Posting solutions outside of Project Euler's site

Announcements, comments, ideas, feedback, and "How do I... ?" questions
User avatar
B-Con
Posts: 2
Joined: Tue Dec 18, 2012 12:24 am
Location: USA
Contact:

Posting solutions outside of Project Euler's site

Post by B-Con »

I'm curious to know if there is an official position by Project Euler on problem solutions being posted outside of the Project Euler website.

Specifically, I want to store all my solutions in a GitHub repository as I work, but that gives aways exact solutions to everything I've done. Obviously cheaters will find ways to cheat, but I don't want to serve as an enabler if PE has specifically requested against this sort of thing.

Does an official position exist? I didn't see anything in an FAQ, apologies if I missed something.
Image

User avatar
hk
Administrator
Posts: 10875
Joined: Sun Mar 26, 2006 10:34 am
Location: Haren, Netherlands

Re: Posting solutions outside of Project Euler's site

Post by hk »

We strongly disapprove of posting solutions that are accesible to the general public.
If you feel like having your solutions somewhere else, please use some password only known to you, or use some encryption.
The reason should be obvious: members coming after you are supposed to solve the problems themselves as you have done.
Image

User avatar
B-Con
Posts: 2
Joined: Tue Dec 18, 2012 12:24 am
Location: USA
Contact:

Re: Posting solutions outside of Project Euler's site

Post by B-Con »

OK, thanks.
Image

op=op
Posts: 3
Joined: Thu Oct 04, 2012 5:18 pm

Re: Posting solutions outside of Project Euler's site

Post by op=op »

I noticed that in problem 409 there are
4 solvers on day 3
3 solvers on day 4
14 solvers? on day 5 (still 12 hours to go)

I found on a site a reaction of riemann ("solved" 375+):
Stop guessing who I am . I just want people to learn out of PE.
Don't think, just code:
#409="a number"

How demotivating.

Even those who want to solve a problem and struggle to find a mistake in the code the temptation is too great not to look for the answer on those nasty sites.

Suggestion:
Remove the answers on solved problems and change every problem a little bit.

or: change a few problems a little bit and check if the answer given equals the answer of the old problem. In that case remove the person from PE.

thundre
Posts: 356
Joined: Sun Mar 27, 2011 10:01 am

Re: Posting solutions outside of Project Euler's site

Post by thundre »

op=op wrote:I noticed that in problem 409 there are
4 solvers on day 3
3 solvers on day 4
14 solvers? on day 5 (still 12 hours to go)
I see evidence of at least collaboration on the fastest solvers list. Two accounts from the same country have been within a an hour or two of each other on all problems since 405 (except 406, which neither of them solved).

Can someone suggest a good statistal approach for finding likely collaboration? If we can't stop the cheaters, we might at least have fun outing them.
Image

sivakd
Posts: 217
Joined: Fri Jul 17, 2009 9:37 am
Location: California, USA
Contact:

Re: Posting solutions outside of Project Euler's site

Post by sivakd »

> or: change a few problems a little bit and check if the answer given equals the answer of the old problem. In that case remove the person from PE.

Nothing can be done if the person registers again and continues down the same path, it's just a headache to admins. But at the same time, it might result in false positives. If someone tries a problem and can't solve it and revisits after a few months and just use their hardcoded parameters and solve the problem, they are going to get the old answer and submit it.

I think any system like Project Euler has to accept the reality that there certainly are some people who want to collaborate (just like how they might be doing their class projects or at work). One way to allow this without distorting the statistics is, introduce the concept of a group. So, a user can be an individual or a group. Group users will have an admin who will approve who are the members of that group. When the group solves a problem, the answer should be submitted using the Group username. Once it's submitted, every member of that group will get to see the private forums. Hopefully these members will not submit the same answer on their own. This group idea is just a high-level idea which can be refined with a few more discussions.

Then there are some users who just want to get in for the purpose of showing up in the reports. Identifying such users is actually not the difficult part. Designing a system that doesn't offend them is. Because once they know that the system recognizes them as not solving the problem on their own, they can choose to recreate a new account and start over. One way to easily identify such users is to publish incorrect answer to some of the tougher problems on some random websites.

Overall I understand this is not an easy problem and any particular implementation may not satisfy everyone.
Image
puzzle is a euphemism for lack of clarity

op=op
Posts: 3
Joined: Thu Oct 04, 2012 5:18 pm

Re: Posting solutions outside of Project Euler's site

Post by op=op »

sivakd wrote:> or: change a few problems a little bit and check if the answer given equals the answer of the old problem. In that case remove the person from PE.

Nothing can be done if the person registers again and continues down the same path, it's just a headache to admins. But at the same time, it might result in false positives. If someone tries a problem and can't solve it and revisits after a few months and just use their hardcoded parameters and solve the problem, they are going to get the old answer and submit it.
I don't agree.
There are a lot of ways to prevent registering again.
E.g. change some problems. When an old answer is given, send an email to that person, explain what has happened. After a second old answer give him a last warning(*). After a third old answer, remove the person. React every time after he has given more than 10 answers to other problems, or after some days, so that he does not know what problem caused the alarm. I do not belief in False positives (more than two times) when the answer exists of 10 or more digits.

(*) There are a few extra things you have to consider.

User avatar
Marcus_Andrews
Administrator
Posts: 1517
Joined: Wed Nov 09, 2011 5:23 pm

Re: Posting solutions outside of Project Euler's site

Post by Marcus_Andrews »

Just some comments on what's been said thus far:

1. Altering old problems would likely create many false positives. People who are in the middle of working on a problem may not notice the changes. Similarly, like sivakd mentioned, someone could come back to a problem later using their old hardcoded parameters. While I think the spirit of op=op's suggestion is a good one, I feel like it wouldn't stop a dishonest solver from getting access to source code and just changing the relevant parameters. In the end, it would just serve as an inconvenience to the admins and honest solvers.

2. While this is just my opinion, I don't know if Groups would help the issue much (even though I'm sure that some would still get utility out of that feature). I think the public releasing of answers is a far bigger concern that impacts the integrity of the statistics in a nontrivial way.

3. Posting fake/shill answers is actually a tactic that DavidM of Notpron once tried back in the day, but the problem is that it's easily exposed (especially if you have someone new with no post history tossing out an answer to a very hard problem). People learn to test the waters with dummy accounts and warn others before leaping in with their mains. All this does is muddy up the rankings with dummy accounts, which makes the problem worse (and this is actually what happened with Notpron, too).

DavidM ultimately decided to charge users a nominal fee past a certain point of completion. Serious participants were willing to incur the cost, and dishonest types weren't as likely to follow suit. In other words, the PE-equivalent would be charging users a fee if they wanted to participate in the Fastest Solver rankings. There are many obvious downsides to this idea. I'm just explaining it to illustrate what other communities have done to address their respective problems.

One of the easier things to track is how quickly people are solving problems. If someone is hammering back 344 through 361 in a single day, I think it's safe to say that that user should be flagged. I exaggerate with my example but I hope the point is clear. In the same vein, the users who comprise massive solve-count spikes should be flagged (such as those in the recent 409-surge).
Last edited by Marcus_Andrews on Fri Jan 11, 2013 10:07 pm, edited 1 time in total.
Image

User avatar
hk
Administrator
Posts: 10875
Joined: Sun Mar 26, 2006 10:34 am
Location: Haren, Netherlands

Re: Posting solutions outside of Project Euler's site

Post by hk »

Thanks all for being on the alert.
Very recently the fastest solvers list was extended from 50 to 100. However the Eulerians table is still based on the first 50.
We deliberatedly took the risk of the answer to a problem being revealed before the first 100 were filled. (In the past this also has happened sometimes somewhere between 50 and 100).
I think most of have been aware of it if they were interested.
When I spotted that the answer was spoiled yesterday I deliberatedly used a few accounts I already had with names that hardly could be from a serious user, you will know when you lookup the table after my hint. (the three names are from the muppet show).
I'm still not sure if I did the right thing. However, if I hadn't some other fake people would have entered it, so I'm not feeling particularly guilty. I think it made something clear that otherwise might have gone unnoticed. I don't think I will do it again.
The advantage of the extension of the fastest solvers list to 100 in my opinion is that it's more revealing than hiding the fact that some people cheat.

Just a remark: we've been gradually been gaining lost ground. When that damned spoiler site began about two and a half years ago it wasn't uncommon for a problem to be spoiled somewhere between 20 and 40 solvers and after it was spoiled the number of 100%-ers raised to staggering amounts of 60 or so. Nowadays it's rather uncommon that all problems are spoiled before the next one appears. Also the number of fake members has shrunk significantly. However, there still are some members with different ideas about learning that seem to think it a good idea to spoil answers. In my opinion they are the real problem and not the oafs that think they are learning something by peeking into the fora and probably not even understanding what they are reading. So I'm not sure what flagging "cheaters" will have for positive effect. If outing some cheaters gives you a good feeling please go ahead, but let's not make it a witch-hunt.
We believe in being courteous and considerate to everybody, whatever he does to us. That doesn't take away that we're in our right to express our feelings about such behaviour.
Asking money beyond a certain point surely might diminish the problem. But what is globally a small amount? For us from the developed countries that are not jobless $25 might be neglegible for the fun we have but for a jobless person or a student in a less fortunate part of the world it might be a fortune.
Image

User avatar
Marcus_Andrews
Administrator
Posts: 1517
Joined: Wed Nov 09, 2011 5:23 pm

Re: Posting solutions outside of Project Euler's site

Post by Marcus_Andrews »

hk: I remember seeing those names elsewhere (although with slightly different formats) -- were your accounts solvers 49 and 50 of Problem 344?
Image

User avatar
hk
Administrator
Posts: 10875
Joined: Sun Mar 26, 2006 10:34 am
Location: Haren, Netherlands

Re: Posting solutions outside of Project Euler's site

Post by hk »

Yep. At that moment some suspicious people were filling in the last fifty of that list. So I short circuited that list then.
(actually 48,49 and 50 of that list)
(That was the moment I created some characters from the muppet show for special purposes).
Image

User avatar
Marcus_Andrews
Administrator
Posts: 1517
Joined: Wed Nov 09, 2011 5:23 pm

Re: Posting solutions outside of Project Euler's site

Post by Marcus_Andrews »

hk wrote: Asking money beyond a certain point surely might diminish the problem. But what is globally a small amount? For us from the developed countries that are not jobless $25 might be neglegible for the fun we have but for a jobless person or a student in a less fortunate part of the world it might be a fortune.
To give a point of reference: DavidM charged 1 Euro.
Image

User avatar
Francky
Posts: 90
Joined: Sat May 07, 2011 3:49 pm
Location: South of France

Re: Posting solutions outside of Project Euler's site

Post by Francky »

And what about :
1) to solve a 400+ problem, you need to have solved 50 problems or more. (or another bound)
2) to solve problem N, you need to have solved int(N*0.125) problems or more.
3) something else

This could prevent suspect accounts.

I remember there were some limitations when I create my account, perhaps still active.
I solved something like 40 problems just for me before create an account, and one day I decided to submit them.
I was stopped at 25, with a message (don't remember exactly), and I thought it was a good idea although I wasn't concerned by this guardrail. Those guardrails could be extended, no doubt about that.

Francky (no so active here now, it's true, PE becomes a bit hard for me, but I want to thanks a lot PE, it's here a learn so much, I think I'll come back when I'll have learned how to uses complex structures ...)
ImageEntia non sunt multiplicanda praeter necessitatem

wrongrook
Posts: 396
Joined: Sat Oct 17, 2009 10:39 pm

Re: Posting solutions outside of Project Euler's site

Post by wrongrook »

For what it's worth, I would prefer there not be such limits before people can solve recent problems.

If I recommend Project Euler to someone who wants to learn a new language then I suggest starting from the easiest problems, but if someone just wants to have fun and play with solving some mathematical problems then I suggest starting with recent problems.

Note that the difficulty of problems changes each week (I believe it is roughly a hard/medium/easy/medium cycle) so there are always recent problems of an appropriate difficulty level.

I think it is more enjoyable to solve a recent problem because you have the opportunity to participate in active discussion in the solution threads. Some of my favorite moments have been when I have solved a problem with some dumb ad hoc method, seen an amazingly elegant solution in the forum way beyond my current understanding, asked a question about it, and then been able to read a stunning explanation of some new method/maths that I had not encountered before.

I think it would put people off the site if they felt they had to churn through a huge backlog of questions before they could really participate in the community.

User avatar
Francky
Posts: 90
Joined: Sat May 07, 2011 3:49 pm
Location: South of France

Re: Posting solutions outside of Project Euler's site

Post by Francky »

But for one person who's able to solve a recent one (even an "easy"), it should be a formality of some hours to solve 50 problems. After that "check", He can play only with recent ones.
But I understand your argument.

another possibility :
4) Solve 50 problems before being able to submit for the last 5 ones.
((You can play with recent ones, but not on the hottest ))
ImageEntia non sunt multiplicanda praeter necessitatem

op=op
Posts: 3
Joined: Thu Oct 04, 2012 5:18 pm

Re: Posting solutions outside of Project Euler's site

Post by op=op »

What about sending your solution (code) to PE.

whakamaru
Posts: 47
Joined: Thu May 06, 2010 11:08 pm

Re: Posting solutions outside of Project Euler's site

Post by whakamaru »

Yes, I look for others solutions... 'trouble is... I don't understand either their logic, or their code. Inconsiderate bar-stewards!

martianred
Posts: 6
Joined: Tue Jun 11, 2013 9:12 pm

Re: Posting solutions outside of Project Euler's site

Post by martianred »

When can we have a non-cheater environment? :D

User avatar
hk
Administrator
Posts: 10875
Joined: Sun Mar 26, 2006 10:34 am
Location: Haren, Netherlands

Re: Posting solutions outside of Project Euler's site

Post by hk »

Probably when people like you stop paying attention to what you call cheaters.
When will people like you finally understand that a big part of the fun of spoiling answers is the never ending fretting of people like you?
If spoiling answers stops "cheating" will stop too.
Image

martianred
Posts: 6
Joined: Tue Jun 11, 2013 9:12 pm

Re: Posting solutions outside of Project Euler's site

Post by martianred »

It's not true
I have kept quiet for a long period, but spoiling answers still happened. How do you explain it?

That's enough. You can not stop cheaters but blame me.

Locked