Reddit reviews Elements of Programming

Folks interested in this should really look at Stepanov's EoP, which has the algorithms we actually use.

For example our implementation use the triple reverse version for random access iterators.

Work has been pants-on-fire-mode for the past year+. I'm pretty burnt out, don't have any desire to study after work and feel like someone will get on my case if I do any relevant studying during work hours. I really don't understand our products that well either (RF), but doing that sort of studying at work would definitely be seen as wasting time since I'm in software now.

For the specifics, I'm going to crack open several of my core textbooks from college and thoroughly work through them. At a high level I understand the concepts, but I haven't internalized a lot which slows me down when I need to learn something more advanced on the job. I've tried this type of study for a week previously when I was motivated at home and it was an incredible boost to my job performance. However, that was during a slow period. Additionally, if I'm going all out and quitting my job I expect that I would have relevant projects so studies wouldn't go in one ear and out the other.

Edit: I'm just rambling at this point, but
> Why can't you study the things you need for work as you need them?

Have you ever found that you can't use what you don't know exists? Some books (Elements of Programming) really blew my mind open, but I would have had no idea how useful the contents would have been before going through them.

Start with a good algorithms book like Introduction to algorithms. You'll also want a good discrete math text. Concrete Mathematics is one that I like, but there are several great alternatives. If you are learning new math, pick up The Princeton Companion To Mathematics, which is a great reference to have around if you find yourself with a gap in your knowledge. Not a seminal text in theoretical CS, but certain to expand your mind, is Purely functional data structures.

On the practice side, pick up a copy of The C programming language. Not only is K&R a classic text, and a great read, it really set the tone for the way that programming has been taught and learned ever since. I also highly recommend Elements of Programming.

Also, since you mention Papadimitriou, take a look at Logicomix.

If true, Alex's retirement would be sad, a big loss for the community. Here are a few pointers to his recent work:

• http://www.amazon.com/Elements-Programming-Alexander-Stepanov/dp/032163537X
  
• http://www.amazon.com/Mathematics-Generic-Programming-Alexander-Stepanov/dp/0321942043
  
  Also, his A9 lectures are linked from here: http://www.stepanovpapers.com/

Not only has your statement killed Alexander Stepanov, it has started him rolling in his grave. Clearly you have never had to write serious algorithms, and only had to deal with high-level coding. So, impromptu interview:

1. Please write a parametric sorting algorithm that operates on subsequences of a container whose worst-case performance is O(n log n), with a low constant factor, and whose best case performance is single-pass O(n);
   
2. Please write a naive fourier transform that allows you to select frequency domains; write its inverse that operates on subsets of the texture or volume;
   
3. Please write a framework for a compiler to operate on subregions of straight-line code for peephole optimizations; expand this to subgraphs; generalize the notion of an iterator to a coordinate; and,
   
4. Please write an algorithm for discovering which elements of the intersection of three subsequences of an array, specified by the user, contain non-null-elements.
   
   Let me comment on these, lest you think they are somehow all 'academic'.
   
   
5. This is commonly implemented for the 'sort' algorithm in the C++ standard library;
   
6. The field of computational protein crystallography (PX) revolves around these functions; due to specificity in data-handling there is no 'good' way to abstract away from these algorithms, i.e., until we solve the phasing problem anyone writing PX code will have to know how to implement these algorithms;
   
7. I write compilers for graphics drivers. We have a laundry list of peephole optimizations and abstract interpretations which requires an entire shelf of books and papers to describe. Each and every algorithm uses this basic idea;
   
8. I also debug the DDI layer of WDDM (UMD) DirectX drivers. The particular problem this comes from was ubernasty; thanks Microsoft!
   
   Since you appear to be totally ignorant of what iterators are, please read Alexander Stepanov's Elements of Programming. It will explain in excruciatingly detailed mathematics the development of iterators and why the iterator is one of the most important concepts in all of computer science.
   
   Also, whoever is downvoting this guy needs to stop. He holds a common viewpoint that is, unfortunately, a misunderstanding. Instead, downvote Fabien4.

Elements of Programming
and
From Mathematics to Generic Programming

I don't know why.

https://www.amazon.com/dp/032163537X/
https://www.amazon.com/Mathematics-Generic-Programming-Alexander-Stepanov/dp/0321942043

Aside from the other excellent choices people have recommended, here are a few I liked that I haven't seen in the thread yet.

• The Art of the Metaobject Protocol
  
  This one sounds super-obscure. It's basically the design notes for the Common Lisp Object System, which isn't exactly a manual you need to read to get your work done. However, if you look at it less a book about how to use CLOS and more a book about how an object-oriented language can be built from scratch, it's really a fantastic little read.
  
  
• Coders at Work
  
  It's what it says on the tin -- interviews with several programming icons. What makes this one better than the other half-dozen or so similar titles is how well the author runs those interviews.
  
  
• Elements of Programming
  
  If I'm honest, I didn't find this one to be that engaging of a read, but it's worth the bit of effort to get through it just to absorb Stepanov's vision for how to express algorithms. He's got a newer book as well that I have high hopes for, but I haven't had a chance to read it yet.

Elements of Programming