OJ Board

This is getting too hot ...
Suman

Well, I am in a fix. What to do now. Continue with solving? help me..

Do what you like to do. If you like problem solving, why should you stop? And if you don't like it, and just wanted that number one place, then think yourself number one, because if the name of the number one is "Cheater" everyone knows he is, so why bother about that?
And it seems sometimes you also couldn't resist to send in printf programs.
Look at this statistic: http://acm.uva.es/cgi-bin/OnlineJudge2?ProblemStat:2358
This is a program that reads only the input:
[c]
#include <stdio.h>

char cube[70][70][70];
int n,m,k;

int main() {
int l,val,x,y,z;
while(scanf("%d %d %d %d",&n,&m,&k,&l)==4) {
if (!n || !m || !k)
break;
while(l--) {
scanf("%d",&val);
z = val/(n*m);
y = (val%(n*m))/n;
x = (val%(n*m))%n;
cube[x+1][y+1][z+1] = 1;
}
}
return 0;
}
[/c]
It gets WA in 0:00.102, I just tried it.
So in my opinion, all people who got a time that is better with a factor of 4 (this is probably the best factor possible with IO- optimisation) send in only a program that prints:
The number of faces needing shielding is 223202.

The number of faces needing shielding is 6.
The number of faces needing shielding is 6.
The number of faces needing shielding is 6.
The number of faces needing shielding is 6.
The number of faces needing shielding is 6.
The number of faces needing shielding is 6.
The number of faces needing shielding is 6.
The number of faces needing shielding is 6.
The number of faces needing shielding is 54.
The number of faces needing shielding is 62.
The number of faces needing shielding is 72.
The number of faces needing shielding is 150.
The number of faces needing shielding is 26.
The number of faces needing shielding is 214.
The number of faces needing shielding is 150.
The number of faces needing shielding is 206.
The number of faces needing shielding is 150.
The number of faces needing shielding is 616.
The number of faces needing shielding is 7270.
The number of faces needing shielding is 54.
The number of faces needing shielding is 56.
The number of faces needing shielding is 56.
The number of faces needing shielding is 56.
The number of faces needing shielding is 56.
The number of faces needing shielding is 56.
The number of faces needing shielding is 56.
The number of faces needing shielding is 56.
The number of faces needing shielding is 56.
The number of faces needing shielding is 56.
The number of faces needing shielding is 56.
The number of faces needing shielding is 56.
The number of faces needing shielding is 56.
The number of faces needing shielding is 8802.

(This is the output file of the regional contest).

Hi, Adrian!

You are right with it. I was always upset looking at such 'offensive' ranklist

I like to optimize programs and feel delightful getting first rank but very often I see that there are a crowd of cheaters standing at first ten places with 0.000 runtime...

But still I go on optimizing as long as I like it although it makes my AC rate rather low (989/3437). Luckily there are quite many problems without published IO to beat the cheaters

By the way, I dont like to call that people cheaters. It is too much for them. Cheater is a man who is very smart and cunning, I suppose. And there is nothing difficult to send a program printing published output. There are just bots following silly strategy, I guess

Best regards,
Andrey.

Yes, concept of cheater is very vaste.

Peoples may think that Kugel, is a cheater, because he study in ULM, he has a local training, a good coach, a time for programming, and access to excellents materials.

Is Kugel come from AFGHANISTAN, he was solved 1xyz problems ?

Personally, i think that Kugel is a good programmer ! and we lost our time searching cheater, we will never end

Peoples may think that Kugel, is a cheater, because he study in ULM, he has a local training, a good coach, a time for programming, and access to excellents materials.

About "time for programming". In Ulm, there is no course related to ACM, so all time I take for solving problems is time in addition to my normal course work. Ok, I have an advantage because I don't have to work to finance my studies.
And about "excellent materials": yes, we have some prewritten functions available, but it was all written by previous ACM teams from Ulm. And my team has also written some new functions. I think, every university with ACM teams participating regularly in the ACM finals can build this code library.

Maybe, The solution must be, eliminate ranklist pages, AND programming only for fun and learn.

If ranklists are eliminated, this would be a solution for the cheater problem. But the competition character of this site would be lost.
So best thing would be, if the judge output used for the online judge is different than that of the regional contest. Just a permutation of the test cases and adding some other test cases are sufficient that a cheater has no chance.

If ranklists are eliminated, this would be a solution for the cheater problem. But the competition character of this site would be lost.

Today, Sort account by problem solved is not a good idea, because we sort:

-teams
-multiples account for single person
-cheaters
-goods programmers
-Problems setters
-

It is a great shame that those who has access to the judge's data use it improperly..

.. I was always wondering how some people gets certain programs AC in 0.00 s time... and where just taking the input requires a significant time....
... Well now I know why ( thanks to A.K ).

It was no surprise to see the cheater is also at the top of this problem -- http://acm.uva.es/cgi-bin/OnlineJudge2?ProblemStat:2358

,'

Hey Anupam!

Why do you not create additional Input/Output for these problems?

That will help to eliminate cheaters from ranking, you will have the possibility to solve correctly these problems and all will be happy.

Yes, I am sure that Anupam solved a lot of problems himself.
And as you can see for example submission list of 2358, Caesum aka Cronos also send in a printf program, and he is certainly a very good problem solver who wouldn't need to cheat. It happens when people become too ambitious and impatient.
If different judge data would be used than that of the regional contests, then nobody can be led into temptation to send printf programs.

Why do you not create additional Input/Output for these problems?

That would be great

In BUET, there is no tranning on programming, algorithm, competitin. We all, introduced ourselves own in programming contests.

My team also organized the training ourselves. Most of the time it was just participating in uva online contests, and the other times we simulated the regional contest with old regional problems. And after the trainings I tried to solve all problems that I hadn't done during training. Everyone in a team can do this.

SAMS wrote:It is very easy to see who is cheater and who is not.

He wants to say that number one "cheater elimintated by another cheater" is the ONLY cheater here.

SAMS wrote:So, there is only one cheater at archive.
And the account should be eliminated.

And in a flush what's this?

anupam wrote:Again, after participating in many local contest in BUET, we thought that we will go for a regional. But, for partiality we couldn't go. That is why, with frustation I started to cheat.

God damn it, Well done, purpose served. That's what we wanned to listen from you.
Listen, I never tried to take any awfull credit for accepting ( that's what is said) those problems with printf; which YOU'VE ALWAYS TAKEN in your posts.

As Adrian said about the problem 2358 Space Station Shielding, there would be another nice example, 2788 Fax Regions
And there're a lot more like this.

Moreover, I can see that on the 24hr judge you've done so many good things, Solved almost all the Waterloo contest problems which is most of the time the `judge solution' provided by the Waterloo site. The timing does say that. Surely.

Let's end with your own quote...

anupam wrote:That is why, with frustation I started to cheat.

Man o man, I had that frustation too when you were in number one, in fact a cheater in number one, BUT disguised. So with frustation I too started to cheat , how's that make?
Still there's no difference in the ranklist, number one is cheater too, but not disguised at least.

Adrian Kuegel wrote:And as you can see for example submission list of 2358, Caesum aka Cronos also send in a printf program, and he is certainly a very good problem solver who wouldn't need to cheat.

Yeah, and once a printf-program has been sent, it won't go away regardless of how many zillions of real solutions to the problem you submit. For all I know, maybe all of the 0.000 solvers of 2358 have sent real solutions as well.

The ICPC Live Archive is still young, and I think that with time it will evolve and that just like with the regular problemset, this won't be a big problem. It is a good idea to permute the order of the test cases of the data sets and/or add some simple test case(s).

cheater_finds_cheater wrote:Man o man, I had that frustation too when you were in number one, in fact a cheater in number one, BUT disguised. So with frustation I too started to cheat , how's that make?

It makes you the same?

Couldn't you find something more useful to do with all that time you must have spent finding all those data sets? It seems wasteful.


People, try not to get too upset.
For me, the most important thing with this site is the fun I have solving problems.

Why do you not create additional Input/Output for these problems?

I have permutated all the in and outs of the problems I have sent to judge. So, printf can't be used there.

the other times we simulated the regional contest with old regional problems.

I have told many times to simulate a contest by the new problems of our senior brothers.

Man o man, I had that frustation too when you were in number one

So, why didn't you warn me by mail or in front of me? Why spending a lot of valuable time surfing the net?

judge solution' provided by the Waterloo site

Most of the teams participated in waterloo contests. I went to my senior brothers for those problems I can't do during the contest and I went to my senior brothers, they made me understand, solved the problems in front of me and sent them to my account. I didn't know whether they followed the pattern of the judge or not.

On every site, anupam, everyone has friends and enemies
But I think, that a most of use are your friends

Best regards
DM

Anupam,
Don't get frustated. Man's have ups and downs in life. You shouldn't be concern about what other people say. You should do your work without listening to what other people say. Go on guy. You have all the qualities. Now it's up to you.