Well, I think a tool for problem setters to remove submissions to their problems they judge not legal would be really useful. Cheating has been growing recently and I, as a SpojBR admin, have found several judge solutions to problems, most of the cheaters have been talked to and convinced not to cheat again but a tool to quick delete illegal submissions would be really appreciated.
It would probably require some kind of control though, maybe a form telling the reason of the code's removal.
Wondering about his AC solution for problem DDT (https://www.spoj.pl/ranks/DTT/8) I looked at the submissions of user gmsr (https://www.spoj.pl/users/gmsr9) and was wondering even more: It seems very obvious that (s)he has been cheating ... 
....
161115593562260183597018076262500259385225118963936327496691227156776984827584194180',
'637841185472509493966277041641953081675754238090104091048544721209706145413312768740',
'2525330407789119221009341756704875466226455554887350891090156651320061065513932186440',
'9998943371381242321023474793439574481139884832189105555262377011307809353994353116580',
'39593131470570019928884900188787576804513637926117934749025519709205419589642069387800',
'156788800623457278918384204747598804145874006187427021606141058048453461574982594775688',
'620925183926009621146978506218967449531342090729015621989883130549504437230725772687824'
....Source : spoj.pl/status/MCIRGAME,numerix/12
This kind of solution is call cheating or not? It is almost the same with TEXT solution.
I dont understand why some people who can solve only 10,20.. easy problems but like to blame other people cheating. Why they do not use their time to study some algorithms or do some harder exercises?
For me, I only take care about how many problems I can solve, I dont take ecare about others.
But I am wonder why you can assume he is cheating? Because of his high AC rate ? Could you explain it for me?
our idea is to leave the decision to the problem-setter: if some subms are illegal - he can ban only selected submissions,
is it ok?
vdmedragon, from the fragment of numerix's solution that you posted it looks as if he precomputed all possible values locally and pasted a large array that contains the solution to every possible input case. His program simply looks up the solution from the array and prints it. Of course, I may be mistaken (after all, I cannot see all the code).
IMHO this is not the same as a TEXT solution. The user does not know, and is not using, the judge output. Granted that this is not the solution that the problem setter intended, but it is also not using someone else's code, neither is it making use of judge output.
If the problem setter wants to stop solutions like this, it is up to the problem setter to choose the constraints in such a way that makes such solutions infeasible, or simply set a much smaller source limit. Otherwise, if it is the user's own code/solution and it meets all requirements specified in the problem statement, I see no reason why this should be called cheating.
If you are not satisfied with the situation, you can simply resolve this by changing the problem so that these solutions no longer work.
EDIT:
Don't get me wrong. I am not criticizing the problem or the the problem setter. Just pointing out that calling numerix a cheater is not fair, and that you can put restrictions if you don't like this kind of solutions to this problem.
I can't understand that!
According to MCIRGAME, I spent hours to analyze and at least found out, how to construct a sequence that solves the problem.
Then I checked the restrictions of the problem and decided to precompute to get a good runtime. I can't see any kind of cheating here.
And: If you look at my link to the user I named before you can see, that he submitted some TEXT solutions. So he must have known the solution to the problem. How can he know that without any kind of cheating?
Yes, you are right: I only solved about 50 problems, and mostly not so difficult ones.
And: Yes, I should spend time in learning more about algorithms.
I just want to distinguish the difference between TEXT solution and other type of precomputed solution. Sometime they are very similar.
But in all the circumstance, I dont take care about cheating or not, submitting TEXT solution or precomputed solution or not (to get some points???) - but I have seen some people complained about other who submitted existing solution (not TEXT solution) -(which existed a lot in internet).
For me, may be it was good for them because we have different level of official/unofficial training. We can see a lot of coder here who can solve only very easy problems because the lack of knowledge/experiences, so the source code may be useful for them. Even us, sometime we need to verify the solution work or not, and how well they work so we submit it.
The one thing I want is that after having some solution, they can adapt it to other problem, so it will be really useful for them. For example, I cant understand and solve some segment,interval tree or BIT problems if I havent had a pascal solution and some document described these data structures (in vietnamese). But I know there are a lot of coder here who dont have any source code/documents about it. So I am right if I blame them when they submit some existing source code for their purpose of study.
I graduated from university 6 years ago and I just started to code about 1 years when we have a judge online in Vietnamese. So i think I can understand the difficulty other people have (lack of documents/training/dont know how to find a solution for a problem) ... but because of others experiences/advantages I have so I dont need to post a lot of question in forum to obtain the assistance.
Sorry if you misunderstood, because I did not understand how many TEXT solution he had submit so you assumed him "cheating", I saw 2. But anyway, we dont talk about it any more. 
Well, I understand your point of view about someone else's code source, I didn't mean to say people are forbidden to look at a correct code and learn from it, but what advantage do you get by simply submitting the source code, if you already know it is already right. There are some people who submit judge's solutions and don't even care about removing the comments the person who did the code made. This kind of atitude should not be tolerated, at least I think it is a really bad practice, you don't have to submit someone else's right code "to make sure it is right".
Learning from other people's code should encouraged, in case you are not capable to have the idea, but not the submission of a code you didn't touch.
I think the point is not, whether they are precomputed or not, but how to get the precomputed version. If one (e.g. me) analyzes a problem by himself, working at it, implementing an algorithm that produces the needed data and than - because of loose restrictions to a problem - submit a precomputed version, doesn't anything wrong.
No doubt you can learn a lot from others source code. But I learn by analyzing it, making experiments with it and try to understand. But I would never submit someone elses code just to check whether it works. If someone posts e.g. code here in the forum telling us it was AC - I will believe that and have no need to submit that code again.
In the case of the user I named before there was obviously something wrong. Anyone who looks at his submissions will see that. And IMHO it is in the interest of all honest SPOJ-coders to hint at those cases so the admins can have a look at it and decide, if there should be some kind of reaction. That was all I wanted to do.
The main difference between a TEXT solution and a precomputed solution is that the TEXT solution must have used exact knowledge about the contents of the judge data.
The submitter of TEXT knows exactly what the judge output lines are and their relative order. The precomputed solution does not rely on knowledge of the judge data, since the user is reading the input and producing corresponding output. It also uses a list of all possible cases, not just the ones in the output file. The user may have used a slow/suboptimal solution to come up with the precomputed values, but it is still the user's own solution (unless you have proof otherwise) and can not be called cheating.
Frankly, I don't understand why TEXT is selected by default in the list of allowed languages in the problem editing page. Except for a few rare and unconventional problems problems (e.g.problems with no input like BCEASY), I see no way that someone could submit a TEXT solution and not know anything about the judge output.
DOC, PDF and PS are deselected by default, maybe deselecting TEXT is also a good idea to prevent cheating.
spoj.pl/users/humblecoder/13
I think this person is cheating.
He is i think submitting judge solutions, which does a lot of discredit to those who solve hard problems with a lot of effort.
What is noticeable is that he is solving hard problems from COCI and University of Ulm contests (infact he has solved entire sets)
However he cannot get easy problems like KROW and BISHOPS.
How i noticed this was that person was submitting solutions very fast, i think he solved 3 COCI problems in under 20 mins (which is hard even for strong coders). Consequently i emailed him asking him for a hint to POLQUERY to which he said... he used DFS to check for connectivity... which is naive and will not clear the tests.
I request the problem setter (vdmedragon and AdrianKuegel) to look into the solutions and take appropriate action if found guilty.
He submitted judge solutions for several problems.
At the moment, the function to disqualify the solution doest not work well. I cant disable his submissions.
I did not find his email to suggest him to disqualify these judge solutions himself.
give me a private message.
Thanks.
humblecoder disqualified himself these submissions.
You can find more here, spoj.pl/users/humblecoder/11
Before: Current world rank: #146 (64.1 points)
)