Hacker News Reader

Comment by systemvoltage | Hacker News Reader

How do you algorithmically code like this? In my software engineering career, I am a plumber at best. Would fail hard at anyone who would ask me to code anything more than fizz buzz under time pressure. That said, I've built some incredible production worthy and robust systems that move mountains as a plumber. I am damn good at that.

View on HN

Norvig teaches a really cool, and challenging, class on Udacity called the "Design of Computer Programs"[0] in Python where you can learn his thought process on how to write reliable and beautiful software to solve interesting problems.

[0] https://www.udacity.com/course/design-of-computer-programs--...

I'd recommend taking a look at his book Paradigms of AI Programming [0]. It really shows his thought process in developing solutions to problems. The only other way is practice, and in particular practice with languages that offer functional features. Notice his use of lambdas, assignment of existing functions/constructors to more accurate names (Passport instead of dict, day 4), and higher order functions (quantify).

Those are things that really help in algorithmic code by increasing accuracy of the names of things to the application (almost creating a domain specific language).

[0] https://github.com/norvig/paip-lisp

What kind of experience do you need for this book? For most ML you need linear algebra and whatnot, but this doesn't appear necessary here. If you've worked through something like SICP, do you think that would be sufficient?

This is more classical AI, if you have a foundation in algorithms and data structures you can follow along. More experience, it'll be easier. If you've worked through SICP and that's your primary study of programming, then you can approach PAIP, but may need to take your time and look up related material to help you out.

I am a plumber too, and a good one, but I notice (if you forgo the time pressure: my fast solutions are always plumbing) that the language I use matters in my thinking and that is probably how I learnt them. When I have the time, I think solutions up (mostly in my head or on 'paper'(scratch buffer)) in lisp/scheme, haskell/idris or apl/k. The result, when I consider it elegant, is usually very nice after translating to another language.

For Scheme I think my thinking came from working through SICP many years ago while the other languages in the list force you basically to first think about a solution before you start typing.

I am not saying, obviously, people cannot think up these things straight in rust, python, c# etc, however, for me personally, it works much better if I first work it out in one of the others; it is way too tempting to grab for my plumbing wrench otherwise.

If you're ok with throwing time pressure out of the equation, then I think the main thing is just an aspiration to find a beautiful solution to a problem, where "beautiful" is being concise, efficient, short (and, rarely, original) in some proportion that suits your fancy.

After that, it's just practice interspersed with data structure and algorithm research.

I wouldn't worry about it. Jobs that require this are few and far between - so the necessity isn't there. HN have a wideon for this leetcode stuff. In reality, unless it's your hobby, you don't need to be able.

> In reality, unless it's your hobby, you don't need to be able.

I agree with you completely here. I also feel like I fit the description of a plumber presented here. The most challenging algorithmic problems that I've had to confront in my career, which came when I was working in DSP for an RF company, were things that were tackled by the team -- a group of very talented and very experienced engineers -- and we worked on these problems collaboratively and over a period of time that is not measurable in minutes or hours.

But, having said that, isn't it also undeniable these things, these leetcode-style algorithms, are used to gatekeep entry to many of the "better" jobs in our field, despite the fact that we know that these things are not strictly necessary on the job? At the very least, this has been my experience and I read the same anecdotes on HN frequently.

Practice.

Reply on news.ycombinator.com