/configuring/administration of computers/networks with an automation (i.e., shell scripting, tool development, etc.) focus.
There's a serious problem with how people learning languages... their exposure to their first language has a strong tendency to color their thinking (read: biases) towards/against other languages for the rest of their careers.
Another serious problem is that students end up with very distorted views of what "programming" is by getting only one narrow view of the how things work. Some curricula are over-focused on the abstract, "math" side of the field while others are stuck on the basic engineering side. There's not only a need for both but also the very hands-on administration/operations side precisely so that people get a much more well-rounded experience of the entire picture.
Yes, but the Harvard course’s treatment isn’t nearly as deep as MIT’s SICP course was, skipping most of the coolest parts, and they put it after a C course.
It's truly sad to me that Berkeley and MIT seem to have dumped Scheme as the introductory language for the "job skills" approach of using less flexible but more popular languages.