CodeCracker National

How everything panned out

My college mates will immediately know what this page is about and an intro should not be necessary.  But let me assume that people unknown to me may also visit this page. The thought makes me feel good! 

Linux Users' Group NIT Durgapur organised a national level programming contest , CodeCracker, on the 8th of February,2008. It was our 1st time and we thus joined the class of colleges that have an online judge and have organised an online programming contest. That list includes 5 IITs and 5 NITs i think.  It was organised on the CodeCracker software developed by us. The software itself was selected for the finals of Opensoft, the software designing contest at IIT Kharagpur (but we dint win anything :-(  )  . Everything went almost perfectly well, except ofcourse a few hiccups, which i shall elaborate on later. 

You could call it an international contest, coz everyone cud, and did, participate, but prizes were only for Indian nationals. The contest began at 10:00 PM and ended at 04:00 AM IST.

We installed statistics tracking code on all of the pages. Here are the resulting screenshots. 

The day before the contest saw 90 page loads and 50 unique visitors. The day of the contest saw 17 095 page loads, and 1500 unique visitors approx. Here are the places from where people visited us:

   


The Ukrainian visitor not only visited but also registered and participated. His handle was "vlad89" and he was the first person to solve a 500 point question. He ended up in the 31st position. here is the mail he sent me after the contest got over.

"Hi

My handle at Code Cracker was vlad89. I tried to solve 8th problem,
but received Wrong Answer. Actually, it doesn't play something for me
(I compete in the contest just for fun, and solve only hard problems,
so I anyway have low score), but maybe for others it would be
important. Thus, I test my solution on the test cases, that you have
published, and my program returns right answer. Number of operations
in my output is coincide with the authors', but operations list that
my program outputs differs from one that author have. And I have
checked both lists (mine and authors') and both of them are correct.
So I suspect that the error is in your checker program, please look at
it more patient, and I hope you'll quickly find a bug.

Thanks, Vladyslav Simonenko"

And there was indeed an error in the 8th problem, but more about it later. 



Here are visitors from Santa Clara California. As you may know, Google headquarters are at Santa Clara Cali. and also several other big Companies such as NVIDIA.


Here are people from NVIDIA.

Well, why, you might imagine, do people from these places visit our contest. It is probably because employees there are interested in spotting talent in international arenas. Or maybe simply because some of them young employees actually participate!! Which ever the reason maybe, but they now know a college by the name of NIT Durgapur, INDIA.


It was an interesting mix of countries visiting our arena. Initially the communist countries, such as Cuba, Russia, China registered. They tore apart the practice sessions with their international standard skills. Then teams from USA, UK, Mexico came along. There was also a participant from Bangladesh.


Here are visitors from France and that too 5 returning visits.
I guess people found our questions challenging. Thanx to Swapnil for his original and tough problems that even world class programmers had problems in solving. The problems were mainly from places where coders dont visit. There were also original problems. That saved us from problems that IIT Kanpur faced 2 days later. Their contest was over very quickly as coders copy pasted solutions found on popular coding sites such as Topcoder.com. Many top teams were then banned and then problems ensued. One of the banned teams were Aryans, winners of CodeCracker 2008.


Another screenshot of visitors stats.
 

 Yet another.

Regional stats.
 
Taking snapshots of the visitor stats was very tricky. There were so many visitors and the stats were changing so quickly that we could not keep up with the changing percentages of visits from different parts of India or the rest of the world. We missed taking  snapshots of several visits from Redmond, MIcrosoft Headquarters and also a substantial number of visitors from HELMOND, NOORD-BRABANT, NETHERLANDS (CHELLO.NL). 

To be honest, a lot of international coders visiting our page is not much of an achievement. People also visit porn sites but that doesnt mean they are doing great work. We are proud of the fact that people visited our pages, and then registered and participated and stayed till the end, even though there were no prizes for them. The system and the software worked without fail for the 6 hours. Thanx to Shreyank and Supratik Da for administering it along with me.


One of our aims was to make it possible for everyone to participate. Even our second year students who maybe just beginning out on programming.  Hence the format of our questions. 4 easy, 3 medium and 2 difficult problems. Here is a screenshot of Vineet Pathri getting a problem correct. I dont mean to say that he is a bad programmer or anything... just that he wants to program and he doesnt program a lot. He found an opportunity to code on a platform such as this and it excited him and he tried and got one correct.



I would also like to show you some system statistics collected during the contest.

74% users were on Mozilla firefox, and 34% were on Linux. Thats pretty good i think. Of course, most people in our college were on windows, but with the administration's help, LUG plans to change that very soon. Having said that, the contests that were organised on the lan brought many new users to Linux.

Now about the problems that we faced:

  • Question number 8 was an awesome question. But it had a small flaw. There were multiple correct sequences possible for an input and we only had the most obvious one in our answer. Anyways, there was some protest by some of the teams, but we decided that since it was problematic for all the teams, there was no bias towards anyone. So that problem was solved.
  • The LUG server is not really under our control, and the college's lan architecture is horrible. We had to make changes to the subnetting to make the contest server accessible to the public on the lan. This was necessary since the net speed in our hostels is very poor and people would not have participated otheriwse. Ultimately and unfortunately too, the girls' hostels could not be connected, but they accessed the arena via the internet.
  • We were very quick to respond to all the problems that the contestants faced. I was always online on Gtalk and many people started to bother me just out of fun. I had decided that i would be polite to all of them so that they come back next year to participate. But ultimately it turned out to be irritating.
  • I was feeling very hungry and a little sleepy too !!

in hindsight, we should have approached sponsors for the event. Given the number of users that visited the site, it would have been easy. We will do it next year for sure, if LUG still remains an independent club that is.

The motive behind organising this contest on the lan was manifold. We wanted to develop the software to be capable to handle a national online programming contest during Mukti. We also looked at the tremendous capability of our college's network infrastructure and the programming talent distributed over it. It was a major part of a new LUG which aims to focus on software development in the future and hence needed to identify talented programers.
 
Thats all... too much typing for today.