Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Telgin]

Here's one dedicated to that purpose:

Stack Exchange's Code Review Section

If you're not familiar with the Stack Exchange sites in general, then I'd highly recommend them.  Stack Overflow in particular is a great wealth of information on specific programming problems, but I do advise you to be careful about searching for questions on that particular site before asking any new ones, since most things have been asked a dozen times and closed as duplicates with extreme prejudice.

The code review site is much newer and obviously subjective, so it should be more accessible.  Having said that, I've never asked for a code review there before, so I don't know exactly what requirements they have for posting code snippets.  Be sure to read the rules.

[jaked122]

Recursion is a powerful and compact notation, but it's very costly in terms of memory and speed. There's basically no situation where you're not better off just using a loop in terms of computer resources. But it could have readability advantages some times for code that you know won't get called very often, so saving in programmer time.

Tail call optimization.


 That's all I'm going to say. Lots of languages have this capability, I think that gcc can do it for some circumstances in C and C++, Haskell is useless without it, Ocaml is also useless without it, Scheme is generally supposed to use it when available(not all implementations get around to doing this), common lisp seems to do it sometimes... if you use the right lisp(Clozure common lisp, SBCL, maybe CLISP).


Of course,while you can implement DFS using a loop in any language, not using recursion is pointless because you need to use a stack anyway.


Anyway saying that you shouldn't use recursion isn't quite good advice. Recursion should only be done if it makes sense to, I.E. you use the stack in order to expand upon computations compactly.


It shouldn't be used where loops make more sense, but if it is harder to implement as a loop than as a recursive function, then perhaps it might be a case where it makes sense to use.

[breadman]

Recursion is a powerful and compact notation, but it's very costly in terms of memory and speed. There's basically no situation where you're not better off just using a loop in terms of computer resources. But it could have readability advantages some times for code that you know won't get called very often, so saving in programmer time.

I ran into a situation once where it was called for.  It only went six or seven levels deep, but had a messy state at each level, calling the next level many times before returning.  Yes, I could have used an array for each variable, but recursion was much easier to understand.

I've since lost all copies of the program, but its result is prominently displayed on a U.S. Patent.  My father even decided to frame it for me.

[Bauglir]

i would like to buy your patent

i will put it into production, and call it the recursatron

name your price, but know that the only currency i have available is honor and good wishes

[Dutrius]

I can think of a function where recursion is required. The Ackermann function.

Code: [Select]

int Ackermann(int m, int n)
{
    if (m == 0)
    {
        return n + 1;
    }
    else if (m > 0 && n == 0)
    {
        return Ackermann(m - 1, 1);
    }
    else
    {
        return Ackermann(m - 1, Ackermann(m, n - 1));
    }
}


[TheBiggerFish]

Patent?  Seriously?
Good job?

[DrunkGamer]

Hey guys, I want to hear your opinion on something. (I'm annoying everyone with a knowledge of programming better than mine with it  )

What do you believe is better? To read a bit about any language and then teach yourself the rest by googling, reading snippets and all that stuff (Slower, but "easier", so to speak) or reading a book entirely and doing what it says?

[Putnam]

the former

it's not really slower, you get results faster that way too

[Orange Wizard]

Yeah, I prefer the former. Books can be really useful too though, especially with concepts you're having trouble with on your own. Just following them 100% is a bit much.

(I'm annoying everyone with a knowledge of programming better than mine with it  )

Don't worry about it. The thread exists to discuss programming, at all levels. If people are annoyed by someone making an effort to learn then they can sod off.

[Antsan]

Most of the time I learn via Internet, but to learn unusual concepts I find that most of the time books are better for that purpose.
To get behind Shen's type system I read "The Book of Shen" (of course) and to achieve a deeper understanding of macros in Common Lisp "Let Over Lambda" was quite useful.

So, stick with small-scale stuff for now but if you encounter something that interests you and you want to get a little bit deeper into that specific topic, try to find a book on it.

[Moghjubar]

The best way to learn is to make something.  And not something insane, something like a mini-text rpg with dice rolling and input/output, etc.  Yes, this is simple, it doesn't have to be Cataclysm.

Theres definitely merit in following tutorials or reading a book all the way thru, but IMO: get your basic setup done, figure out what you want to do, then start doing it and learning all the actual references you need to look up.  You will run into roadblocks, sure, but you can get past them.   Your drive to complete the project will help, instead of just doing checkmarks and saying 'I read 5 pages today' or 'I did the exercise section' or whatever.

Past that, you can just keep making stuff all the way up, and maybe getting some use of what you make too, instead of just dumb tutorial projects.

[EnigmaticHat]

Hey guys, I want to hear your opinion on something. (I'm annoying everyone with a knowledge of programming better than mine with it  )

What do you believe is better? To read a bit about any language and then teach yourself the rest by googling, reading snippets and all that stuff (Slower, but "easier", so to speak) or reading a book entirely and doing what it says?

I find I'm motivated by constant progress.  So googling stuff is great because once I get it to work I can immediately see a functioning result to my code.

[Reelya]

I can think of a function where recursion is required. The Ackermann function.

100% Required or just preference? In my opinion it's extremely rare to find an algorithm which absolutely cannot be converted between iteration and recursion if you put some effort in.

A quick google shows that it's possible to write a loop version, and I'm pretty sure that the loop version could be optimized faster.
http://stackoverflow.com/questions/10742322/how-to-rewrite-ackermann-function-in-non-recursive-style

[Bauglir]

Fussing around with SMTP for a class. I can work my way through sending an email manually over Telnet no problem, but for this assignment we need to automate it. I can get through most of it, but for the life of me I cannot figure out how to end sending the message body. I know you send a "." on a line by its own, mind. I just never get the expected 250 reply from the server.

Code: [Select]

if (send(sockFD, ".\r\n", 3,0) != 3)
{
        perror("Error sending message.");
        (void) close(sockFD);
        exit(1);
}

I have tried every variation on the above that I can think of - writing the text to be sent to a fixed-length buffer and calling send() on that instead of a string, every combination of the characters ., \r, and \n that I can think of from . alone to being flanked on both sides by \r\n, and appending it to the last part of the email body instead of sending it in its own packet.

halp

y it no work

[Dutrius]

I can think of a function where recursion is required. The Ackermann function.

100% Required or just preference? In my opinion it's extremely rare to find an algorithm which absolutely cannot be converted between iteration and recursion if you put some effort in.

A quick google shows that it's possible to write a loop version, and I'm pretty sure that the loop version could be optimized faster.
http://stackoverflow.com/questions/10742322/how-to-rewrite-ackermann-function-in-non-recursive-style

Ah, Ok. I just couldn't think of one myself.