Project Euler Forum

hk, thanks for the thoughtful response. I do like the update to the description for the Veteran class. The last bullet,

Continuing to enjoy the profound sense of self-satisfaction in cracking another "hard nut".

really nails it for me.

Yes if they were few you would perhaps be right. But they seem to concentrate on particular countries where they aren't few.
For instance out of the top 10 of your own country about 7 are at least of very dubious quality.
So I hope you haven't been comparing yourself with people that haven't been doing anything to place themselves above you several times during the last year.

If I may, I would call this a slight philosophical disagreement: should PE provide the data (in this case, the number of problems solved) and let the members beware of the attendant noise, or should PE in some sense try to vouch for the meaningfulness of the data, and if there's no way to do that, cease to provide it?

Yes, in the case cited, I recognized some dubious entries, and discounted them, though I hadn't put the number as high as 7 out of 10. Actually, it was kind of fun trying to make sure I hit 100% before even the cheaters could.

Thanks again for the update to the Veteran's description. PE continues to provide irresistably intriguing problems!

hk wrote:Regarding following who is consistenly solving new problems you can extract this from following the forum posts.
After all, if members are giving as much as they receive then they will make it a point of sharing their insights when they solve the problems. If they cannot be bothered ever sharing their methods then who cares if they continue solving problems or not; they haven't really engaged the spirit of the Project Euler community.

That's a bit harsh, especially when there have been many discussions here of trying to dissuade people from posting identical approaches in the problem forums. In the rare cases when I'm amongst the first few people to solve a problem, I still normally find that my algorithm is similar/identical to the couple of people who have already posted it, so I don't repeat the same thing to allow people with new approaches to post instead.

Does that mean I shouldn't be considered part of the Project Euler community? Or are we now encouraged to post identical approaches so that we aren't considered cheaters?

hk wrote:Regarding following who is consistenly solving new problems you can extract this from following the forum posts.
After all, if members are giving as much as they receive then they will make it a point of sharing their insights when they solve the problems. If they cannot be bothered ever sharing their methods then who cares if they continue solving problems or not; they haven't really engaged the spirit of the Project Euler community.

I think one should be careful to interprete a lack of involvement in problem threads as not being engaged in the PE Community. Many users do not have English as their native language, which is the de facto standard when writing in problem threads. One thing is understanding a problem written in English as a non-native speaker, another thing is to concisely express an own solution in a foreign language. The latter is much more difficult. The few times I have posted in problem threads, I have spend a lot of time drafting my input due to this barrier. I am so lucky that my native language (Danish) is related to English, so my own barrier is manageable. If I had Japanese, Russian or Chinese as my native language, the language barrier would most likely be larger.

Secondly, I get the feeling that sometimes it is forgotten that all the time PE gets new members starting from scratch. For the first 100 problems they only encounter a few open problem threads, and there is typically very little to add as compared to the realizations of the many previous users who have solved the same problem. The value of the realizations by the new user is not being acknowledged though, because another user also wrote about it four years ago. The new user may be just as skillful and bright repeating this insight, but has nowhere to vent this joy of mathematical or programmatically realization. Nor do they have the possibility to show solutions using tools and langauges, which did not exist when the thread locked. As a consequence several, out of frustration, start posting solutions elsewhere, in public, against the spirit of PE, because they are alienated by the locked threads and simply need to vent their insights somewhere.

@TripleM.
I'm apologising If I gave you this impression. I'm sure that people visiting this forum regularly will have noticed that you are doing a great job answering questions in the Project Euler thread here. But perhaps you have been too restrictive for yourself regarding publishing in the problems forums.
@Slaunger
The guy you asked to stop publishing solutions on his weblog used more arguments than "there is no room for me on the problems forums". His main argument was: everybody is doing it.
I don't think that opening up the closed problems fora again is the way to go.
When we closed the forum for problem 1 PE had about 20,000 members (that is about a year and a half ago) and the forum for problem 1 had about 1,000 posts.
Nowadays PE has about 120,000 members, so my guess is that that forum by now would have had 6,000 posts if we hadn't closed it, all containing something like the naive solution on the Wikipedia entry about PE that there is followed by a O(1) solution.
Do you really think that an open forum for problem 1 would have withheld this guy to start publishing his solutions on his weblog?
He then would have argued that his posts would have been drowned by the deluge of posts coming after his.
I also noted that some people are writing blogs about some very recent problems (as recent as 299!).

The whole point is that when PE was set up by Colin you had to pay to get some webspace. And if you had that who noticed it?
Nowadays everybody can start a weblog for free. What is more: Google considers a domain like wordpress.com as one domain, and one that is changing very often, so it crawles it very frequently.
If you put an entry on your log there it is findable within a few days or hours.
So my very personal view is that PE has to adapt to a changing world and an increasing "popularity".
And let's be honest: were the adaptations we made so shocking?
The main objective has remained the same: providing quality problems on a regular base and nothing has been changed to that. I consider the ever growing number of members as a proof we are succesful in doing just that.
However, the same growing number of members, the growing popularity and the changing world force us to do some adaptations.
I'm convinced that at some time in the future, if we are surviving/ are to survive, more adaptations will be necessary. Which? Time will learn.
And I truly hope that the members will try to follow the changes we make after ample consideration.

Hi hk,

hk wrote: @Slaunger
The guy you asked to stop publishing solutions on his weblog used more arguments than "there is no room for me on the problems forums". His main argument was: everybody is doing it.
I don't think that opening up the closed problems fora again is the way to go.
When we closed the forum for problem 1 PE had about 20,000 members (that is about a year and a half ago) and the forum for problem 1 had about 1,000 posts.
Nowadays PE has about 120,000 members, so my guess is that that forum by now would have had 6,000 posts if we hadn't closed it, all containing something like the naive solution on the Wikipedia entry about PE that there is followed by a O(1) solution.
Do you really think that an open forum for problem 1 would have withheld this guy to start publishing his solutions on his weblog?
He then would have argued that his posts would have been drowned by the deluge of posts coming after his.
I also noted that some people are writing blogs about some very recent problems (as recent as 299!).

The whole point is that when PE was set up by Colin you had to pay to get some webspace. And if you had that who noticed it?
Nowadays everybody can start a weblog for free. What is more: Google considers a domain like wordpress.com as one domain, and one that is changing very often, so it crawles it very frequently.
If you put an entry on your log there it is findable within a few days or hours.
So my very personal view is that PE has to adapt to a changing world and an increasing "popularity".
And let's be honest: were the adaptations we made so shocking?
The main objective has remained the same: providing quality problems on a regular base and nothing has been changed to that. I consider the ever growing number of members as a proof we are succesful in doing just that.
However, the same growing number of members, the growing popularity and the changing world force us to do some adaptations.
I'm convinced that at some time in the future more adaptations will be necessary. Which? Time will learn.
And I truly hope that the members will try to follow the changes we make after ample consideration.

I agree that locked threads were not the only argument used by that groovy user, but it was one of the arguments, and I think that if something can be done to lower the barrier or incitament for publishing elsewhere, it is worth considering. I acknowledge, that the one site, which is most destructive for PE concerning cheating is beyond reach from a reasoning point of view although they are using some of the same arguments from time to time when being critized about hosting public solutions.

I do not think either that opening the thread for problem 1 would solve that. Problem 1 is special, I am mostly thinking of threads following problem 10, and no, just opening them, would give chaos. One would have to introduce some kind of structure or other means of. E.g. that you make a post within a category, which could be either general or one of the preselected approaches (pencil/paper, Java, C++, Python and so on), and then a maximum number of posts within each category. And then a wider seletion of programming languages/methods than currently selectable, like, e.g., "Groovy". If the posts were categorized it would also be easier to find solutions in the preferred language and not be overwhelmed by a large quantity of "spam".

I think we all know that the quality of the posts in the problem thread is of very varying quality. There are lots of old posts, which are trivial or publications of obviously terrible brute force solutions. Now why should they be more qualified for staying there than new posts in a perhaps new programming language (I'd love to see problem 1 done in one of the esoteric programming languages "Whitespace", or "Brainf***"!) except for editorial convenience and the lowest administrative maintenence load?

I do not know how to solve that elegantly, as all solutions seem to bring extra work, like letting the problem thread posts be moderated and filtered by a problem editor, letting users vote on solution and those getting low votes are pushed out when new posts arrive in a "full thread". I just think there is a point in letting users share their revelations small or large in a "legal" manner here instead of venting it elsewhere, and that it is worthwhile discussing how that can be done.

Perhaps the way to look at it would be not to concentrate on the particular programming languages but at the algoritm behind it, like we did in the (albeit few) overviews. There we adopted the idea of using pseudo code.
Everyone that wants to make an overview for a problem is more than welcome. We simply haven't the time to do so. Problem development takes most of our time. (This is a standing invitation to everybody!)
But I understand perfectly that people are using the lower numbered PE problems to get practise in a particular programming language. So after all, a collection of decent solutions for a few lower numbered problems in a wide variety of programming languages would be a very good addition to the site. And if someone wanted to add his solution to this collection in a not yet covered language with the same degree of decency that should be welcomed.
However, before starting to collect solutions in different languages a overview should be made first, so that the implementations use the same algoritm.
But to be honest: that would not in the least cover the wish of newcomers to publish their own experiences.

As always, my head starts spitting out ideas the moment I read such a (maybe not intended) call for inventiveness in the problem thread structure. One is confronted with the fact that anything like voting or such (I think) will call for a huge database and lots of modifications, severing the load time and accessibility of the threads (and maybe PE as a whole). Therefore I think we need to search for a different way to categorise solutions. For example, pseudocode could be developed for different algorithms for a problem, after which users can select the algorithm they used. Since the number of different algorithms is usually quite concise, there could be a possibility for users to shortly describe any peculiarities they encountered (say 250 chars, or other suitable limit). It reduces the huge [ code] tags in the solutions thread, as well as giving new algorithms the chance to be added (how satisfactory is it to be with 5 people using a fast algorithm when 1000 users have solved the problem brute-force, for example). Unfortunately, to monitor new algorithms, there would be need for moderators with considerable knowledge of programming to determine if the used algorithm is new.

As for newcomers, I think the wish to publish your solution should vanish if you see that hundreds of people before you took the same approach. However, you would be allowed the small post (250 chars or so) to say how happy you were solving it.

This idea of mine has yet to deal with the frequent 'discussions' of problem threads where people ask for clarification/proof on fast algorithms (though I suppose it could be added to the pseudocode part).

I hope this is not too far fetched and detached from reality

Interesting thoughts hk and Lord Farin,

How about totally reinventing the way the problem threads are strutured and do it the wiki way!?!

If you have solved a problem you get access to a wiki page. One page titled "Problem XXX", where the equivalent of the pdf document developed for the first few problems are gradually drafted by those, who get access. This is a page, which can be edited and developed throughtout the lifetime of the problem by any editor who has solved it, containing a description of the mathematics, relevant figures and forumlas as well as pseudocode. It could have Implementation subsections dealing with specific implementations and their performance in different programming languages.

For each article page (which is initially blank, when the problem is new), there is also a "discussion" page as in any Wikipedia article, where the general solution can be discussed, people can present their variations thereof, vent their joy of having "made it", the stupidities they have made, realizations they have overlooked and so on. I.e, present the more social side of the problem solving process. The pages should never be locked. As on Wikipedia, archives should be setup when discussion pages grow too long, thus keeping only the most recent discussion immediately visible, letting older discussion be available via archive links, just as in Wikipedia. Of course these pages can in principle be vandalized, but just as in Wikipedia, some users would gradually attain a patroller or editor role, keeping the main problem articles clean and tidy. Users repeatedly abusing their rights to edit by vandalizing could be blocked for editing by administrator users just as in Wikipedia.

A transition could be made, where the existing problem threads are transferred to the wiki discussion pages, letting ofiicial problem pages grow from the editors interested. This is a much more dynamic and flexible solution than just having the static pdf documents, which was a good idea at the time, but not very flexible.

Of course it would be a large task making such a transition, it would need a lot of testing and surveying to pick the right wiki framework for PE. It could be a long term possibility though.

A wiki approach would allow for about everything I suggested above, however I am still a bit struggling with the quality control, in the sense that vandalising will probably be frequent. This then should be countered by the ability to flag documents that don't meet the quality standard, and also, a quite strict policy wrt. vandalising. Of course, this calls for a lot of moderation to be done. There would be a great need for a pool of moderators whom have access to all threads and watch over quality. Since our problem development team already has enough to do, this would call for a new set of moderators, or people who take over moderation of this forum. I don't think it would be a good idea to just up the work load of a very small set of people, whom also have other things to do on a day. Furthermore, the risk of the wiki page being copied to eg. Wikipedia would be quite high, since Ctrl-C, Ctrl-V from the editing screen would be enough.

Still, this might just be it, if executed properly.

Vandalism is possible, but i do not think one should be too worried about it. The frequency of school children wanting to write the name of a genital on some random page would not be a problem as it is on Wikipedia. Most wiki frameworks, like, e.g., Mediawiki has excellent tool for reverting or rolling back vandalism, and the few vandals can be blocked by a very few number of admins. I do think it would require a lot of moderators having access to all the problems. The potential editors who have unlocked a particular problem should be enough for assuring that the quality is OK using tools such as the edit history of a page and watchlists. The risk of vadalism and low quality edits would tend to lower the higher the id in good balance with the number of editors who actually have access. Thus I do not think it would be an extra burden for the moderators, who also develop problems.

Experienced editors could assist the current moderators with other boring tasks, such as putting code with dollar signs in code tags and actually release load from their shoulders.

Copying to Wikipedia would in theory be easier, but it would very fast be noticed by Wikipedia patrollers and deleted on there by admins as being out of scope and not fulfilling their notability criteria for inclusion. Such copies could be oversighted such that did not even appear in page histories, so I would not be too concerned about that.

My experience as an editor in the Wikipedia world (and also administering a company wiki at work) is that one should not be too concerned about giving edit rights to all who should have legal access. Of course the PE Community is very different from the Wikipedia community, and things could evolve differently. One could make a test with one problem and give access to a wiki page for that problem and see if it would work. Use taht as an example to see if this could be a good path to follow and learn from it to see how it should be taylored in its setup for PE. It could be one of the problems where we are past the most basic stuff, like, e,g,, the monopoly problem of finding the most probable places to land (do not recall the problem ID).

I am quite anxious to know how the development team regards this new suggestion... It seems as though we have a sort of workable concept here, and I would be happy to see a test case coming up. That is, of course, if the team isn't too busy with important stuff like designing problems. Thank you Slaunger for the wiki idea.

The team continue to discuss new ideas and we never ignore suggestions put in here. But lack of time is the issue. You cannot imagine how much time the rest of the team devote to problem development. As I work full time as a school teacher, and apart from the very occasional problem suggestion, I don't have much time any more to help with problem development. Everything that is done for you is handled by a team absolutely committed to delivering the very best problems in terms of style, challenge, and enjoyment. However, I do prioritise and spend what little time I have developing the scripts on the website. I have looked into "wiki" ideas, but I've found that the general software - and there are a frightening number of them! - are quite bloated and over-powered for our single purpose. Plus it would take far too much time and effort to integrate them into the website so that access is restricted depending on the data in the user database.

Having said that, it does not rule out the possibility of me creating an interface which would operate in a similar manner to the forum on the main site, allowing a "public" page associated with each problem to be edited. One of the challenges is how editing rights are managed; for example, what happens if two members are simultaneously editing the same page? But more important is managing the quality and efficiency of content. It could just turn into a single page that resembles several pages of the forum. Is that going to be any better than the forum we already have? As you can see the purpose of the overviews, when we introduced them, were a step in this direction. But as has already been explained, lack of time has restricted the output of these documents.

I really do like the idea of opening things up and I can assure you that we are not ruling out your suggestion. We will certainly continue discussing the matter.

You and the team have my deepest respect for the resources you put into PE, especially the development of new problems. Each and every problem I have encountered so far was been very professionally thought out. Personally, I have learned things about mathematics (and expecially combinatorics and number theory) and programming methods, which I would have thought I could never had learned. This has only been possible by going through the gradual process of first solving the first problems and gradually build up the skills. I really think it is true that if your are patient and persistent with the problems you can in the end solve all of them by learning along the way. It requires a lot of skill and time to develop problems which makes this possible!

I fully understand the workload issue. Maybe delegation of certain types of assignments/tasks to other PE Community members could be an idea? To ask if anyone could help solving this or that technical issue, such that the core team could better focus on its core capability, to continue developing these cool problems. For instance, I agree with you that implementing a wikiidea and integrating it with the PE pages is a *very* big task. It is not simply a matter of installing a mediawiki server on some server and boot it up. There are very many wikis around and another one may be better suited for the task. It takes studies of various technical possibilities, consideration of the special needs (access only for those who have solved a particular problem), the technicalities of the interface to existing systems, test and adjustment of the process as one gets some experience with how users would use it.

It is also not an urgent need. It is ideas for which there is time for them to evolve, e.g., by community discussions such as these. Of course hearing the opinions from more than four different persons could also add new perspectives, which could help mature some ideas.

At first I have been a bit disappointed by these new "hidden" profile pages... But the ranking by speed for the fastest solvers is definitely the best possible ranking, and (I hope) prevents from cheating.

Small suggestions:
- I only care about the profiles of my friends. Maybe we could have the possibility to have (up to 10 or 20) "friends" from which we can see the profile...
- Maybe you should reserve the access to the "Veterans" section to people who at least once have been in the fastest 20 solvers (which I confess hasn't been my case since problem 273)...

I have really learned a lot by solving problems on PE: thanks a lot to all the PE team for this wonderful website!

I do agree and like everything of last PE modifications.

If I can say only thing, I'd like that first fifty solvers could get points from 50 to 1.

Twenty is really too challenging (at least for me ).

Ciao.
Taifu.

I would like to echo Taifu's suggestion. If we can't see the total number of problems solved anymore once we reach Veteran level (which I still don't like), then at least let's increase the number in the Fastest Solvers tables to include more people. It's just not quite as much fun anymore without seeing publicly the 100%'ers. I'm really not inclined to start cancelling social engagements on the weekend to have a shot at the 20 fastest.

cheers, Stache

I agree with Stache.
As an alternative suggestion, perhaps a list containing the number of veterans having solved n problems (300 <= n <= max problem) would be acceptable for you to provide. It's anonymous like the "solved by" numbers in the problem list, while it doesn't leave the veterans floating in a complete vacuum (zero-info) Euler space.
For me, I would appreciate such a list very much.

Currently we are discussing what further changes to the website will be made based on reactions in this thread.
From problem 304 on we will record the first 50 solvers.
However, for the time being, we will keep publishing the first 20 only and the Eulerians table will be based on the first 20 for the time being.
In any case no data will be lost from problem 304 on by starting logging them now.
What we will do with these data and if we are going to do anything with these data we'll decide later as not all of us are available at the moment and we want to think about changes thoroughly before the necessary implementation steps are taken, that by themselves also need their time.
So don't expect many changes before early next year.
Everybody interested can make a rough guess about his place on the fastest 50 list by himself.

riemann wrote: As an alternative suggestion, perhaps a list containing the number of veterans having solved n problems (300 <= n <= max problem) would be acceptable for you to provide. It's anonymous like the "solved by" numbers in the problem list, while it doesn't leave the veterans floating in a complete vacuum (zero-info) Euler space.
For me, I would appreciate such a list very much.

I did a quick survey:
(298 8)
(299 10)
300 12
301 16
302 8
303 22
total 58

Update 8 oct 2010
300 6
301 13
302 6
303 12
304 25
total 62

Update 16 Oct 2010
300 6
301 11
302 4
303 8
304 10
305 30
total 69

Update 23 Oct 2010
(298 7)
(299 6)
300 2
301 12
302 6
303 3
304 9
305 14
306 26
total 72

Update 6 Nov 2010
(298 6)
(299 6)
300 3
301 11
302 6
303 3
304 7
305 8
306 7
307 18
308 18
total 81

Update 12 Nov 2010
300 4
301 10
302 6
303 5
304 7
305 6
306 7
307 18
308 3
309 18
total 84

Update 19 Nov 2010
(298 6)
(299 5)
300 5
301 6
302 7
303 5
304 6
305 8
306 8
307 11
308 7
309 8
310 18
total 89

Update 26 Nov 2010
300 6
301 6
302 5
303 4
304 5
306 10
307 10
308 5
309 1
310 18
311 14
total 91

Update 6 Dec 2010
300 6
301 3
302 10
303 2
304 6
305 4
306 8
307 9
308 4
309 4
310 5
311 2
312 15
313 18
total 96

Update 14 Jan 2011
300 8
301 4
302 5
303 4
304 6
305 5
306 2
307 6
308 2
309 3
310 6
311 4
312 5
313 3
314 4
315 2
316 4
317 4
318 25
319 9
total 111

Update 22 jan 2011
300 10
301 3
302 4
303 1
304 5
305 9
306 2
307 6
308 1
309 4
310 6
311 4
312 3
313 2
314 4
315 4
316 6
317 1
318 21
319 7
320 10
total 113

update 28 March 2011
300 7
301 5
302 5
303 4
304 3
305 2
306 6
307 4
308 5
309 3
310 6
311 2
312 3
313 1
314 3
315 2
316 4
317 2
318 4
319 4
320 1
322 4
323 5
324 14
325 3
326 4
327 9
328 9
329 3
330 5
total 135

Update 6 April
300 4
301 7
302 3
303 2
304 6
305 3
306 4
307 5
308 3
309 4
310 6
311 2
312 3
313 3
314 3
315 1
316 2
317 2
318 4
319 4
320 2
321 3
322 3
323 6
324 11
325 3
326 6
327 5
328 3
329 6
330 10
331 4

Update 13 April 2011
300 3
301 7
302 4
303 1
304 2
305 5
306 4
307 7
308 4
309 2
310 6
311 3
312 3
313 3
314 1
315 4
316 3
317 1
318 3
319 4
320 1
321 4
322 4
323 5
324 12
325 1
326 7
327 2
328 4
329 3
330 10
331 4
332 7

Update 20 April 2011
300 5
301 7
302 4
303 2
304 2
305 2
306 6
307 4
308 6
309 3
310 6
311 1
312 5
313 1
314 4
315 2
316 4
317 2
318 0
319 6
320 1
321 3
322 5
323 3
324 10
325 2
326 8
327 3
328 4
329 2
330 7
331 6
332 6
333 7

Update 7 May 2011
300 5
301 6
302 5
303 1
304 1
305 3
306 4
307 4
308 3
309 2
310 5
311 1
312 6
313 1
314 4
315 1
316 5
317 4
318 2
319 2
320 4
321 4
322 5
323 3
324 9
325 2
326 6
327 2
328 3
329 3
330 4
331 0
332 17
333 2
334 1
335 4
336 7

Update 14 April 2011
300 4
310 7
302 4
303 2
304 0
305 3
306 5
307 3
308 2
309 3
310 6
311 0
312 5
313 1
314 2
315 5
316 3
317 3
318 2
319 4
320 3
321 3
322 8
323 3
324 9
325 1
326 5
327 4
328 3
329 2
330 2
331 2
332 18
333 2
334 2
335 2
336 3
337 6

Update 21 May 2011
300 4
301 7
302 4
303 0
304 2
305 3
306 5
307 2
308 3
309 3
310 5
311 1
312 5
313 1
314 1
315 4
316 3
317 2
318 2
319 2
320 5
321 5
322 5
323 6
324 10
325 1
326 4
327 4
328 3
329 0
330 2
331 3
332 17
333 3
334 2
335 2
336 1
337 5
338 4

Update 28 May 2011
300 4
301 7
302 4
303 0
304 2
305 3
306 3
307 3
308 3
309 4
310 4
311 0
312 6
313 2
314 1
315 3
316 4
317 2
318 3
319 2
320 3
321 5
322 5
323 6
324 11
325 1
326 2
327 4
328 5
329 0
330 1
331 3
332 16
333 4
334 3
335 2
336 0
337 2
338 5
339 4

Update 10 June 2011
300 5
301 7
302 5
303 1
304 2
305 2
306 3
307 2
308 1
309 4
310 5
311 1
312 5
313 2
314 3
315 1
316 4
317 1
318 2
319 2
320 2
321 6
322 4
323 5
324 7
325 6
326 2
327 2
328 4
329 1
330 2
331 2
332 17
333 2
334 2
335 2
336 1
337 0
338 0
339 3
340 3
341 8

Thank you for both replies.

EDIT:
... and thanks for the updates.
This page is now in my favorites.