Printable View
Ok just to continue on the existing thread (It was getting very cluttered) Ok only suggestions here please, no chitchat ;)
Im working with a few ideas myself but please post here with suggestions. I will have a contest up and running in the next week.
Pino
How about something in Malboge?
Looks pretty die hard, plus... It's a completely different language, so it's open to any developer from any language. Learn it and develop something impressive (no set project, just show how quickly you can learn and build something in a given amount of time).
I dont think that would be a good idea, firstly languages like that ( I can name more) are pointless, also it doesnt give our younger members much of a chance.
This is a VB foum :)
Thanks for the ideas though.
They're not pointless, PASCAL was pointless; this is educational. ;)
That's cool though, just throwing it out.
Not thrown it out, just looking at 2 facts.
1) people most likly wont be motivated to do this kind of contest
2) 60% of people here work full time and wont have time.
Lets just stick to vb unless this gets alot of interest expressed in it!
No, I mean "I" was throwing it out... Like "to throw something out there," pitch an idea? :)
Sorry :(
O i get it, makes sence now :p
Anywasy, dont appoligise, thanks for the suggestion :)
I like Merri's idea of an ongoing contest. I think it should go more like this though:
1) You make the code and post it when you are done
2) You get judged and your score is posted
3) Your code is NOT made public
4) Other people make code and their score gets added to a table or something
5) You have a chance (Every 3 weeks or so) to post a new submission and you get a chance to imporve your score
This leads to less competition and more sharing of ideas, but doesn't allow other people to steal your code and call it their own. We could even have a few of these contests going at the same time with different goals.
Any takers?
Merri, can you think of any speed related challenges we could do this with?
well that would put alot of stress on the mods. They probably dont want to look at lots and lots and lots of programs. Now I hope this isnt chit chat, but I know im looking foward to being in a contest, although it will probably be hard, and my code will suck, but I do think that some of the judging sould be based on the way the program looks! - and the speed. But the user interface of a program is always important.
Hehe.. im all for the programs that are plain and gray... with one button.. where everything you needed is done by clicking that one magical button :) Who cares what it looks like, as long as it does what I need it to do. Just my opinion :D . It can be the most beautiful interface in the world, but if it cant do what you want, whats the point :)
As for a contest suggestion... how about.. an unjumble algorithm... fastest way to unjumble a string of characters to output all the possible words from a word list that is made with those letters... (simply because I have already done something like this hehehe) I need a faster algorithm for Anagrams, actually :)
I think people are far too confused where to post... As the first post says, this thread is not for future contest suggestions. It is in general for discussion about how to make a contest that works. Apparently Pino has something on his mind, but which will be running on the old contest format. But here we're looking for an alternative way to run a contest.
With my idea, I meant it to be very lightweight on mods: the contest could be run by a regular member thus ensuring bigger interest on the contest (mods are kind of "forced" to run a contest, thus running a contest along with the normal moderator work can get bothersome).
Also, eyeRmonkey: I don't agree with code kept private at all. Code optimization has a lot of emphasis on coding techniques, thus making different techniques and algorithms public is much more helpful for sharing ideas than just showing "look ma, my code runs super fast and others don't know how I do it!" - this is exactly what I don't want to happen and it makes sure everyone can't join in. Pure discussion is not enough, many people aren't either able or willing to tell how their code works... and the main point for the contest, as I see it, is educational: make people analyse what others have done, why it works, see if a concept is good and if it is possible to use an idea from another code in your own code to make it faster or if it is possible to make a new better code with an idea seen in another code. Also, as everyone will see everyone elses codes, it is easy to see if someone is just trying to steal a code. I wouldn't even call it stealing: I'd call it improving a code (as long noticeable changes are done). "Stealing code" and customizing it is an effective way of learning new things. I wouldn't be as good as I am now if I never "stole" code to analyse it.
- Contest starts with initial example code and instructions
- You submit your better code for rating
- Code is benchmarked and rated and placed on the table and made public
- You're allowed to submit one new code every day (although contest maintainer isn't forced to update daily)
I also thought it would be good to have a few different categories, ie. code that assumes input is always valid vs. code that assumes it may get invalid input. I also think that there could be separation with a function made into a class module vs. a function made into a module. Class module allows to initialize arrays and clear memory and stuff whereas module function has to always do everything by itself and it can't have any precalculation.
The first contest I have on my mind would be very easy and it should be relatively easy for anyone to join into. The function would be rather pointless (in real life application it wouldn't require any speed optimization), but I know from experience that it is a real good one for training and learning.
Where, I said this thread is purley for suggestions no chitchat, Might be grammer though. I'll make it clear!Quote:
Originally Posted by Merri
No, i'm open to any changes, the contest format is not set just yet, I just want to get somthing running.Quote:
Originally Posted by Merri
I am allways up for running a contest, I wouldnt mind part running it with a well established member (Like yourself) but I really wouldnt mind taking the full responsibilityQuote:
Originally Posted by Merri
Sounds interesting....Quote:
- Contest starts with initial example code and instructions
- You submit your better code for rating
- Code is benchmarked and rated and placed on the table and made public
- You're allowed to submit one new code every day (although contest maintainer isn't forced to update daily)
Can you PM me with this? or contact me on MSN? I'm interested.Quote:
The first contest I have on my mind would be very easy and it should be relatively easy for anyone to join into. The function would be rather pointless (in real life application it wouldn't require any speed optimization), but I know from experience that it is a real good one for training and learning.
Also as to posting code public, wouldnt this mean we would end up with 20 entrys all the same?
Thanks
Pino
Hmmm, not entirely necessary....some one who has a pretty good knowledge about perl/php - could design a system where uploaded submissions are automatically benchmarked/tested and updated? (provided that the submissions have an specific format)Quote:
You're allowed to submit one new code every day (although contest maintainer isn't forced to update daily)
Just an idea - :) because i certainly can't program in perl or php :P - i should learn them sometime though :D haha :)
Cheers.
No, nice idea but i'd prefer to test each manually. :)
Raedwulf: Such script would require a computer which has both Visual Basic and PHP installed... basically coding it and setting it up overall would be more work than it is actually worth and it would cause a big delay. Not against the idea, just think it isn't going to happen in practise due to the complex and time consuming nature of it. Another problem of automatical benchmarking and scoring would be the incapability of the system to figure out what kind of code it is.
Pino: I thought you wanted to take out the contest future discussion out of the suggestions thread so that the subjects would become much more clear and separated: these are two different things. As it is now, this thread isn't much better than the old one: it can get just as out of topic and it is even harder to find the actual suggestions... though in the end, I think suggestions don't work. Very few of the suggestions (maybe 0.001%) have come to be a contest, so looking for a contest that could work instead of an idea (suggestion) for an existing contest model might work much better.
Oh, and I opened MSN.
contest suggestion: design a program to solve a given rushhour board (of any size) in the shortest time possible.
this would make a good coding challenge as the general brute-force method of solving these puzzles does not work. i have done this previously for a uni assignment (in Haskell) and it is a lot of fun. it would also be easy to find the contest winner: the person with the fastest solver.
Play the game to try it out.
that sounds like a good idea! Sounds hard though...
it would involve finding the best pathfinding algorithm.Quote:
Originally Posted by damasterjo
not just "who can create the best code"...
I wouldn't call it pathfinding. The red car can only move to left and right or up and down, directly to the exit. It is just a matter of finding the minimum number of moves required to allow the red car to move to the target position. Doing such moves on a 6 x 6 field isn't all that costly. Constest rules would limit the most: if the contest was run using similar rules to what was in the pathfinding contest with penalty point scoring for each thing done, then we'd have the "random vs. logic" dilemma. If we're only looking for the fastest code to solve the puzzle, then it'd be up to the programmer to figure out how to make a fast algorithm + code combination.
it can be solved (after a long time) using simple methods such as dijkstra's algorithm, but there are much faster (heuristic) algorithms that could be used.
I think some small contests should get going.... just to spark interest in the program.....
Like simple algorithms graded on accuraccy and time::
Easy contests like:
1. Suduko puzzle solvers
2. Number Recognition
3.A suduko solver where an image is inputed into the algorithm with a few squares filled out and a completed one comes out..(impliments number recognition..)
4.Given A black & white image with a puzzle, find the way from point a, to point b .......and draw the answer with a red line...
5. A maze generator....
I know these sound childish, but I'd think they'd be fun little contests to have to get contest interest started...
I agree. Even though we have already done the basic sudoku solver, the others sound like fun. We just need to get some interst in them. The last contest (cannon) didn't follow through.
I like the sound of this. Sounds like a bit of fun. there are a few different ways of getting the code to preform the actions required so should give a few varied entries.Quote:
Originally Posted by Modderman
and it would also be nice to kick off a contest. I know its hard to think of good contests but we could have done two not so good contests in the time we have spent talking about what contest to do.
sounds good to me!Quote:
Originally Posted by makster246
Sounds like the 2nd contest, the pathfinder one...
I think you could well end the whole contests forum now. Last time I was willing to run a contest all by myself, but then I was asked not to say anything the contest and since then it has been all silent (for reasons not told to me; I first thought they just wanted to pull out something themselves but it has been many weeks now).
It seems there is no interest to run any kind of contest, there is no encouragement to held a contest and even people who are interested to run a contest are turned down. You are constantly worried about if enough people participate, but the less people do that if this keeps going on as it has. By worring about if the contest idea is good enough to get "enough" participants kills the interest totally.
Btw, I think contests forum would work much better if there was clearly one person responsible of it. Now I don't know who really is the person who decides about it. And it seems people responsible don't know either, like if everyone just keeps pushing it out of their thoughts: "I'm not responsible so whatever". I don't know who I should really write this message to so I complain in general and hope someone cares.
We've been told for the last couple of months to not start any contests.. see here for a hint at the reason (we cant tell you more than that - we dont know!):
http://www.vbforums.com/showpost.php...0&postcount=21
Contest idea:
Removing noise interference from a noisy WAV file using C# / VB.net.
Points for speed of execution and quality of resultant audio (selected by judges ear, assuming they have decent speakers).
The wav file would have to be standardised for everyone of course. Maybe 30 seconds long and contain music and speech and some random sound effects.
I like the above steps mentioned by eyeRmonkey, he made a good point in the starting to offer a less competiive solution with healthy competition and good chances for win for every candidate.Quote:
Originally Posted by eyeRmonkey
Modderman's sudoku puzzle idea and wossnames's wav file idea are also appreciable.
Carry on guys!!
Brad mentions in there about a contest in the next couple months, but it has been way more than a couple months and I haven't seen any contest. Just curious what's up.Quote:
Originally Posted by si_the_geek
It happened a while back - it was posted as a forum-wide announcement (can't remember what it was tho, I think it might have been an extension to VS.Net :( ).
Well that just doesn't work. We've already learned to not read the forum-wide announcements because all they always contain is advertisements... (advertisements that aren't related to my interests.)
I just don't see a good reason why we can't have a nice user-to-user contest. These tend to attract different people than the commercial contests and it could have VB6 included, which is what you can't see with the commercial contests. They don't really eat participation from the "more official" contests.
There have been a few forum-wide announcements for other purposes too.. it would be wise to at least have a brief look at them when new ones appear (unless the title is obviously an Ad for something you arent interested in), as they can contain useful/important information.
Back to the competitions.. as the big inter-site one has finished, we can have another "user-to-user" contest. I'm afraid I don't have time to oversee it in addition to my current workload, but hopefully another moderator will be able to.
The first step tho is for you guys to decide what the subject of the contest will be, and preferably the rules too.
Unfortunatenaly that isn't how human brain works: I only now noticed that I haven't paid much attention to them since I noticed they often contain ads; or contained them when I did pay some attention to them :) So basically once it is being used for something that is seen somehow negative/uninteresting, it tends to generalize automatically to "things to skip".Quote:
Originally Posted by si_the_geek
Anyways, I guess we can drop this subject now and get more on the contest (I guess we both see each others points by now, being better than Artificial Intelligence), so... does anyone else besides eyeRmonkey and webguru like the idea I proposed many months ago?
If it would be anything speed related, it might be better if I'd be a judge: I could give feedback on the works so that the codes could be improved, atleast the VB6 ones :)
vrooom vrooom, speed! :D
Unfortunatenaly this is still dead, so I did something and put up a QuoteSplit challenge. No prizes, no victory, just different solutions to this classic issue that comes into picture from time to time (most often when someone wants to get long filenames passed via Command$ or when drag and drop support is done for dropping files).
What does a Solitaire cardgame solver sound like?
No real point in it, but it might give atleast some degree of challenge for brains to figure out how to do it.
An idea for a challenging contest:
Solve a Sliding puzzle of specific size. I think a 3x3 or 4x4 grid (http://en.wikipedia.org/wiki/N-puzzle) would be large enough for this...
I think that to make a contest(s) that everyone can join you should have a beginner, inter and adv group so that the new people can make something simple and the others can do slightly harder things?
I think that it should be done using C/C++Quote:
Originally Posted by Pino