tea.mathoverflow.net - Discussion Feed (Could we make "unanswered" work better?) 2018-11-04T13:38:42-08:00 http://mathoverflow.tqft.net/ Lussumo Vanilla & Feed Publisher Ilya Nikokoshev comments on "Could we make "unanswered" work better?" (521) http://mathoverflow.tqft.net/discussion/47/could-we-make-unanswered-work-better/?Focus=521#Comment_521 2009-11-19T09:57:47-08:00 2018-11-04T13:38:42-08:00 Ilya Nikokoshev http://mathoverflow.tqft.net/account/22/ You converted me to the idea that 7-day bounty period is too limited, so I posted a feature request on Meta.SE. @David, Any arguments against the stumper tag? no arguments against, but I think ... You converted me to the idea that 7-day bounty period is too limited, so I posted a feature request on Meta.SE.

@David,

Any arguments against the stumper tag?

no arguments against, but I think it's better to go slowly and explicitly ask people for comments with first few retags.

]]>
Andrew Stacey comments on "Could we make "unanswered" work better?" (506) http://mathoverflow.tqft.net/discussion/47/could-we-make-unanswered-work-better/?Focus=506#Comment_506 2009-11-19T00:38:59-08:00 2018-11-04T13:38:42-08:00 Andrew Stacey http://mathoverflow.tqft.net/account/4/ Oh no, it's back. Oh no, it's back.

]]>
Andrew Stacey comments on "Could we make "unanswered" work better?" (505) http://mathoverflow.tqft.net/discussion/47/could-we-make-unanswered-work-better/?Focus=505#Comment_505 2009-11-19T00:37:16-08:00 2018-11-04T13:38:42-08:00 Andrew Stacey http://mathoverflow.tqft.net/account/4/ I notice that the "have you considered a bounty" message has disappeared! I approve. I notice that the "have you considered a bounty" message has disappeared! I approve.

]]>
Andrew Stacey comments on "Could we make "unanswered" work better?" (503) http://mathoverflow.tqft.net/discussion/47/could-we-make-unanswered-work-better/?Focus=503#Comment_503 2009-11-19T00:34:17-08:00 2018-11-04T13:38:42-08:00 Andrew Stacey http://mathoverflow.tqft.net/account/4/ @David: I think that if bounties could be open-ended then I'd be more accepting of them. I think that we work on different timeframes to programmers and forcing us into their model isn't always ... @David: I think that if bounties could be open-ended then I'd be more accepting of them. I think that we work on different timeframes to programmers and forcing us into their model isn't always going to work.

@Anton: I want some finer way to find out if there's something still outstanding in a question. I think that the "hasaccepted:0" is too coarse for this as I think that on MO there'll be many more questions that just cannot have an accepted answer but still can be "resolved". The idea of writing a community wiki answer and then accepting that is good, but I suspect most people won't do that. If there was some way to get people to retag their questions after a week then that would be great. Rather than the "Have you considered accepting an answer or offering a bounty" message you could say "Is this question still unresolved? How desperate are you for an answer?".

]]>
Anton Geraschenko comments on "Could we make "unanswered" work better?" (497) http://mathoverflow.tqft.net/discussion/47/could-we-make-unanswered-work-better/?Focus=497#Comment_497 2009-11-18T17:57:42-08:00 2018-11-04T13:38:42-08:00 Anton Geraschenko http://mathoverflow.tqft.net/account/2/ The right thing to do if you don't get an answer you're satisfied with is not to accept an answer. Please go vote up my request for SE to include advanced search (already implemented in S[OFU]) so ... The right thing to do if you don't get an answer you're satisfied with is not to accept an answer. Please go vote up my request for SE to include advanced search (already implemented in S[OFU]) so that you can use the search term "hasaccepted:0" to find questions without accepted answers. See this post for additional ninja search options available in S[OFU].

]]>
David Speyer comments on "Could we make "unanswered" work better?" (496) http://mathoverflow.tqft.net/discussion/47/could-we-make-unanswered-work-better/?Focus=496#Comment_496 2009-11-18T17:23:30-08:00 2018-11-04T13:38:42-08:00 David Speyer http://mathoverflow.tqft.net/account/23/ I find bounties a little useful, in that they tell me when someone was just tossing out a question and when they are seriously interested. But a serious question should be able to stay hot for more ... I find bounties a little useful, in that they tell me when someone was just tossing out a question and when they are seriously interested. But a serious question should be able to stay hot for more than a week. It is frustrating that if someone comes to this site a month from now and wants to sink their teeth into some interesting but difficult questions, they will have trouble finding good candidates.

Any arguments against the stumper tag? If not, I'll probably start using it, and do some retroactive tagging, this weekend.

]]>
Andrew Stacey comments on "Could we make "unanswered" work better?" (490) http://mathoverflow.tqft.net/discussion/47/could-we-make-unanswered-work-better/?Focus=490#Comment_490 2009-11-18T11:58:25-08:00 2018-11-04T13:38:42-08:00 Andrew Stacey http://mathoverflow.tqft.net/account/4/ I really hate the bounty idea. There are several aspects of the SO model that I feel don't carry over from the programming world to the mathematical one. One is the difference between a workable ... I really hate the bounty idea. There are several aspects of the SO model that I feel don't carry over from the programming world to the mathematical one. One is the difference between a workable solution and a correct solution. The fact that bounties have a time-limit works for programming, but it just feels wrong for mathematics. If I don't get a correct answer then I will not accept a partial answer.

On the other side, I have no particular interest in working on problems with a bounty or not. I'll think about a problem if it's interesting and if I think I can contribute. Knowing how much the other person needs the result would help me decide on how much effort to put in, but a bounty ain't gonna cut the mustard for me.

]]>
Ilya Nikokoshev comments on "Could we make "unanswered" work better?" (488) http://mathoverflow.tqft.net/discussion/47/could-we-make-unanswered-work-better/?Focus=488#Comment_488 2009-11-18T11:36:45-08:00 2018-11-04T13:38:42-08:00 Ilya Nikokoshev http://mathoverflow.tqft.net/account/22/ I've wanted to "accept" several answers to questions, or to say that this, that, and the other together make up the correct answer. One possible solution would be to post yourself a ...

I've wanted to "accept" several answers to questions, or to say that this, that, and the other together make up the correct answer.

One possible solution would be to post yourself a community wiki answer, summarize other answers there and accept it. You can then make a note that others can add to it directly.

]]>
Ilya Nikokoshev comments on "Could we make "unanswered" work better?" (487) http://mathoverflow.tqft.net/discussion/47/could-we-make-unanswered-work-better/?Focus=487#Comment_487 2009-11-18T11:33:10-08:00 2018-11-04T13:38:42-08:00 Ilya Nikokoshev http://mathoverflow.tqft.net/account/22/ would it be a good idea to create a tag for questions that have resisted serious thought, but have not been found as open questions in the literature? I'd like a way for the questioner to say ...

would it be a good idea to create a tag for questions that have resisted serious thought, but have not been found as open questions in the literature?

I'd like a way for the questioner to say "I'm still not happy with what I've gotten here"

The asker can post a bounty on the question. This mechanism signals to community that the question is a stumbler and that the asker cares about it (which is important since one possible reason why the question has no good answer could be because there was not enough work done to ask a well-defined question).

Usefully, there is a special tab for questions with bounty which seems to solve the problem of finding these questions + a special thrill one may feel when receiving a bounty should be a good motivator.

David, your idea about a tag seems to me to differ from the bounty mostly in the fact that the bounty question can last for at most a week and then one answer gets chosen automatically. One may see this as a bug, I see it as a feature -- if the question survives for a week without substantial progress, it's essentially an open problem; if there is some progress it probably makes sense to post a different question.

To summarize, I think the questions without an accepted answer should be an anomaly so participants should be encouraged to accept "this is the best we all could come up with" answers.

]]>
Scott Morrison comments on "Could we make "unanswered" work better?" (480) http://mathoverflow.tqft.net/discussion/47/could-we-make-unanswered-work-better/?Focus=480#Comment_480 2009-11-18T08:47:45-08:00 2018-11-04T13:38:42-08:00 Scott Morrison http://mathoverflow.tqft.net/account/3/ I agree this would be great, but can't think of any fix for now. We should put this in the long-term box. I agree this would be great, but can't think of any fix for now. We should put this in the long-term box.

]]>
Andrew Stacey comments on "Could we make "unanswered" work better?" (478) http://mathoverflow.tqft.net/discussion/47/could-we-make-unanswered-work-better/?Focus=478#Comment_478 2009-11-18T08:34:49-08:00 2018-11-04T13:38:42-08:00 Andrew Stacey http://mathoverflow.tqft.net/account/4/ I agree with the general point, but have no idea what's possible - from the outside it's impossible to tell what would require something deep in the code and what can be done by the admin. I find ... I agree with the general point, but have no idea what's possible - from the outside it's impossible to tell what would require something deep in the code and what can be done by the admin.

I find that the "a question has one answer" model just doesn't fit. I've wanted to "accept" several answers to questions, or to say that this, that, and the other together make up the correct answer.

So I'd like a way for the questioner to say "I'm still not happy with what I've gotten here" and for that to be what comes up as "unanswered". It'd also be useful if the questioner has to actually do something for this to apply, since if they aren't bothered enough about the question to keep coming back and saying "I'm still not satisfied" then I'm not bothered enough to try to answer it.

]]>
David Speyer comments on "Could we make "unanswered" work better?" (477) http://mathoverflow.tqft.net/discussion/47/could-we-make-unanswered-work-better/?Focus=477#Comment_477 2009-11-18T08:08:15-08:00 2018-11-04T13:38:42-08:00 David Speyer http://mathoverflow.tqft.net/account/23/ I'm thinking out loud here, and would like to hear other people's thoughts. There is a tab on the main page called "unanswered". If you click it, you get a list of questions for which no ... I'm thinking out loud here, and would like to hear other people's thoughts. There is a tab on the main page called "unanswered". If you click it, you get a list of questions for which no reply has a positive score (I think). If a question receives several partial answers, none of which completely solve the problem, it will be taken off the unanswered page because people will vote up the partial solutions. For example, the question about power sums, the question about 2-adic coefficients of modular forms and the question about BSD and the class number formula are all treated as answered, although no one has solved them. The open-problem tag deals with this for problems marked in the literature as open; but none of these are.

I think it is probably good that the unanswered button behaves the way it does. I think the intended use of the "unanswered" button is that someone clicks it and scans through to see whether they have anything obvious to add that others missed, and these questions won't meet that description. But it bothers me that these questions are being called answered when they are not.

So, two thoughts. First, can we rename "unanswered"? My first thought is "needs attention", but hopefully someone here can think of something better.

Second, would it be a good idea to create a tag for questions that have resisted serious thought, but have not been found as open questions in the literature? I'm thinking of "stumper", but maybe someone can do better. The trouble is that I can't think of a tag which is self explanatory, and there is no way to make a tag link to a FAQ.

]]>