3 ms·
I think this what people really are looking to stop: "Tim just got a new phone. Tim needs his IT friend/family member to help him add wifi to his phone. Howeve
by tsumnia 10y ago
I think this what people really are looking to stop:
"Tim just got a new phone. Tim needs his IT friend/family member to help him add wifi to his phone. However, by help, Tim hands the friend/family member the phone and now it works!"
Like others have said, the layman wants "computer science", but really wants "computer literacy". Phones, software, technology changes so fast that current teaching methods become outdated in years (I'm looking at you Intro to Computers classes). As computer literate users, we get you go to "Settings" and poke around til you see something like "Network" and then something like "Wireless" or "WiFi" and are set. However, would I do this same behavior with a car? Pop the hood and poke around?
Computer Science seems like the route to people because of the side benefit of "figuring it out on your own" comes naturally after 3 weeks of programming. Don't understand something? You have to pool every resource you have (instructor notes, slides, books, websites, etc.) to 'get' it. And even then its not an 'AHA' moment. Its a pray and 'whew' when whatever you did works.
But somewhere along the way, most programmers pick up that 'it', that computer savvy-ness that people are looking for. They want to stop being dependent on IT people because all we do is huff and puff about how simple a task is (regardless of program, if someone asked you how to bold text, how would you get them to bold the text WITHOUT pointing at the screen?).
Personally, it doesn't have to be "learn computer science", but it needs to be a complete redesign to the "Intro to Computers" style class.
- DonHopkins 10y agoThe fad when I was taking CS 101 in 1985 was "program verification", so the first thing they taught all the freshmen was "programming calculus" [1], to write code in a stripped down "CF Pascal" dialect, along with proofs of its correctness. I suspect the real reason was to weed out as many people as possible from perusing a degree in CS, because at the time everybody was signing up for CS because they thought it was the ticket to make lots of money. If you didn't already have practical programming experience, you were fucked, because you had no frame of reference for what the proofs they wanted you to write were about. If you already knew how to program, it was frustrating and useless, because you already had an intuitive understanding of the concepts they were trying to teach, but you'd never actually use those "programming calculus" proof techniques in real life -- the scope of the proofs was microscopically small, and couldn't scale to real-world non-trivial problems. https://www.cs.umd.edu/~basili/publications/proceedings/P39.pdf https://www.cs.umd.edu/~basili/publications/proceedings/P39....
- collyw 10y agoAs a software engineer, I rarely use the algorithmic computer science stuff. It is useful to know (and it is a requirement for my job occasionally), but the only times I have ever had to implement a sort routine were for interview questions or exams.