Not signed in (Sign In)

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

    • CommentAuthoran_mo_user
    • CommentTimeMar 31st 2011
    This is a question prompted by the recent 'lament'-incident and similar things; I thought about the subject described below for a while already on other occassions but did not really think it through.
    (I never got to read the 'lament' post, as when it was visible, I had no time, and then it was already gone;
    I say this not as a complaint of any form, merely to point out that this question has nothing to do with the specific content of this 'question' that is, as said, unknown to me)

    My question is essentially in the subject line, I repeat it in expanded form:

    Should there (and if so, can there) be a place for 'discussiony subjects' linked to MathOverflow?

    First, I stress that I really mean 'linked to' or 'close to' and not 'on'. This is not at all about advocating to relax the policy of MO, in some sense the outcome could be the contrary. It is amply documented that a significant subset of the community does not want this type of content on MO-main and I think even somebody who actually would like to have it, would, looking at some of the somewhat discussiony existing questions, arrive at the conclusion that the site, on a technical level, is (by design) not well suited for discussions. So, certainly not 'on'.

    Second, the 'can there' part (if the answer to this is no, the rest is essentially obsolete).
    What could I (somewhat naively) imagine: There is already a forum/discussion board prominently linked to from MO-main. This one. So, could there be a second one? This is not clear to me, but also not impossible, as my understanding (possibly flawed) is that MO-meta is not part of the software that nobody here can change, and adding an additional link on MO-main might also be feasible. It is also not clear to me, not even roughly, how much work it is to set up such a forum (for somebody who knows how to do something like this). Another option would be to allow discussion on MO-meta, however I do not think this would be a good idea, as it raises the issue that somebody who is an intense user of MO-main might need/want to follow the MO-meta but only for actual MO-meta content and not general math-related discussions. (A strict tagging could partial remedy this, but then things like this tend not to work sufficiently well to be satisfactory.)

    Third, the 'should there' be part. Actually, I don't know, but perhaps yes, this is my main question. I can see pros and cons. The main points in favour of it: if this existed, it seems much easier to get rid of that type of content on MO-main. Second, and more importantly, I know of no other place than MO where there is such a large and quite diverse community of mathematicians gathered on one site. To duplicate this from scratch seems difficult (to say the least). Moreover, mathematicians seem to like to organize the 'softer' and more 'discussony' parts around more serious mathematical content (just think of all the birthday-conferences). So, basically what I envision could be the analog of the hallway at a math-conference: evidently general discussions in the lecture-room during a talk must not happen, but then a conference without any opportunity for general discussions might not be the best conference.

    Since this post is already long (thanks for reading, if you made it until here), I will omit to discuss my own objections to the idea, and leave this to others.
    MO does something well and in my opinion it's best to keep things simple and avoid trying to do everything for everyone. So I'm pretty much against any support of discussiony meta-math, at least, on MO.

    I agree with Ryan; this is edging towards unnecessarily complicated. In the past I have adopted an MO-unsuitable question on MO for discussion on my blog and others could do the same if they wished.


    For what it's worth, I think this is possible and "should" happen exactly to the extent that it happens on meta naturally. If there were some third forum dedicated to things that don't belong on MO or meta, in what sense would it be "linked to" MO? It seems like the connection would be as tenuous (or as strong) as if such a site existed completely independently. People on MO and meta already link to blogs and other sites when appropriate. I don't understand what functional difference the proposed forum would have.

    Note: if we notice that there is a strong connection between MO and another site, we can help it along. For example, when somebody links to an arXiv paper, we submit a trackback so that the arXiv links back. What makes this work well is that we're just facilitating a connection between independent sites that are good at what they do, rather than trying (somehow) to force a MOrXiv into existence.


    Believe me, you're not the first to come up with such an idea. I rate the possibility of such a thing coming to existence with an "official" link to MO as quite low. I can definitely say there's little appetite amongst the moderators, and I have yet to see the wave of popular support amongst users necessary to create such a site. I personally see little upside and many pitfalls in trying to create such an extension of MO's "brand." Of course, anyone is free to create such a site without MO branding. I don't think it would be all that difficult to get MO users suggesting that discussiony questions go there; certainly that started happening really quickly for math.SE. I would recommend that interested parties email you about trying to set up the website, but...well, anonymity has its pitfalls.

    By the way, you might consider just using reddit or some similar site. No point in reinventing the wheel.

    I understand the interest of an_mo_user. Though I agree with Qiaochu and Ryan (blogs are a good place and the MO admins have enough work), I think this does not mean that the discussion has to stop there. Some thoughts.

    1) I would go with Qiaochu and say: start a blog (it takes 5 minutes on and blogspot; look at for what's out there already). Then simply leave a comment at an MO question with a link to your post to invite people to a discussion.

    2) If you prefer the Q&A format, look for more open places than MO. Quora comes to mind. If you look for back-and-forth situation, why not use twitter (#mathoverflow should go a long way). Also, there are research oriented social networks such as researchgate or mendeley that could be a good place for discussions.

    3) If nothing fits but you have a concrete idea, then start a thread presenting it and looking for other MO users to join you (I'm pretty sure the MO overlords will look gently upon such an effort). If you can get a project off the ground, I'm sure the MO overlords would also be open to give "easy" support, say a prominent link location or similarly simple things.

    This idea has been aired many times here on meta. I'm not going to search back and try to find all the discussions, though. Each time it surfaces both Scott M and I offer to set up such a place but say that we wouldn't want to moderate it. At that point, the discussion tends to stop!

    Here's some points to ponder.

    1. The correct setting for a discussion is a forum. They are a bit more adaptable than blogs, and more easily threaded than Q&A. It is also easy to have different levels of membership so that only certain people can, say, start discussions or whatever. Discussions do not work in comment threads of blogs.
    2. The technology is already in place. The nForum (which I maintain) uses an adapted version of the same software as this place and one of those adaptations is that I can add a whole new forum with about three clicks. Separate forums (fora?) are almost completely separate, sharing just login information (the forum for John Baez's new venture, the Azimuth Project, is hosted there). In fact, although it doesn't get used much, the nForum can emulate a blog when it wants to.
    3. The main problem is moderation. You would need to assemble a team of moderators to keep an eye on things. Discussions are liable to erupt in flames at a moment's notice and it is important to have people and policies in place to deal with them when they do. On the nForum, we have a "Cave" where I can put discussions that get out of hand. It doesn't show up on the front page, and no-one gets notified of replies there, so any discussion sent to the Cave tends to simply die of its own accord. It doesn't get used much, partly because the nForum has a distinct purpose that is external to itself: discussion of the nLab. A "mathematical discussion" forum wouldn't have that so would need much more time and energy to keep it clean.
    @Andrew, great to know the nForum would be open for such a thing! (although an_mo_user seems not to follow the discussion *sigh*)

    I would nevertheless defend that a blog is more efficient for discussions coming from MO. In my experience, a forum works best for long-term discussions, mid-sized groups and a fixed common interest. I'd expect the opposite for discussions "beyond MO", much more like a twitter discussion.

    A blog is in my eyes perfect to "invite" a small group from such a large group as MO for a short discussion. Somebody personally takes charge, they automatically volunteer to moderate that discussion because (for better or worse) they get the visibility personally. Also (very important), they can close the discussion at any point. It's simply more personal.
    • CommentAuthoran_mo_user
    • CommentTimeApr 1st 2011 edited
    Thank you to all that responded so far.

    Peter Krautzberger, I am following the discussion; I merely did not want to appear as trying to dominate it and thus stayed silent for a while.

    I will respond to more points later (some hours), but not yet (for this reason). Just one remark:

    Regarding the suggestions for blogs. On the same grounds, one could question the relevance of MO itself: everybody having a mathematical question could simply post it on her/his blog.
    The problem with this is of course to reach a critical mass of audience/participants. And, I think this is at least as important if not more important for topics that are a matter of opinion than for questions with a precise answer (where there is either no answer or one person knowning it is suffucient, for opinions one needs some relevant sample size, else the outcome of the discussion seems almost worthless). What makes this even more difficult is that it is not merely a question of having some audience, but it seems key to me to have a diverse one. (I really do not need to read the opinions of my colleagues and friends on my blog, asuming I had one; they typically tell me in person or email.)
    As a frequent reader of various maths-blogs, I would say that the number of different people that tend to comment on one given blog is almost everywhere quite small. While I like to read various blogs, also perhaps even particularly smaller ones (so the following is not meant as negative as it may sound) many blogs have more an 'open diary' feel, where once in a while somebody else leaves a note.
    Finally, I agree with Andrew Stacey's first point, from a technical point of view a blog does not seem optimal.

    Yet, I have to agree that the idea of 'inviting from MO to a given blog for a given questions' seems like a good suggestion. Perhaps this would work. However, how to communicate this invitation in a non-intrusive way? To ask a question with the content of the blog post, with a postscript: don't hesitate to close this, was just ment as an advertisment for the discussion on this blog [link], might be efficient, but is evidently an absolute non-option in particular if this should be something more or less common. (Though it seems at present this more or less works like this, just with the questioner and the blogger being different persons.)

    Maybe a 'Planet MathOverflow' ?
    Wouldn't that simply be a duplication of another website, "Planet Math"? Keep in mind the administrators are grad students and postdocs and are operating on little in the way of funding. So if you want to create something you largely have to do it yourself. Moreover, it's not clear how this thread is helping in that regard.
    @an_mo_user: looking forward to your post! (sorry for suggesting you had abandoned this :) ) I'll wait until you post before I reply.
    • CommentAuthorMariano
    • CommentTimeApr 1st 2011

    A Planet Math would not be utterly pointless.

    @Ryan Budney, first actually I was unaware of 'PlanetMath', but a cursory glance suggests that this is something very different than what I meant with 'Planet MathOverflow' sorry for not making this more precise. What I meant is something like this 'Planet Debian' (and 'Planets' exist for various projects). This is a mere aggregation of blog posts from certain appropriately tagged posts on blogs. In the case I linked to the blogs of Debian developpers that they tag as relevant to Debian.
    As various prominant MO-users have blogs, it might be interesting.
    Regarding the second point you raise 'time/work': yes, I am very aware of this and I am really grateful for all the time they invest. I even thought about adding a pre- or posttext to my original post regarding this issue, but then decided against it (perhaps this was the wrong decision), as I can well understand that somebody (and then even anonymously) making suggestions/requests could be perceived as somewhat odd. (I will say something more to this extent, in my longer answer, after which I will also stop except others keep up the discussion).

    @Peter Krautzberger: No problem. Thanks for your interest.
    • CommentAuthorRyan Budney
    • CommentTimeApr 1st 2011 edited
    We appear to be getting off topic? This is a forum for discussing things related to MO but it seems this thread is about the creation of a general social-media framework for all of mathematics. As Ben mentioned, websites like Reddit can serve that kind of role. Does mathematics have any special needs in this regard?

    For example:
    @Ryan I don't quite follow your objection :( -- the question was "off topic" from the start, wasn't it? What's changed from your point of view?

    Most mathematicians are extremely unaware of the social media tools available. As a scientific community, we're years behind the scientists in this respect (though ahead in other things like arxiv and MO). *But* a lot of mathematicians hear of MO (for all the right reasons).

    I would argue this gives us the a responsibility to discuss such things somewhere on MO. I wouldn't know what other place than meta.MO (well, if there was an admin/moderator-blog to add to the static FAQ...).

    If the moderators nevertheless feel like closing the discussion, let me again point out *Quora* ( for Q&A like discussions. I'm not a user (yet) but it's had a big impact in other fields already.

    Maybe the MO-FAQ could point to more such sites.
    @Peter: maybe I misunderstood the original question. The original question was talking about a "corner" of MO, which I interpreted as something that is part of or closely affiliated with MO. It appears now we're talking more of a general social media framework for mathematics. The former seems relatively on-topic (but people aren't so interested in it), the latter is more off-topic but there seems to be vague support for.

    I actually quite like the idea of a MathOverflow aggregation blog. Set something up to collates blog posts tagged "mathoverflow", throw a nice theme on it, disable comments, and then wash your hands of it. Anytime anyone wants to "adopt" a question that's a bit too discussion-y for MathOverflow, they just create a post on their own blog, tag it as mathoverflow, and post a link on the closed question. The central blog serves only one purpose --- to direct the attention of whichever people want more discussion-y\ content in and around MathOverflow to all the individual blogs that they wouldn't otherwise follow.


    • CommentAuthoran_mo_user
    • CommentTimeApr 1st 2011 edited
    Having now read what Scott Morisson wrote, I remove this last contribution.
    He describes exactly what I mean, just much better.
    Didn't someone recently set up a nice math blog aggregation page? I remember seeing it, but can't seem to find it at the moment.

    Sorry for my late reply (research had happen at some point...).

    @Ryan: I understand. I'll keep it short and final this time ;)

    @an_mo_user, @Noah, @Scott Morrison: I'm one of the people behind I'm pretty sure my co-conspirators are happy to invest some time to create the kind of aggregation that @an_mo_user and @Scott Morrison described.

    Tag aggregation is on our to-do list for the main site anyway and it wouldn't be difficult to implement a "planet debian" like subsite after that. We're absolutely open for help but all our code (including our database) is openly available so feel free to build on our buggy stuff ;).

    @an_mo_user and anybody else interested: just contact me directly (my name's easy to google... email, twitter, website).

    On a final note, I would love to try to convince the moderators & power users to start a group blog (something MO specific, I'm aware some are involved in The Secret Blogging Seminar), but perhaps that is for another time. I'll definitely bug Francois when I'm back in Ann Arbor next week :)

    • CommentAuthoran_mo_user
    • CommentTimeApr 1st 2011 edited
    Some hours ago I said I'd write a longer response, then I said some small things in between, and then the subject changed a bit.
    [Having finished the post I see that it is again a bit long. In case the discussion in one way or the other continues, I will certainly read it, but will only contribute if somebody addresses me directly. I see and understand that being annonymous in such discussions is not optimal; and thus will reduce my activity on meta, as along as I remain annonymous.]

    First, in retrospect I think I should have started differently. Perhaps the question itself was ambigous due to a poor choice of words. The 'should there be' was meant in the sense 'assume it were technically feasible and somebody would be willing to do the work, would it then be desirable to have this' , certainly not as 'there really should be something like this. please add it.'

    Second, why did I even ask this. As I was interested in the general opinion on this, and thought it might be a nice idea but actually was not sure myself whether it was. And, as I saw this idea sometimes come up here and there, but only somewhat implicitly or in a context that made it difficult to follow up on the idea. (Namely, in discussions on the appropriateness of specific questions, somebody, I believe mainly Andrew Stacey, said something like 'please discuss these subjects somewhere else, I would even help you building the site' where then nobody followed up.) However, I never saw a public discussion dedicated to it. Possibly, it happened before I was around and while I read parts of the meta-archive I certainly read not all of it. Finally, for example math.SE seems to have a built-in chat. Now, I do not know whether anybody uses this, and also I do not think that for the discussion-y content a chat is the best medium, but somebody apparently thought to have a chat attached to such a site would be a good idea. So, perhaps not a chat, but something vaguely similar, a place for slightly more chatty subjects.

    Third, what has this to do with MO and why not simply do this completely independently. As Ben Webster put it, it would mainly be about using the 'brand'. However, using a 'brand' for something else is certainly a risk. And, to moderate something like this is definitely not easy. So, most likely the idea of doing this is in the end really not a good one.

    Fourth, the blogs. Frequently people gave the advice to other people to use blogs for the questions they asked but that got closed. I always found this a not too useful suggestion, as while creating a blog is (I assumed and Peter Krautzberger said so) not a big deal, to get the readers to the blog seems non-trivial. For example, I read a question I find interesting but it gets closed. How will I know somebody decided to take this to a blog? If somebody decides quickly, fine one could leave a comment on the question. But, what if somebody decides after three days; pumping the question in some way for this type of announcement seems inappropriate, and the comment who will read this on a couple of days old closed question? Anyway, because of this I always thought this 'go-to-blogs-idea' is not feasible. However, during this discussion I then thought as people mentioned blogs so frequently and then remembering the various 'Planet' sites, perhaps this could be interesting. Unfortunately, I somehow at least partly failed to convey what I meant. But, fortunately Scott Morrison came to the rescue.

    Fifth, when starting this discussion I thought blogs are (from my point of view) irrelevant to the discussion. However, now not being convinced of this anymore, the question somehow arises what is an MO-related post; of course one that talks about MO itself and an 'adopted' question. Yet the questions that need to be 'adopted' actually should eventually 'die out' as a consequence of this entire excersise. So, then only the post about MO itself remained, but these might no be too many. Thus there might be a need for more content, such as the posts on blogs of users of MO that are of interest to a broad mathematical audience (and thus tagged as such); opposed to more individual blog content, be it of personal or mathematical nature. But then there is no very good reason to limit to the blogs of MO users, and then the clear link to MO really dissapears. And so this stops to be a relvant question for this place. And, might rather be a topic for the site linked to by Noah Snyder.

    Whether to still have an aggregation of MO related blogs would be a nice thing, say to present the community of users in an additional form, I will leave this discussion to others (being annonymous and at the moment without blog I am really not the right person to discuss this).

    ADDED: Again I missed a post before posting... @Peter Krautzberger, thank you. I will contact you. (Just not now, as I am in Europe and it is thus already a bit late.)
    • CommentAuthorfedja
    • CommentTimeApr 2nd 2011
    I'm one of the few people here who would support such an idea but I completely agree that efficient moderation is an issue. The relation to MO is not the "brand name" but, rather, having the same audience. So my suggestion would be to restrict posting in such forum or whatever it may be to the people with, say, 2000+ reputation on MO. I do not see anything wrong with general discussions related to the craft of mathematics (moreover, I like them a lot) but I really do not want the "random cranks" and other low quality posters to jump in in herds (that's one of the reasons I do not run a blog, by the way).

    @fedja: on WordPress (and, I'm sure, other sites) you have the option of total control over comment moderation, so you have complete discretion regarding who is allowed to comment or not. I had to start doing this after a troll invaded my blog...


    Just on those last points, forums are far more configurable than blogs on that. One can have a variety of levels of membership, and it's much easier to promote or demote people. This forum here is not very ... er ... sophisticated and doesn't show all that is possible.

    I should say that I would also welcome a place to discuss such things. What I would want was to be sure that the people that I was discussing with were also professional mathematicians (or on the path to be such) since those are the ones whose experience I could benefit from. For example, I would really like to discuss teaching ideas with other mathematicians.

    That's why I'd be happy to set up such a place; it's just that I don't think that I have the time to devote to moderating it, since I'm already technical support for the nLab, nForum, Azimuth Project Wiki, and Azimuth Forum. But since I already run two forums (actually three and a half, but let's not quibble), adding another wouldn't add any time.


    Just a small and final (on this thread) comment from my side of things.

    The crew working on has started to build the basics for "Planet MO" or "MathOverflowOverflow" as @Scott Morrison called it (I really like that). We hope to have something running next week. I'd love to discuss it at that point in a new thread.

    • CommentAuthorMariano
    • CommentTimeApr 7th 2011

    MO² ?


    When you pass 100 reputation at, you gain the privilege to "create chat rooms"...