I've found that I have to switch to decaff long before afternoon tea rolls around, but here's a thought or two that slipped between the cup and the cookie.

All the most pressing questions I have these days seem to get tagged as “soft-questions” here, and one of the hardest questions I know got deleted as “Not A Real Question” by the NARQ Squad. Well, maybe it's the time of year, or maybe it's the time of man, but there doesn't seem to be time for the reflective practitioner at all.

Now, I fully understand, in every population there's always a sub-population that would be kings of the hill, and the wouldbe kings of the MO-hill — sorry, Ladies, it's mostly a high-testosterone thing — have decided that “research question” means “question that can be settled before afternoon tea”.

It's that old Tragedy of the Commons again, and I don't see much chance for another cup o' tea, no matter how we partition the room.

Jon ;)

Old Boys Network or Omni Broadcasting Network?

Jon ;)

I love this idea. afternoontea.mathoverflow.net (ATMO) has a nice ring to it too.

By the power vested in me as chairman of the naming committee (if I'm the only member of the committee, I should be chairman, right?) of the Old Boys' Network, I officially rechristen our ancient organization as the NARQ Society, or the Not A Real Question Society. Our enforcement arm is hereby known as the NARQ squad. Individual officers of this division will be known as NARQs.

And there was much rejoicing.

Afternoon tea discussions (at least those that I enjoy) are dialogues. SO software is pretty much by design not great for dialogue.

@David: Setting it up here makes it part of the "MO brand", which means that it's reasonable to expect that the people who participate there will be a subset of the people who participate on MO itself. That way, it's an extension of this community, instead of its own new community (just like meta!).

This has been suggested before, I think more than once though I can't be sure and I can't be bothered to hunt back through the threads!

The main hurdle to get over for this is the question of moderators. Who wants to moderate the "tea party" or "math pub"?

Software's easy enough. I've tweaked the software the under-pins this forum so that it has decent mathematics support (by which I mean MathML for those that can and pictures for those that can't) and would be happy to help anyone who wanted to install it do so. The nForum is an example of what's possible. In fact, I can set up a forum for this in a matter of seconds alongside the nForum. The main reason that I haven't as yet is that I have no desire to moderate such a place. So if there's a group of people willing to moderate but don't want the hassle with the software, talk to me.

I would support such a site, but for slightly skewed reasons. Firstly, if it did use the same software then that would be good for the development of it. Secondly, for the same reason that I welcome math.SE: I don't go there myself, but it's useful to be able to point others there and get those questions off MO!

Difficultly and time-consumption scale with the number of moderators.

As would I. But I think we need actual (i.e. PhD or grad student) mathematicians for this not to look like a joke.

I like afternoon tea, but I'm not interested in hosting or moderating such a forum. I can imagine too many reasons it could fail:

1. It doesn't solve a problem. To formulate this concern as a question, how/why would a new discussion forum be different from math reddit?

2. It doesn't draw a dedicated audience. When Dave Brown and I first imagined MO, it was with the intention of making mathematicians better at doing mathematics. I like to think that people regard time on MO as research time. I really like collecting math factoids and I really like working with many different people to solve "little" problems. What's more, that's a large part of my job as a mathematician. It's harder to justify spending much time on a math chat forum. If mathematicians are the target audience, the subject matter should be mathematics. A math chat forum will draw people who identify themselves as "math chatters". I don't think there are a huge number of people who think of themselves that way, but I could be wrong. Professional mathematicians might not be the target audience, in which case you should clarify who the target audience is.

3. It could easily degenerate, or be mostly low quality.

   • It's really hard to pose a good discussion question, but it's really easy to pose a bad discussion question: "how can we improve mathematics education in the US?" Blogs solve this problem by (self)selecting people who start high quality discussions. You only follow blogs you like. Having to wade through an aggregate of all blogs would be really painful.
   • With precise questions, the threads are short, so there is little energy barrier to joining the conversation. With discussions, the threads can quickly become huge, at which point people either don't join because they don't want to read everything or they start making points that have already come up, annoying the people who have been following the thread and intimidating others by making the thread even longer.
   • With a bad but precise question, at least everybody knows when the discussion is over: the question is answered or essentially proved to be unanswerable. With discussion questions, even good ones, there's no good way to finish them. If you engage in the discussion, it's liable to become a drain on your mental/emotional energy with no end in sight. Blogs solve this problem by putting control in the hands of a small number of people. Once those people have moved on, the thread is dead.

4. I like in-person afternoon tea conversations. I think it's nice to have a small group of people to talk to. I feel more comfortable making/admitting mistakes and assuming a position for the sake of argument. It's easier to avoid people who are just looking to disagree, so it's actually possible to change people's minds (including my own) and to move forward. In person, you get much faster feedback, and it's of higher quality. For example, I can usually make out emotions in person, a task which is (unfortunately) often very difficult online.

All that said, I'm happy to encourage anybody who disagrees with me. If an idea is obviously awesome, it probably won't work, otherwise somebody would have done it already. Plenty of smart people expressed serious doubts about MO before it existed. If somebody wants to host and moderate an afternoon tea forum, I'd link it in the MO FAQ. I could even be talked into redirecting tea.mathoverflow.net (or whatever) if enough MOers support the project.

One could, at least in theory, restrict participation to people who actually know some math (measured, say, by having reputation >= X on MO, where X ideally is small). This would keep out cranks and spammers, and most of the people who participated would be mathematicians (or students, physicists, etc.). I don't really like this approach, but I can't see any other obvious way.

@Anton: I think the difference would be that "afternoon tea" is intended primarily as a discussion board for mathematicians. I haven't spent too much time on Math Reddit, but my impression is that most of the users there are not mathematicians (or students), and many of the threads there are just about sharing links. So the proposed forum would be a lot more like an actual afternoon tea.

If there wasn't any filtering, I don't know how it would be essentially different from math reddit.

I don't know whether the filtering is technically feasible though. One thing that could be done is to make people register with their email addresses (or openID), then check that they match with the ones of a registered MO user.
Or, if they have a respectable .edu email address, that should also be okay, to include mathematicians who don't use MO (I recently met a very well-regarded mathematician over dinner and asked him offhand if he had seen MO--he had never heard of it, apparently).

On the other hand, the ones that have never heard of MO will probably never hear of the afternoon tea forum (if it actually ends up existing). Presumably discussion-y questions (Fields Medal gossip, arguments over the merits of Bourbaki, etc.) could be moved over there (i.e., the OP could be asked to move them), which might get a few people to read the forums. But on the other hand, I also think that it's reasonably likely the serious mathematicians would, in view of their busy schedules, all stay on MO (or meta), resulting in the new forum being over-run by enthusiastic undergraduates. (Though a serious undergraduate math forum would not be a bad thing.) I say this especially because the only people who have expressed interest in moderation are undergraduates.

Test answer: you don't go with Wiley, CRC or WS if you can get any of the others, for reasons that are obvious. [Append gossip and bickering in test comments.]

Anyway, there was a proposal before to use javascript to check flair to ensure minimal rep for something or other. Perhaps if Anton could be convinced that someone else would do the work of implementing this along with setting up a forum there might be some room for traction?

Finally, the issue of privacy/attribution is probably nontrivial. If I was a traditional academic I might not want my test answer to be publically viewable, for instance.

Also, how about Asterisque (SMF), which also publishes english monographs?

Personally I wouldn't see the point (Evans tea suffices if I want to talk to mathematicians over tea, and Raleigh's and Jupiter's, neither of which are exactly a "pub", suffice if I want to talk to mathematicians over a beer.)

On the other hand, if someone puts their hand up and says "I'm in charge", I'd be happy to give them tea.mathoverflow.net or pub.mathoverflow.net (although it's Anton who actually matters here - he has the keys to the DNS box). Andrew has said he's willing to do the technical hosting side (as am I, for the matter, if all you want is a basic vanilla installation).

Requiring a mathoverflow account with some minimum rep sounds nice, but I don't think it would actually be a good idea. Too much hassle (for both hosts and users), and not that much clear benefit.

So I kind of like this idea, but there are two obvious hurdles:
- The idea needs to be carefully thought out ahead of time. If someone just throws up a poorly planned site, it will probably fail.
- Somebody needs to be willing to commit the (not insubstantial) time to set it up and run it.

I am also imagining it as more of a secondary site -- a place to chat about the kinds of things people on Moe might want to chat about, but which are just too discussiony for MO itself. To this end, I do like the idea of restrict to people with at least some small amount of reputation -- essentially bootstrapping the quality control on MO. I imagine there must be an elegant way to do this, from a technical PoV. I also would want to see some sort of moderation/voting type of system for sure.

As far as why you would want such a thing, rather than just going to Jupiters, a couple reasons spring to mind:
- Just a different and in some ways more diverse audience. (more vital if your local mathematical community that's not as big as Berkeley's.)
- The ability to focus on, or ignore, topics at one's leisure.
- A natural place to move interesting but too discussiony topics off of math overflow.

@Anton---my experience is that my most productive times coincide with the times when I'm either working on my own problems or reading research papers which are closely related to the problem I'm thinking about. MO is fun, but much too general. I don't look at it when I'm at work---it's very much a "when the kids are in bed" site for me. Just my opinion, of course!

Andrew---I think your summary represents the situation very well. I think the only difference between Anton's view and my own is that I attach almost no worth to the bottom 2 categories, insofar as "being worth doing during work hours" is concerned, whereas he attaches some worth to the middle one.

Kevin: what about getting a push in the right direction on research? I think that counts as serious and worthwhile.

@Andrew: a coauthor is one who still wants to work with you on the next project. A contra-author is one that doesn't want to have anything to do with you after this one.

I don't really have a clear distinction between "work hours" and "non-work hours." If I knew I was going to work on math for x hours a day, then certainly reading MO wouldn't be part of those x hours. However, for me that x isn't fixed and as Andy says, MO is a good way to use math to substitute for breaks that would otherwise be non-math. Something else that has a similar affect for me is skimming through papers that are in my field, but not directly related to any of my research projects.