Not signed in (Sign In)

Vanilla 1.1.9 is a product of Lussumo. More Information: Documentation, Community Support.

    • CommentAuthorquid
    • CommentTimeOct 5th 2012

    @Greg Kuperberg: I do not disagree that much with your final comment but just one point:

    The answer is interesting, even influential, and it wasn't going to be posted to some other web site.

    How do you know it was not going to be posted to some other website? Did you check with Dimitrov or somebody close to him? If not this seems like pure speculation, and to me (also as pure speculation) one that seems unlikely to be justified. Why wouldn't he for example have put up the thing on a personal website and posted a link in a comment on a blog, for instance?


    Greg, it seems to me that in the end things worked-out the way you want them to. It wasn't pretty but I don't think community moderation is ever going to be pretty.

    Ryan - Yes, that's fair. Maybe I mainly hope that this case can serve as a useful precedent for other such exceptional cases in the future, since it was my impression that the posting was thought to be at variance with "rules".

    And I concede that part of the problem was simply that version 1 of these Mochizuki questions may have been written in the wrong tone.
    • CommentAuthorgilkalai
    • CommentTimeOct 14th 2012
    Quid wrote: "In general, I think appropriateness of MO questions should not be a majority decision, by contrast if enough people 'veto' a question it should stay closed. I understand that Gil Kalai just formulated the opposing view that if enough people 'endorse' a question it should stay open, but I disagree, in particular for controversial question like the current one. If enough people see a problem, it is a problem (at least for these people) and the others should have a very good reason to ignore the concerns of these MO colleagues."

    Quid's position does not make sense in my view. MO enables people in our community to ask questions and get answers to questions that they find interesting. There are quite a few cases that there is no controversy at all about closing questions. There are fairly rare cases that there is disagreement among the eligible MO participants, and in these cases there is no symmetry between right of the OP who asked the question and other people who are interested in the answers to have the question remain open and be served by the MO framework and the rights of those who are against it to have it closed. There is no symmetry between the right to ask a question and the right to close a question. Moreover, the most common argument against some kinds of questions is based on the fear that there will be many similar questions. If we leave controversial questions open only after 10 people voted to open this will be a serious barrier.
    • CommentAuthordeane.yang
    • CommentTimeOct 14th 2012
    I endorse Gil Kalai's position on reopening questions. For me the value of MathOverflow is greatest when it pushes the limits of its original mission and allows people like Bill Thurston (alas, no more), Minhyong Kim, and Vesselin Dimitrov give beautiful discursive answers to not-sufficiently-narrowly-stated questions.
    • CommentAuthorquid
    • CommentTimeOct 15th 2012 edited

    @Gil Kalai: possibly you read my statement stronger than it was meant (which would be my fault as it is really not formulated well) but it might also be the case that you really think very differently about this than I do. In any case, let me retry:

    My statement was meant in the context of this discussion; so for this controversial type of question (at least mainly, or even only; to use 'at least' instead of 'in particular' would have been better). I agree as a completely general principle it would not be so good. And by controversial I did not mean each question that causes some discussion, but only those where some people express quite clearly that they are actually concerned about the (implications of the) existence of a particular question, and if you like add "in an immediate way" (so not just some long term effect, and so on).

    Conversely, it is hard to believe for me that you truly believe that even if say 50 users would say they find some question very problematic possibly even close to offensive, as long as 10 think it is good enough it should stay. I really do not think it is a good principle to moderate an internet resource according to the standards of the 10 most resistant users. In particular, having an absolute number for permanent openness would particularly help controversial things. (Numbers refer to 3k+ users, of course.)

    Side comments: while on the one hand I find the information Dimitrov gives quite interesting, on the other hand I am really not sure that it is appropriate to have this continuing near-real-time documentation (see the last couple revisions) and to possibly force the author into rapid responses. [Side remark, for those claiming it is so important this was/is on MO: do you think Mochizucki would not have posted the comments if it were not for MO? If no, it seems hard to maintain the importance claim. If yes, it seems there is a need to argue that MO should really be used to pressure authors into public comments.]


    Deane, the time just doesn't seem to be ripe yet to comment authoritatively on Mochizuki's work. However "beautiful" you think Vesselin Dimitrov's answer is, you can see for yourself that there is a recurring need to update his answer(s) (with an apology of sorts now at the bottom), and there is great potential for confusion here, certainly in the popular mind of general mathematicians.

    Update: that "apology" is now gone as of five minutes ago, since VD has updated again, retracting a lot of a recent update in view of comments by Mochizuki.

    • CommentAuthorgrp
    • CommentTimeOct 15th 2012
    I think it is past time for a referendum, or at least opinion poll, on what changes to
    make to the mission, intent, and basic policies on MathOverflow. If we are going to
    turn it into something like present day sci.math, let's do it with foresight and planning.

    Many may object to that last statement. I will agree that it is objectionable, but I say
    that it is highly relevant. When the community is strong and thriving it can
    survive the occasional disagreement or bit of spam. When the noise level rises high
    enough, (I assert without proof) the health of the community declines quickly, and
    my fear is that that threshhold is not far away.

    I appreciated Justin Moore's information regarding amenability of F (and was mildly
    saddened to hear of the error), but I thought MathOverflow was the wrong place
    to do more than the most basic status updates (which I think I should be "I think
    I have it; see link", and/or "No I didn't, see link"). While I am glad Vesselin Dimitrov
    is contributing time to clarify the work of Mochizuki, I still think MathOverflow is not the place for a
    blow-by-blow reporting of the analysis. MathOverflow can and perhaps should
    have a link to such, but many of us do not have that detailed of an interest in the
    situation, and do not like the question to keep bubbling to the top as a result.

    Right now, having such happen for one question is tolerable if not desirable; what
    about for five such questions? Or ten? How will that affect the community?

    I am all for having a place for such discussions, even if people want them collected
    and not on separate blogs. I am considering setting up a forum for such when I am assured
    I will get enough moderation help. However I see encouraging such efforts on
    MathOverflow a potentially destructive force. With apologies to and some sympathy
    for Greg's position, this to me is why the rules are there and the sake for which they
    should be followed.

    Gerhard "Ask Me About System Design" Paseman, 2012.10.15

    I am considering setting up a forum for such when I am assured I will get enough moderation help.

    That's been the problem all along with this. The number of people who have offered to set up such a forum is considerably larger than the number who have offered to moderate it (in proportion, that is).

    • CommentAuthorHJRW
    • CommentTimeOct 15th 2012

    As the number of edits to Dimitrov's (undoubtedly mathematically excellent) answer tends to infinity, the position of those who think that MO is the right place for this seems, to me, less and less tenable.

    • CommentAuthormghjfg
    • CommentTimeOct 15th 2012 edited
    Vesselin Dimitrov's oft revised answer has been copied to the nLab:

    Some characters don't display. It was a pure copy and paste so there must be differences between MO and nLab.

    Thanks for saying so, mghjfg. Wow.

    • CommentAuthordeane.yang
    • CommentTimeOct 15th 2012
    I'm willing to accept the consensus here (which appears to be different from my view). I just want to say that I don't understand why it's a problem for Dimitrov to edit his answer or add new answers continually. Even wikipedia, which tries to be a a definitive reference, is being continually edited. What's wrong with MathOverflow answers, at least the ones to open-ended questions, being edited a lot?
    • CommentAuthorgrp
    • CommentTimeOct 15th 2012
    Deane, in principle, there should not be much problem. In practice, the software will convert his answer to community wiki, which may not be a smart idea. Also I (and I suspect many others) have been conditioned either to accept bumps to the top as substantial edits, or to ignore such questions entirely but wish they would go away. The disappointment level or frustration level with the question (and then the interface, forum, and community) then grows. This would not be a major concern if MathOverflow had a million users with 1% dropping out and then returning on a periodic basis. The community supporting this forum is much smaller, and such annoyances will add up and have a greater effect than many here might suppose.

    Gerhard "Ask Me About System Design" Paseman, 2012.10.15
    • CommentAuthorgilkalai
    • CommentTimeOct 15th 2012
    I think that Quid made an important distinction. When it comes to offensive questions I think that those should be closed and even deleted. MO cannot be a stage for offensive behavior and offensive behavior should not be tolerated. When it comes to problematic questions (but not offensive) such that there is a large number of voting members that want to see open, then I think that the question should remains open. This is a very basic matter of tolerance. In such a case there is no symmetry between those, even a minority, who wish to see the question answered and those who think that it is problematic for MO. Moreover, often the main reason for not allowing a question is that it may encourage similar questions and that it will be hard to draw the line. This argument does not apply to rare cases of open/close "wars", and the proposed rule also draws a line. (OK, maybe the number of such cases will go up if my suggestion is adopted but not by much, I think.)

    (It is especially unreasonable to talk about opinions of members of the community and majority/minority decisions in cases of questions with strong records of upvotes and views. )
    • CommentAuthorJDH
    • CommentTimeOct 15th 2012

    In regard to Deane's comment about "consensus", let me say that I agree with Greg and Deane. I think MO is great for the purpose as in this question, especially when it attracts the participation of extremely knowledgeable users. Since that seems to make at least three of us with this view, I think there is no consensus here, and this discussion is at best inconclusive.

    My view is that MO is strengthened by such questions as the one under discussion and many similar ones. The software appears to handle closing disputes in an elegant, robust way, without much need at all for meta-discussion about closing. Namely, controversial questions are closed and reopened for periods of time roughly in correspondence with the depth of their support by the community. If you oppose a question, vote to close it, and if you support it, vote to reopen. The question will settle into a status that reflects the community norms.


    without much need at all for meta-discussion about closing

    JDH (Joel), I think rational discussions in case of disputes are generally a good idea. It can happen that someone gives reasons that hadn't been considered by others; it's so easy to click on buttons irreflectively.

    • CommentAuthorquid
    • CommentTimeOct 16th 2012

    A couple remarks:

    1. Just to underline what grp said on edits. The issue with repeated edits is that the default is that everybody (visiting the site in a certain not very short time-frame after it) gets notified of the edit (via the question being bumped). These notifications via bumps can be (and are for me, sometimes at least) a considerable annoyance, and perhaps more importantly take away visibilty from other questions (if it is not on the front-page anymore it is gone forever, seems almost true for some more typical questions). True, if the question is sufficiently high profile and/or this is sufficientlty rare it is not a big deal, as grp said. But it clearly shows that MO is (by design!) not made for this type of content; and if it is not very rare it will be problem.

    2. I strongly agree with Todd Trimble's last comment. Moreover, I do not at all share JDH's point of view that the software handles these disputes in an elegant way. To me frequent changes in the open/close status are really the worst case scenario. It feels absurd to have some answers but not others only due to the fact when the users wanted to post them (on a relatively short time scale), or to make it necessary to make some strange effort to get an answer in. So then I need to watch the site closely not to miss the one hour when the question was reopened before being reclosed to get in my answer. Perhaps initially only a very sketchy answer so that even after the question was closed I can expand it. Or maybe leave a sequence of comments on a closed question with the request that somebody when the question happens to be open reposts this as a CW answer. Or perhaps then usesrs could start coordinating/announcing when exactly they will vote to reopen, so that then people wishing to answer can prepare and give the answer while the closers prepare to reclose as quickly as possible. This is a bit exagerated but effects of this form do occur. For example Minhyong Kim asked to be notified in case of reopening. And in general the argument with the preriods of time open and close seems somewhat besides the point. Since there is no symmetrie at all. To get the point across another exageration, if a question were open/closed every other hour say, to me this seems almost as good as it being open, and it being closed sometimes nothing but a useless annoyance (without any actual effect).

    3. Yes, Gil Kalai, "often the main reason for not allowing a question is that it may encourage similar questions" but not always. And for these rare cases I am in favor of a more veto-based system. Also, tolerance is not a one-way road. Where is the tolerance towards or consideration for those that get majorly annoyed by seeing certain question on MO that are there just because some others find them somewhat interesting or amusing?
      (The present case is not a good example since I think some really also care a lot about having the question, but for some question it seems really hard to make an argument that anybody has more than a idle or passing interest in the matter, though many might share this interest.)

    • CommentAuthorbsteinberg
    • CommentTimeOct 16th 2012
    I think Dimitrov's answer has become the exception that proves the rule. I think it confirms why papers and issues with papers should be discussed first privately with the author unless the author shows an unwillingness to respond to issues. The current situation seems to be that Theorem 1.10 claims something too strong to be true. It seems the author now states that this is the case, but that his proof gives a weaker statement which is enough for the main result. This may or may not be the case. But having it be public knowledge has probably slowed down the processing of Mochizuki papers. Most likely a number of people going through the paper have halted because of the possible error and probably some will not resume to check if his proof really gives ABC because some credibility has been lost. Had instead this been handled privately and with the result being that Mochizuki puts a replacement on the ArXiv thanking Dimitrov for pointing out the misstatement it would less likely deter people from reading ahead.

    Regardless of how this particular example works out, it highlights why some people had concerns about using MO or other public fora as places to discuss the correctness of papers. It is not all that uncommon that papers state theorems in stronger forms than what the proof gives. I could easily imagine that if ArXiv papers had a mathoverflow style commenting/reviewing system that you will get comments saying Theorem X is wrong here is a counterexample. But maybe Theorem X just overstated what was really proved, which is a good result. The comments saying Theorem X is wrong may kill future readership of the paper and doom its chances for acceptance in a good journal. It would be much better if the author was privately contacted so that they can fix these problems in a non-public forum so as not to put a shadow on the final product which may be good.
    • CommentAuthorWillieWong
    • CommentTimeOct 17th 2012

    The number of people who have offered to set up such a forum is considerably larger than the number who have offered to moderate it (in proportion, that is).

    If all is needed is a Jim Lehrer style of moderating, I am sure many (including myself) would not mind volunteering our services. But I suspect that many expect the moderators to have some degree of authority in the subjects being discussed. This drastically limits the pool of available candidates.


    Jim Who? There's only one Lehrer and his name isn't Jim.

    (Seriously, I really do have no idea who Jim Lehrer is so don't know what sort of volunteering is being volunteered here.)


    Jim Lehrer.

    He moderated many presidential candidate debates.

    • CommentAuthorWillieWong
    • CommentTimeOct 17th 2012 edited

    Jim Lehrer is an American journalist who moderated the first debate this year between the US presidential election candidates. One of the few things that many pundits from all sides agreed about (concerning the debate) was that his performance was rather pathetic. His self-professed moderating philosophy is to (I paraphrase) "get out of the way and let people talk."

    • CommentAuthorEmil J
    • CommentTimeOct 17th 2012 edited

    (never mind)

    • CommentAuthorgilkalai
    • CommentTimeOct 17th 2012
    Quid: "Also, tolerance is not a one-way road. Where is the tolerance towards or consideration for those that get majorly annoyed by seeing certain question on MO that are there just because some others find them somewhat interesting or amusing?"

    This is a familiar issue regarding tolerance - tolerance towards the intolerant. "Being annoyed" by some question, which is often just a vehicle to force one's opinions on others, is not a good reason to close the question.

    In the rare cases where I voted to open is was because I thought that the academic merits (or clear potential) of a question are strong enough to leave it open and get those interested MO participants have the opportunity to answer, and read answers by others.
    • CommentAuthorWillieWong
    • CommentTimeOct 17th 2012 edited

    The number of people who have offered to set up such a forum is considerably larger than the number who have offered to moderate it (in proportion, that is).

    (Seriously, I really do have no idea who Jim Lehrer is so don't know what sort of volunteering is being volunteered here.)

    BTW, I think the reason that not many volunteers rose to the call is that most people would be hesitant to commit to work whose details are rather sketchy. It is one thing when the volunteers are "free to fail". It is a totally different thing if the volunteers are expected to be able to adhere to certain as-of-yet unstated expectations.

    You know, come to think of it, I will volunteer. Because if we just sit here and talk about it all day, nothing will ever happen. But on the one condition: Just like the original MathOverflow, such a forum should be set-up as an "experiment". We should have no expectations that it will succeed. And the first, say, 6 months (or a year, just some finite-and-not-too-large number of days), should be understood as a trial period; if it works, great. If it doesn't, the people running it should be free to shut down the forum (after dumping all the discussions to, say, nLab) at the end of the period by saying "sorry things didn't work out..." Seriously, why are we all so risk averse? Is it because MO is such a success that we are now afraid to fail? Is there actually something that we'd lose if we try to set up one of these sites and things don't work as great as we imagined, and we end up shuttering the place?


    You know, come to think of it, I will volunteer.

    Great! Any more takers?

    Just need a name, a basic colour scheme, and a logo.

    (Actually, we have an area on the nForum called "Preprints & Publications" which is precisely for discussion of preprints and publications. So we could do this without even setting up a new forum. But new fora are cheap anyway and maybe the nForum is too tightly bound to the nLab for this purpose.)

    • CommentAuthorHenry Cohn
    • CommentTimeOct 17th 2012 edited

    I think the first step towards setting up a forum like this would be clearly defining its scope and intended use. For example, would it be open to requests for clarification in cases where someone is having trouble reading a paper but there are no serious worries about correctness? If so, it would be extremely important to distinguish between these cases and questions of correctness. (Plus it would bring up the usual MO issues of how much effort someone should be expected to put in before asking.) If more or less arbitrary requests for clarification are allowed, then there's a risk they will function primarily as a way of publicly pressuring authors into helping readers with issues that really shouldn't require the author's intervention. That would make me unhappy,

    What I'd propose is a forum intended for only limited use. For example, it might be aimed only at looking into correctness for papers of exceptionally broad interest and importance (like Mochizuki's work). The ground rules might be that a discussion on a paper can be started only if the forum managers (and perhaps also the paper's authors) approve and if several well-known experts volunteer to moderate this specific discussion. That would be pretty narrow, but it would be very easy to run and it would avoid most or all of the potential problems.

    • CommentAuthorquid
    • CommentTimeOct 17th 2012

    Gil Kalai: This is starting to get a bit of a broad debate. So, in your opinion it is alright if somebody does not want to be offended.

    First, would I be intolerant if I would be annoyed by a question or find it very problematic because I think somebody else could find it (or reactions it created) offensive, while not finding it offensive myself?

    Second, not to be intolerant, do I in your opinion really have to wait until the potentially offensive content appears or is it alright to shut-down something before somebody actually gets offended?

    Third, I could well imagine that Mochizucki meanwhile already got offended by some contribution (if he should follow this). [To be clear what I mean: the way his "comments" where initially discussed is in my opinion potentially a lot more offensive than many other things where some people here say it is/could be offensive.]

    So, am I now intolerant if I do not want this question?

    • CommentAuthorYemon Choi
    • CommentTimeOct 17th 2012

    Seriously, why are we all so risk averse? Is it because MO is such a success that we are now afraid to fail?

    Well, I can't speak for anyone else, but I barely have time (in semi-work mode, that is) for MO, let alone anything else. (This is one reason I get somewhat exercised about what should or should not be on MO.) To quote the words of the Sage of Intake:

    Well I used to do the I Ching

    But then I had to feed the meter

    Now I can't see into the future

    But at least I can use the heater

    • CommentAuthordeane.yang
    • CommentTimeOct 17th 2012
    I seem to arrive at the opposite conclusion to Yemon. I barely have time for MO and definitely none for anything else. I'm assuming that this is also true for many other senior mathematicians. So I want MO to accommodate as flexibly as possible a broad range of substantial mathematical discussions and not limit itself to its original purpose of answering mainly precise specific questions that have a known unique answer. But as I've said before, if the consensus is to protect MO's original vision and purpose, I will abide by this, since that is still a very useful thing.
    +1 Deane Yang
    • CommentAuthorYemon Choi
    • CommentTimeOct 17th 2012

    Thanks for the alternative perspective, Deane. I think I meant something along the following lines: I barely have time for that subset of MO which I find most appealing, and hence -- perhaps unfairly, and certainly somewhat irrationally -- feel other "off topic" questions as a distraction. However, I am very aware of this being just a personal preference. I have been sitting on the fence regarding VD's updates regarding the Mochizuki papers.

    • CommentAuthorYemon Choi
    • CommentTimeOct 17th 2012

    Also, although I think Deane understood this, my remark about not having time was more directed, tongue in cheek, at Willie's rhetorical question "why have people not set up a forum as an experiment". I have to feed the meter.

    • CommentAuthorquid
    • CommentTimeOct 17th 2012

    I find it potentially misleading to bring in, albeit indirectly possibly even inadvertently but still, seniority here, since there are quite a few senior mathematicians quite strict against various kinds of content.

    In addition the argument that it takes more time to visit two sites seems in the end quite weak, although I can see why subjectively one sees it so, since going to the site is really not what takes the time in the virtual world (and using feeds or alike this could even be further optimized).

    Yet, as I somehow expected, it seems a main obstacle to all ideas of creating something new (and in this way also avoiding friction on MO) could be outright refusal by some (since I am somehow replying to Deane Yang, I would like to explict exclude him from the some in view of his balanced position on the matter) to move this content there.

    One can already notice this with academia.SE, essentially always there is some reason constructed why tranfering to there is supposedly a non-option.


    In reply to Deane's comments:

    I seem to arrive at the opposite conclusion to Yemon. I barely have time for MO and definitely none for anything else. I'm assuming that this is also true for many other senior mathematicians. So I want MO to accommodate as flexibly as possible a broad range of substantial mathematical discussions and not limit itself to its original purpose of answering mainly precise specific questions that have a known unique answer.

    I find MO less and less useful. Part of it is that my real questions don't get answers (though now that Peter Michor has joined I'm hopeful once again). However, I feel that in principle it is still a useful resource and so I keep an eye on it and answer questions when I'm able to do so in a timely manner. What makes it possible for me to do this is that I can quickly scan through the list of New questions each morning to see if there are any that catch my eye. Because the level is meant to be high, if a question looks interesting from the title and first line then I can assume with a reasonable probability that it will be interesting and so worth my time clicking through.

    If it fills up with garbage, the likelihood of me being able to maintain this "light touch" will converge to zero. It will take more of my time to filter out the garbage than the few shiny stones that are there are worth. I've almost given up reading maths blogs because I don't have time to filter out what is interesting-but-ultimately-useless from the gold. Instead, I concentrate my attention on those places where I know that I have a high chance of fruitful mathematical interaction.

    I can understand the "let's have everything in one place" attitude. I once shared it. But I don't any longer and the reason I don't is that I found it impossible to fit the kind of discussion that I wanted to have into the machinery. Rather than fight it, I figured I'd let MO be what it was best at and find other venues for other discussions: for maths, I've set up the nForum; for publishing, I've set up the publishing forum; for climate change, I've set up Azimuth's forum. (Anyone notice a pattern there? If anyone's interested in teaching get in touch.) Now I regard the "let's have everything in one place" attitude as really laziness. It says, "I can't be bothered to find out where the best stuff is so I want it all in one place, no matter what comes with it.". Shall we integrate MathSciNet with MO? The arXiv? The BBC? Where does it stop?

    No, I want MO to remain focussed on what it does best because there is something that it does best and that nothing else does. If MO gets too broad, it will stop doing that one thing well and then we'll have nowhere for asking focussed, answerable questions.

    Thanks and +1 to Andrew Stacey. I disagree with your position, but I now have a fuller understanding of why people disagree with me.

    Thanks and +1 to Andrew Stacey. I agree with your position.

    For those who want to be able to find everything in one place, am I right in assuming that it's possible to follow MO and the various fora Andrew mentioned via RSS? (I'm not interested in doing so personally, so out of laziness/busy-ness I'm not taking the time to find out the answer to that question myself.)


    Mark, yes. MO has lots of feeds and so do the various fora that I run. Just about everything has RSS feeds nowadays.

    Frank. Great. That's what I was aiming for. I don't explain my views in order to convince anyone that they are right. I explain them with the aim of showing that they are reasonable and that therefore my actions, being based on my views, are also reasonable and so should not be dismissed out of hand.

    • CommentAuthorgilkalai
    • CommentTimeOct 18th 2012 edited
    Andrew said:
    If it [MO] fills up with garbage, the likelihood of me being able to maintain this "light touch" will converge to zero. It will take more of my time to filter out the garbage than the few shiny stones that are there are worth. I've almost given up reading maths blogs because I don't have time to filter out what is interesting-but-ultimately-useless from the gold. Instead, I concentrate my attention on those places where I know that I have a high chance of fruitful mathematical interaction.

    We talk here about the rare cases of controversial questions, probably less than 1% of all questions that were closed. I am not aware of any single question of this kind that can be referred to as "garbage,"
    and it is not clear how setting a more reasonable attitude for such rare cases of controversy can "fill MO with garbage". (It is certainly completely unreasonable to refer to questions like the one about the ABC conjecture as garbage.)

    @gilkalai, I didn't think the main point of Andrew's post was to discuss "garbage". I read it more in terms of the reality of time limitations and "getting bang for buck" in terms of fruitful mathematical interaction, and particularly the kinds of online discussion he wants to be having. MO serves a particular and sometimes very useful function, but in view of the way it's set up, it's not well geared towards having lots of back-and-forth discussion. Indeed, from the faq:

    MathOverflow is not a discussion forum. As a side-effect of being very good for to-the-point questions and answers, the Stack Exchange software is bad for disscusions and designed to minimize them. There's a place for discussion about mathematics, but it isn't MathOverflow. Blogs and threaded discussion forums are a more appropriate place for discussions.

    Regarding the test case of the popular post on abc and Mochizuki, it seems clear that however good the answers are considered to be, there's not an actual dialogue or multi-person conversation taking place. Yes, the answerers refer (usually briefly) to one another's answers, but they are not asking questions of each other. What we have is several monologues written at an expert-level about partial understandings of what is going on in a paper nobody can speak about with definitiveness, amid a chorus of general celebration about how good the answers are. It's nothing like the back-and-forth high-flying discussion that you would expect in a working seminar, and that the paper really needs in order to be properly evaluated. And that's to be expected -- the software is not geared toward having such a discussion.

    • CommentAuthorgrp
    • CommentTimeOct 18th 2012
    I would like to reassure Gil that we are not considering the content of Vesselin Dimitrov's commentary as garbage, nor the edited question. The original question, while also not garbage, had material that was questionable for MathOverflow, and Dimitrov's updates are being done at a frequency which (in my view) is detracting from his efforts.

    Again, a small amount of this can be tolerated, but I and others argue against not the activity itself, but the activity occurring on MathOverflow in a larger than tolerable amount.

    Gerhard "Ask Me About System Design" Paseman, 2012.10.18
    +1 Gil, Frank - agree, +1 Andrew - disagree :)

    @Andrew, I think everybody is thankful for you for setting up forums, and yours voice will be heard and respected in any discussion. But I honestly do not quite understand your point, you write : "I find MO less and less useful. Part of it is that my real questions don't get answers (though now that Peter Michor has joined I'm hopeful once again). "

    What do see as a reason for "your real questions do not get answers" ? (By the way what you mean by "real questions" ?). "Less and less useful" - what changed in MO ? in what time period ? (I am here for 1 year, I do not see changes).

    I do not think that a small amount of "soft-questions" may affect "yours real questions do not get answers" as well as make MO "less useful". So I cannot accept as a valid argument for vote-to-close that small amount which deviate from being technical question.

    May be you mean that level of questions is going lower on MO ? If yes what is the reason you see for this ?

    More generally - what kind of questions you think are appropriate for MO ?
    (Can you give some words similar to those that given here
    by Joseph O'Rourke and JDH)
    • CommentAuthorquid
    • CommentTimeOct 18th 2012

    Gil Kalai: perhaps let us focus on "interesting-but-ultimately-useless" (admitting we use a not to broad definition of useful) as a more precise description of a certain type of content; and perhaps let us also not focus too much on ABC, since for this (type of) question plenty other arguments can and were already made why one might not want it on MO.

    Let us perhaps look at Music: mathematical point of view somewhat recent with ten reopen votes one from you.

    I was not even totally against this question, but the way it started and developped is really not something to be encouraged, IMO.

    And in addition it raises a curious technical issue if ever your rule should be implement: it might turn out to be a good strategy to start with a really bad version, and to incrementally improve to accumalate reopen votes on the way, to then be 'save forever'.

    More generally, it will be a good strategy to create lots and lots of noise around questions to over time 'find' ten votes (it is after all only about 3% of the voting population) of not so frequent visitors and/or voters.

    In any case this rule (with an absolute number!) seems the more unreasonable the more I think about it.

    • CommentAuthorgilkalai
    • CommentTimeOct 18th 2012
    Hi Quid, yes it is a good idea not to focus only on ABC. When it comes to the ABC question, my impression is that this question (and its answers) is among the most impressive successes of MO. So even for those who objected to the question in the beginning it does not seem reasonable to insist on these objections now (for this question) and even more unreasonable to still support closing the question.

    The music question is a good example, but we probably need more than one example, can you mention a few more? (Let us put aside big list questions and very soft questions that we often discuss elsewhere.)
    There were not that many questions which were controversial and had open/close war (and my overall position is to keep every such question open unless it is offensive).

    Maybe you should explain what you mean by "interesting-but-ultimately-useless."
    • CommentAuthorquid
    • CommentTimeOct 18th 2012

    Hi Gil Kalai, briefly regarding ABC, as I tried to explain partly in another thread regarding ABC I find it important to distinguish between a) the actual question in its original form, b) the edited version and the initial answers c) the about current situation d) the future of the question.

    a) was terrible, b) was admittedly good (could also be we just got lucky this time; also I stopped complaining then) c) some issues surface, d) we'll see.

    Regarding c) I consider it as potentially quite problematic if when some well-known mathematician releases some clarification to his proof, then basically instantly on MO somebody flippantly comments that this seems to be essentially nonsense. (Okay, these "comments" seem to have been unclear, but in my opinion you either need to ask for clarfification privately and wait for the reply or at least give the benefit of the doubt, but not choose an interpretation that makes this look nonsensical.)

    I do not know if this is so unreasonable. Okay, one could say the public interest in these news is so big that side-effects like possibly insulting the author need to be tolerated, but then perhaps one does not have to share this opinion.

    Regarding the other matter: if you exclude lists, and very soft-questions, and things that are llikely to be offensive from your rule, then I would say that I in principle would accept the underlying idea. (For reasons explained I would see issues with "writing it down somewhere", as then some might start trying to game the system.) But, we could agree on it as some informal guiding principle. (Which in someways for example I try to follow anyway already.)


    I find the question and answer format of MO lousy for discussion type questions. A good argument for disallowing that kind of question is that this will encourage civic minded people to set up sites that work better for discussion questions. In fact, we have such offers on this thread. I doubt that we'll get such a site if MO encourages this kind of question as the site would be an MO "competitor".


    I sense that I am in the minority about the quality of the abc question: "Can someone briefly explain the philosophy behind his work and comment on why it might be expected to shed light on questions like the ABC conjecture?"

    Let me say first that Andy Putman's revision (as quoted above) is about as good a revision that one could make while keeping the essential sense of the original question (in particular, he got rid of the really bad parts), and it was very kind of him to perform that service. But -- to my mind, it remains just a pretty way of asking, "tell me about the work of X". In other words, it's still the kind of question that could be posed by just about anyone without any forethought or prior investigation or even any serious interest. How is it an "impressive" question, or even "admittedly good"?

    It strikes me as a case where the ends are used to justify the means. In other words, the question itself could be penetrating or it could be lazy -- apparently it doesn't matter too much, if it provides an opportunity for experts to write answers that satisfy people's cravings to read about hot new developments. I feel as though I must sound very cynical here! But I think if you were to ask the same type of question about a deep and technical paper but with a lot less buzz surrounding it, people would be far less impressed (and indeed, I seem to remember that questions whose essential spirit is "tell me about X" usually get criticized for being on the lazy side and are soon closed -- and I think there's a point to that).

    • CommentAuthorHJRW
    • CommentTimeOct 19th 2012

    +1 Todd. It's a lazy question.

    • CommentAuthorquid
    • CommentTimeOct 19th 2012

    Just some add-ons:

    1. I forgot to comment on "interesting-but-ultimately-useless." But, for me it was mainly just quoting Andrew Stacey with a different (from 'garbage') formulation he used, to get past the IMO tangential 'garbage' issue. Yet, what the formulation means to me is that there are not few cases where while something on MO might be interesting/amusing/fun to quite a few (sometimes including me), this something does not seem to have any actual relevance or use for anybody. Were I mean this to be understood in the context of thinking 'MO should be a tool for a research-mathematician' (if not 'use' starts to be highly illdefined.)

    2. Mainly, I am in fact rather with Todd Trimble. I did not mean to express 'the edited version of the question is a good MO question' (the edit was good given the context/constraints). What I however would admit is that hypothetically if somebody would have shown me the initial answers in advance, then I would have likely in the end agreed to somehow 'host' them on MO (bending the rules a bit on this occassion). Regarding the 'laziness' and related: Yes! This was infact essentially my very first criticism of the original question. Also, I am too not clear in which sense there was such an impressive success for MO.