May monthly contest

[Sang-drax]

Perhaps 'guarantee' was a too strong word.

I don't see how an explicit formula would be impossible to derive in a similar manner as the formula for the first problem (using the generating function for the difference equation). I'm not able to do it though.

[manofsteel972]

Just found out today was deadline so just posted code for medium.

[PJYelton]

Ok everyone, the contest is over. I'll start going through and grading the submissions tonight, should have the results real soon. I haven't quite decided yet if I'm going to be grading them extensively or not, one of the issues with these problems I didn't realize until after I already started the contest was that you either knew the answer or not, so pretty much everybody had the same solutions. So we quite likely will have multiple winners for each problem.

In the meantime, I'm looking for feedback on these problems. Too hard, too easy, just right? Would you prefer problems that are more involved and require more code since each of these problems could be written in only a few lines? Let me know, don't worry you won't hurt my feelings. We had a LOT more entries this time than the last time which is good, but I'm afraid I really didn't challenge anyone who wanted a serious challenge. Of course, when I did that last time, nobody sent an entry

[Pete]

Challenges were ok, but it seemed that for the more difficult ones, people either knew or didn't, meaning that they didn't nessecarily work out the solutions themselves but already knew. I think challenges that encourage people to think for themselves, resulting in a range of solutions could be more interesting

Pete

[PJYelton]

Agreed that was the issue with the original hard problem which is why I created the second version which did require you to think it through and create an algorithm.

Next time I will be more aware of creating problems that can be solved in many ways. I just didn't realize this until after I had already submitted the contest.

[Sang-drax]

one of the issues with these problems I didn't realize until after I already started the contest was that you either knew the answer or not,

The second version of the hard problem I solved myself. The first version is a standard example of combinatorics.

The problems are more about combinatorics than programming, but I'm not complaining. Combinatorics is very useful to a programmer.

[PJYelton]

Oops, that sentence from me should read "you either GOT the answer or not", which I meant was that if you got the correct answer, you almost certainly had the same impementation as anyone else making it hard for me to give scores.

I think next time I'll bring in some graph problems.

[SirNot]

Personally I liked your challenges a lot, including the shortness of them, as I find hours of mindless coding a little boring in its lack of creativity. As for hardness, I think you could have upped the amount of difficulty on the medium one, as I actually found that one easier than the easy one. Then again, I shouldn't be berating your competitions as I am absolutely horrid at coming up with any sort of programming challenge myself.

In regard to any future competition, if you plan on doing one, do you think you could include something that is based on low-level limits, ie. a timestamp incrementor which dosn't use any of the C runtime library functions. I always loved those type of problems.

[Togra]

I think the challenges were a little too easy. But I enjoyed them nonetheless! Mathematical problems / shortcuts are fun, keep 'em coming.

A totally different kind of contest is "code brevity". I don't know if that's been done before on this board, but I had a lot of fun on the CPP-home (example 1, 2, 3) contests. The problem can be really simple, the fun is to minimize the number of coding tokens. So the emphasis is more on pure C++ knowledge / tricks than creating an algorithm. A 'disadvantage' is that the resulting code will be quite obfuscated sometimes.
Would this be something for the future?

[Sang-drax]

A totally different kind of contest is "code brevity". I don't know if that's been done before on this board,

This is the most recent:
http://cboard.cprogramming.com/showthread.php?t=57330

[PJYelton]

Yeah, hard core or low level C/C++ are not my expertise so I tend to leave those types of problems to the experts. As for code brevity, as long as you aren't using macros 90% of the time IMO the shorter the algorithm the better and more efficient the algorithm and would be judged accordingly. Unfortunately the May problems didn't leave too much room for that, next time though

[The Brain]

Since the contest is over, here is my submission for medium.

[programminggeek]

good, fun problems, i think...

[manofsteel972]

Just a thought. Since you said last time that you thought the problems were too hard why not provide a hint or spoiler. bascially you deduct points for needing a hint but if some people are really stumped on where to start they could ask for a hint each hint being more helpful but deducting more points. You could setup a special page to automate giving out the hints so you don't have to be emailing everyone all the time.
**Edit** either that or send out hints automatically at the end of each week. So you are encouraged to send in your answers as soon as possible. each week that goes by deducts from your score as hints are sent out.