Vanilla 1.1.9 is a product of Lussumo. More Information: Documentation, Community Support.
As one of the participants in (and, unfortunately, contributors to) a small bit of the meta.MSE drama, I would like to voice my tepid support for the proposed migration to SE 2.0. My "support" is for the following reasons: the comment notification feature on SE 2.0 is a real improvement, and tools like migration, comment editing, etc. make life a bit nicer over there. (I am less enthusiastic about the other features: for instance, I have never had real use for a tag wiki. If I saw a MO tag about some entirely unfamiliar field, I would simply search wikipedia, not rely on a short blurb.) In addition, perhaps it is best not to remain on the same platform for years and years as technology advances. I understand many people have been frustrated by our inability to change the software.
My initial worry when the migration was proposed was the following: as has been observed above, the moderators at SE (in particular, Jeff Atwood) have been known to over-rule community moderators, sometimes contravening website norms to enforce network norms (which are much less uniformly shared). On MSE, some of Mr. Atwood's decisions seemed unfortunate, and I feared that if MO migrated, then similar history as on MSE (i.e. the threat to suspend Robin Chapman, the deletion of Pete Clark's comments, etc.) might repeat itself, which could quickly wear through the patience of the very busy and smart people that answer questions here.
One explanation, I think, is that SE is primarily based on StackOverflow, which is like a "big city" with its problems: moderation needs (or so it is said) to be extremely aggressive, complete impersonality must be observed in interactions (to the point where salutations "Dear X" are automatically removed, thanking answerers is discouraged), and a lot of emphasis is placed on suspension. But MathOverflow isn't a "big city" in this sense: it's a professional community where research mathematicians, usually using their real names, ask and answer questions where their peers can watch. There's no need for the same harsh levels of discipline and punitive emphasis. I don't obviously know the extent to which MO is plagued by spam, but it seems that real drama (flags, abusive comments, etc.) are pretty rare nowadays for obvious reasons. If MathOverflow is much more like the quiet atmosphere of a university math department, does it really make sense to have armed members of the NYPD hovering around to prevent fights?
I actually suspect that part of the conflict occurred because of this difference in culture: math.SE, like MathOverflow, is a fairly quiet place (I can attest from a brief stint that moderation was usually light work, involving minor spam deletions), while StackOverflow apparently generates hundreds of flags a day. Actions of the above form taken by SE moderators may become more understandable (though, in my view, still undesirable) on a website where moderators are overwhelmed with problems.
Motivated by these concerns, I initially emailed Anton after this thread was posted with a list of reasons why I did not think the migration was a good idea. I talked to him earlier today, and I am less worried now. Anton and the other moderators do not seem to think unilaterial interventions by the SE moderators are likely on MO if a migration should take place; apparently Mr. Atwood and Mr. Spolsky have said as much. In addition, I found the testimonials of the cstheory.SE moderators reassuring. I would still hope, that if such drama at meta.MSE were to repeat itself, the website would strongly consider using the "escape" clause. (I am still cautious because I am not yet certain that said event has a comfortingly low probability. As I recall, Mr. Atwood's (in my view, ill-advised) interventions on meta.MSE began after most of the early problems had been worked out. I would be more at ease if non-interventionism on the SE admins's part was included in the terms of the migration. In my view, the optimal solution would be to use an open-source platform; however, none adequate seems to exist right now.)
Finally, I hope many more will comment here!
The issue of possibly having SE employees telling me or my colleagues what to do on MO, as happened in a number of affairs at meta.math.SE (Robin Chapman
Maybe I don't know the full story here, but here is the story as I understand it: Robin Chapman was hitting Enter to start new lines at the points in comments that he was used to. At some point, it was decided that hitting Enter would automatically post the comment. Instead of noticing this and composing his comments without hitting Enter, Robin continued to hit Enter at the appropriate places in his comments, causing each line of his comments to be posted separately. Given that only a fixed number of comments start out visible, this was crowding out other comments on anything he commented on, so Jeff Atwood decided to merge Robin's comments to fix this. Robin protested against this, Jeff stood his ground, and Robin left.
Can someone explain to me why everyone else seems to think Robin Chapman is the reasonable person in this situation? Jeff may have overreacted near the end, but I think it was mostly out of complete incredulity at Robin's stubbornness, which I have to admit I shared at the time.
@Qiaochu: As I said at the time, I don't think either of them was behaving very reasonably. All in all it was a pretty bizarre situation. If we are attempting to make some kind of tally / reckoning / whatever of mathematician / SE-team relations on math.SE, I wholeheartedly suggest that we leave this incident out.
If we are being comprehensive though, I would not leave out the part where Jeff Atwood decided, after the end of the moderator election, that Robin Chapman had somehow withdrawn from the election and that therefore Akhil Mathew was the third place winner. Given that Robin Chapman still has the moderator diamond more than six months later, this seems to be an obviously factually inaccurate description of what took place. (It is also contradicted by the official page tallying the results of the election and therefore serves to undermine that page.) If memory serves, the elections were tinkered with in at least one other instance: one nomination phase was lengthened by the SE team in order to encourage more nominations. Although the specific outcomes of the tinkering are not objectionable to me (in particular I myself voted for Akhil and not Robin as a moderator -- I might even go so far as to say that electing a moderator who claims to be unable to stop hitting return was a mistake on the community's part!), the tinkering itself was on principle a big problem for me and a major ingredient in the erosion of my trust of the SE team.
It seems that the place that we are now is that a lot of people are in favor of the migration provided it comes with a list of promises from the SE team, many of which are of the form "We will not behave towards MO.SE in the future the way we have behaved towards math.SE up until now." I find it interesting that people whom I deeply admire and thoroughly trust are convinced by this. Honestly, it makes me look at myself and wonder what I am missing. Please feel free to contact me, either to provide further information or (for bonus points, to be sure) to try to explain why I am feeling so much less optimistic about this than others.
@Qiaochu: I don't consider Professor Chapman's actions ideal (he could have tried a bit harder, to be sure), but, as explained by many respected commenters on the thread, the SE developers' choice that enter should submit a comment was unfortunate. His answers were extremely helpful. I was rather surprised that Mr. Atwood (who, as we all know, has much bigger responsibilities) would threaten to suspend him, especially since he was not a member of the website. As a result of this over-insistence on rules, MSE lost a very valuable contributor. MO surely does not want to repeat that: it is fundamentally those who answer, more than those who ask questions, that make the website great.
As I understand, Dr. Chapman left not because his comments were being merged, but because of the emails he received.
In fact, I hope Anton or somebody could implement something (in Javascript? I know sadly nothing about these issues) to make this the default, assuming other people agree with Joel David Hamkins, T.., and others on MSE, if this migration takes place.
@Pete: I agree that my appointment was unfortunate, and I was uncomfortable with it myself. I accepted it in hopes of not fueling unnecessary drama (which is ironic in retrospect). Perhaps it would be good to ensure that nothing like that will happen on MO.
Also, I suspect from the comments of Suresh Venkat and Kaveh that the actions of the SE team towards MSE were exceptional. I think this is what the MO moderators believe as well (or they would not be considering this at all). I think that the early turmoil (caused by other reasons) on MSE may have motivated Mr. Atwood's repeated interventions: on a site that already runs very smoothly, and has experienced moderators, why would he intervene?
Here is the list of modification in SE 2.0 that you may want to check, also the general SE 2.0 FAQ
Another point is that some of the in house additions to SE 1.0 software on MO are not available on SE 2.0, e.g. hiding reputations/votes/ignored questions.
Just reminded that there is the possibility for users to modify some elements of the site's user interface without altering the SE 2.0 code using apps/user scripts. As an example, here is a user script that would solve the enter key problem in comments mentioned earlier.
ps: I personally feel that right now there is too much emphasis on what has happened on MSE in the current discussion. Although both MO and MSE are about math, the communities are very different. I would expect MO would be treated more like cstheory than MSE, and the need for approval by MO moderators of moderator interventions by SE employees on MO can be put into a formal agreement. I feel that part of the problem on MSE was that Jeff feels and cares strongly (and too strongly in the eyes of some MSE users) about MSE and sees himself part of it, whereas this is not the case on cstheory or stats.SE, I would be expecting a similar attitude towards MO if it migrates to SE network. Here is the list of his posts on stats meta and cstheory meta.
Various thoughts while reading the thread:
Pete Clark: I would feel much more comfortable if we were actually paying something for the services we want SE to provide.
Me too. SE 1.0 was perfect for us. However, that option just isn't on the table any more. They won't take our money.
Pete Clark: I honestly think that conversion of MO to SE2.0 would be a very desirable outcome for the SE team, and we should make sure to get some value for it.
Tyler Lawson: what makes us such special snowflakes that we deserve this treatment?
Pete Clark: This self-rule really has not happened on math.SE
@Pete: what is so desirable about that outcome for SE? I agree that it's desirable for them, but I can't put my finger on why. My feeling is that MO is a special snowflake in some sense, perhaps simply because of the engaged self-ruling atmosphere that comes from its legal autonomy. Another thing MO has going for it is that is probably the most professional SE site. The problem is that I can't think of any real benefit the SE network gets from MO migrating that it cannot have anyway. If I could come up with such a benefit, it would greatly ease my concern about not paying them. It would make it clear what the motivations are on all sides.
Tyler Lawson: Are there any real options other than- staying with the legacy code of SE 1.0, with no upgrades in functionality,- migrating to become a "standard" SE 2.0 site, with the corresponding issues that have been discussed at length above, or- switching software?
Not that I know of.
HJRW: Who's to say that some regular MO users might not prove to be be 'hotheads', who might make a confrontation with the SE authorities worse?
When I talked on the phone with Akhil, one interesting thing that came up is the fact that MO used to have a lot of drama. The other moderators and I have exchanged lots of emails with MO hotheads. It has happened that an awesome user deleted (and later undeleted) all of his posts. It has happened that people left (or threatened to leave) MO because of politics. Conflicts like this are unavoidable. It is kind of interesting that nothing like that has happened in a while on MO, suggesting that perhaps an early "hot phase" is a normal way to aneal a healthy community.
When I talked to Jeff, we talked a bit about how moderators contact users, and about moderation in general. Here are a couple of points I got in that I think Jeff took seriously.
Pete Clark: explain why I am feeling so much less optimistic about this than others.
My feeling is that nobody has extremely strong feelings in either direction. Everybody either has some worries that need to be laid to rest, or is cautiously in favor. Here are two reasons you may be less optimistic than me. (1) You have had negative direct interactions with Jeff. In contrast, my interactions with the whole SE team (including Jeff) have been mostly positive. (2) One of the major concerns is whether SE will mess up MO by trying to moderate it. It's very easy to convince them (they're already convinced) that the other moderators and I can do that job, and we can do it far better than they can.
Of course, I still have some worries about migration. The main one is whether being part of the SE network will cause an image problem for MO. Mathematicians are keen to avoid their professional communications being exploited by some commercial entity. I don't want this concern to exist for MO.
@Anton: thanks as always for a very thoughtful and kind response.
As to the question of why 2.0ification of MO would be desirable for the SE team...maybe I should admit that I do not have the most insight (and certainly not the most charity) into their motives and goals. That part of my response was speculative. But I do think that MO was a big territorial expansion for them: into a very old and well-established academic community.
[Note that people have used the term "professional" to describe both the users of MO and those of SO, and I think both usages are correct, but these are two very different professional communities. I continue ruminating on this to try to understand the issues better. For instance, I was actually really pissed when I discovered that routines had been written to search through questions, answers, and comments to strip away things like "Dear X": I viewed that as a form of censorship. I still do, but this sort of thing bothers me because in my job I spend a lot of time figuring out exactly how I want to write something for it to be clear, correct and concise [ok, the last is not my forte...] If someone starts making changes to a math paper automatically -- i.e., without having a human read it -- disaster will surely ensue. Let me tell another quick anecdote about taking writing seriously: I write about two MathReviews a year, and I spend a lot of time on them. Last week I looked at one of my recently posted reviews and found I was referring to the writer of a single authored paper as "the authors", and I wondered how on earth I had not caught that. I looked at the document I submitted and found that it was not in my original: it had been an unannounced editorial change! Was this a big deal? Actually, yes: I immediately sent an email to MathReviews asking what happened and I spent ten minutes describing the situation to a colleague of mine that I hadn't seen in a while. The next day I got a reply from MathReviews, which said: they didn't really know what happened, there was no excuse for what happened, they were very sorry, and the correction should be visible as soon as the next day. In other words, I was dealing with someone in the same professional culture as myself: they took the whole thing just as seriously as I did. But if I were a professional coder, then what I take seriously is probably my code, less so my writing. Moreover, I'm much more likely to think it's cool to write little pieces of code which make my life more efficient.] But the expansion was incomplete, and they get asked a lot of questions like "How come MO isn't part of the SE system?" To me, trying to view it from their perspective, it looks like a half-converted, half-squandered opportunity. But again, I don't think you should take this opinion too seriously.
When I asked for explanations as to why I was less optimistic than others, what I really meant was that I am willing to listen to persuasion that my position is too extreme. If anything, your answer displayed too much empathy for my position: it's not a thought experiment for me to put myself in my own shoes! :)
About the image problem for MO: yes, I think this is a serious concern. Mathematicians are a pretty idealistic group. I saw on meta.SE just within the last day or so that the SE team has some kind of affiliation with amazon.com and because of this they subtly change links to amazon.com in order to make it clear to amazon which of their orders are being triggered by SE sites, for which SE gets some money. I was again pretty surprised by this, and I decided that when mentioning math books on the site I would not link to amazon anymore (I even went back and changed a few links). Now in the world of corporate stuff this is not much of a shocker: SE is a for-profit entity so they are doing something almost unnoticeable by me that makes them a profit. It's not exactly scandalous. But nevertheless I'm not comfortable participating in that.
Also the SE team just changes things around too often for my staid mathematical tastes. Both their business model and their product are very different today than from less than two years ago -- some changes have been for the better of course, but that's a lot of change. Other great internet mathematics artifacts -- MathSciNet, the arxiv -- seem much more stable. (Even wikipedia seems more stable, ironically, given everyone's early worries that they gave carte blanche for anyone to screw it up at any time.) MO is something that we want 20 years from now almost as much as we want it now. I think we should be looking for a solution which offers this kind of long-term stability.
I've not had a chance to weigh in since I'm staying with relatives and only have internet access occasionally. I'm also slightly conflicted about what to say.
Before Anton had spoken with the StackOverflow people, we had had a lot of discussion of what would be involved in Tyler's Door #3, switching software. And I'll be honest, I found it pretty daunting. I mean, we have a software product we're happy with, and we discussing expending a lot of scarce time and energy on rebuilding it from the ground-up, having endless arguments about which features to tweak, and almost certainly ending up with an inferior product. It would be one thing to make such an effort to get a new platform that we can't get any other way, but when it already exists, and can be had on non-insane, if slightly sub-optimal terms.
From the context of those discussions, the terms we were offered for upgrading to SE2.0 sounded fairly reasonable. My paramount concern has always been the escape hatch, and it sounded like they were willing to accomodate us on that point to a reasonable degree, and my feeling is still that provided we make sure that option is ironclad, I'm basically comfortable with upgrading.
I won't dismiss the concerns people have; most of them are serious, and they give me more pause about the upgrade than I had before reading this discussion. The danger of people wandering on from other parts of the SE network and adding inappropriate content is serious, but I think can be dealt with; in fact, I think moderating will be a lot smoother when we can migrate questions to math.SE. The danger of inappropriate intervention from the top at SE strikes me a bit odd as we are no less vulnerable to such intervention now. I'm at least assuming that, say, Jeff Atwood could assume moderator powers on MO and make trouble any time he wanted, given that the site is on SE servers. What basis do we have to suspect he'll suddenly get the urge to do so if we migrate? Of course, I can't rule the possibility out, but that's why we have the escape hatch.
Of course, I can't say firmly what the psychological effect on users of tighter ties to the SE network will be. Maybe it would alienate a bunch of good users. I haven't seen anyone on this thread say they would quit the site if we migrated, but I haven't been paying a lot of attention and may have missed it. I'm not disagreeing that it would be better to be on a site we have tighter control over, but we need to the software in order to do that. I would love to see a motivated group of MO users build that software, but I'm not holding my breath to see it happen.
I have joined meta for the express purpose of commenting on this thread. I have a small amount of experience in MSE (only in asking a few questions outside my field of expertise that didn't belong here), but a decent amount of experience at MO. I was going to ask the same question as DL: if we migrate, and then need to push the button to escape, where do we escape to?
And also, a much harder question, what if we are like the frog in the pot, with the water temperature slowly rising? By this I do not mean SE super-moderator interference, or an influx of spam or the like, but a slow bleeding of the community of the experts here due to a slow decrease in the overall experience? We won't be able to draw the line and say, 'ok now we get out, we have lost too many people'. Perhaps I am being pessimistic, but would MO qua another SE2.0 site have attracted/attract top researchers/Fields medalists etc?
One thing I would encourage people to do is read the complete features list at
http://meta.stackoverflow.com/questions/59445/recent-feature-changes-to-stack-exchange
There are a number of lovely features that have been mentioned already, replies to comments being perhaps the one I find most attractive, but there are a large number of features listed that I feel are less palatable. One of the most irritating to me is this:
2011-02-21: You can now embed ads for Area 51 proposals to help support and promote proposals and beta sites.
This may be a rare event, but if it allowed by the platform, we have to be prepared for the possibility (is that too dramatic? Perhaps).
One other thing that irks me, perhaps irrationally, is the live (and retroactive) tweaking of reputation rules; how much rep one gets for things, how much is needed etc. are at the whim of SE.
SE's motivation for wanting MO back in the fold is an interesting question. I don't think it's because they're paying us back for being such a good poster child for the VCs.
I am not opposed to the move per se, but I would like to be fairly certain the community isn't broken.
OK whinge over.
would moving to SE2 render the work of this core group of people more onerous, or less?
Unfortunately, the answer to that is going to be mostly guess work. Not only do we not know what the future may hold if we were to migrate, we don't even have nice points of comparison by proxy: the intersection of MO moderators and MSE moderators is the null set.
Though my guess would be that the difference would be likely negligible for any of the moderators other than Anton and Scott, by virtue of the fact that there's actually very little work involved in being a moderator for MSE (speaking from first hand experience). So at the very least I don't think the work load can get significantly larger.
Are there currently other examples of two separate SE sites as related in subject matter as MO and math.stackexchange?
There are currently Gaming, Role-playing Games, and Board and Card Game sites. There is an English site and a Writers site. The main example I can remember of two sites with related subject matter being merged is that Guitars was merged into Music, but that seems to have been largely because Guitars simply did not have enough people to sustain itself. Neither MO nor math.SE have that problem so I don't anticipate this being an issue.
...seems to have been largely because Guitars simply did not have enough people to sustain itself...
A bit more than that. Neither Guitars nor Music was looking too healthy. Music was slightly healthier.
Also, most of the "duplicates" managed to be "killed" at the proposal stage; and a few gets flushed when it is in beta, when it becomes clear it is not viable. For example, there is (AFAIK) still ongoing debate whether the Cryptography.SE (which will be open for public beta soon) deserve to exist as a stand-alone site. Despite what I said earlier, I am actually not all that worried that they would try to merge Math.SE and MO; each has a healthy (in terms of size and traffic) community, and despite the partial overlap in audience, serves two different purposes.
Some more examples of related sites:
But one pair that I think most resembles the relationship between Math.SE and MO is actually
One plausible scenario would be a semi-merge of MO and math.SE. For example, what if they essentially turned MO into a front end for math.SE? All math questions would have a research flag that would correspond to whether they were aimed at MO or math.SE. People on math.SE could choose to filter out either type of question (or see both), and visitors to mathoverflow.net would by default see only the research questions. I'd be opposed to this, since it would tie MO even more closely to the stackexchange network, but I can imagine the SE team might consider it completely reasonable.
I don't see this happening. The SE team is all about community-building, and it knows that forcing two (large, successful) communities together that don't want to be together isn't a good idea. I really, really don't think this will be an issue and I am pretty sure either I or Anton could get a verbal guarantee of this.
(Not that I am unequivocally in favor of migration. Many people have made a lot of good points, and right now my opinion stands at "if it ain't broke, don't fix it." But I just want to make sure we turn it down for sensible reasons.)
Earlier in this discussion, there had been some questions about how SE would benefit from a migration, so I'll offer some baseless speculation.
It is possible that they anticipate benefits from a market-dominant image, i.e., if the SE 2.0 community were seen as "the only game in town", so to speak. If they can convince people that any alternative to the SE 2.0 platform and community, even an earlier version, is unacceptable to the users of a Q&A site, then migration and inclusion to their community would carry significant force, at least among people who pay attention to such things.
Also, I think they may feel it is advantageous to offer a migration now, while we don't have many escape options. Certainly, MO moving to an alternative platform would not cast a positive light on any claims that SE 2.0 is the best or only reasonable option available.
Slightly tangential to the discussion, but since the topic about whether Math.SE and MO can meaningfully coexist on the SE2 platform, a similar discussion is now happening on Physics.SE. Physics.SE can roughly be compared to Math.SE in terms of level, and there is a proposal to make a Theoretical Physics site whose relation to physics.SE would be just like the relation between MO and Math.SE. The proposal seems to be getting close to being axed.
Of course, there are certain differences:
However, the thread does give an indication of how SE.inc looks upon two communities with possibly large overlapping interests.
1) OSQA is bad news. They now have a "open source" product, and a "commercial" product. It's very unclear the relationship between the two, or why they have any incentive to treat the users of the open source product well. I can go into much more detail, but I think Anton and I have agreed that for now, OSQA is off the table.
I thought I understood this, but I don't, really. When you say OSQA is off the table, do you mean the open source version or the commercial version? If the former, is the latter an option?
The commercial product offered by DZone is Qato. One concern with DZone is that they may eventually drop support of OSQA leaving us in a situation much like the one we are in right now.
I'm the CEO and co-founder of Stack Exchange, the company, and I wanted to give you some of my thoughts from my perspective.
We're happy to honor the decisions of the community as to whether to continue MO on SE 1.0 or migrate to 2.0. From our point of view, we'd like to see you move to 2.0, but as I've said to Anton many times, we'll continue to do whatever we can to keep you running on 1.0 if that's your choice.
From reading everything here, I think that the concerns people have about switching to 2.0 all revolve around the issues of governance. This is something I spend a lot of time thinking about. We have a fairly specific governance system in operation at Stack Exchange:
It's unusually open. All our discussion about governance takes place in broad daylight and includes everyone from the community. We don't make decisions in a back room somewhere.
It's unusually democratic at many levels. Sometimes that's obvious, for example, we're the only online community I know of where moderators are elected. More than that, though, there's a pervasive culture of democracy and openness at all levels in the stack, from voting on answers to voting to close and even voting on whether two tags are synonyms.
While I have the utmost of respect for Anton and think he has done an impeccable job, I think history teaches us not to build government systems that depend on any one person, no matter how wise or benevolent. I think SE has a good constitution, and we have good laws, and that they are robust enough to allow professional communities to flourish. That doesn't mean that one day a policeman won't do something that you don't agree with, which may or may not be in alignment with the constitution and the laws--this happens in every society. But I think we've set up the governance of Stack Exchange online communities in a way that should give you confidence that this will be a place where mathematics can flourish for a long time.
Finally, there's a lot of speculation as to why Stack Exchange (the company) wants MO to move from SE 1.0 to 2.0. Our motivation is simple: the 1.0 software is running fine today, but since it is not actively maintained, there are a lot of things that could go wrong there which would take an unpredictable amount of work to fix. We might suddenly discover a security hole which requires us to patch the old code. When this happens, we might discover that nobody remembers exactly how the code was working or where it's running or even how to compile it again. Maybe not today, maybe in 5 years. From an operational perspective, it's just easier if every site is on the same, actively-maintained code base. That's really all there is to it. Sure, there are a lot of great 2.0 features that I'd love for you to use, but as a company our main concern is just making our lives easier by getting us onto a single code base.
@Rachmaninoff: 1) bans are site-specific, 2) "founding father" is a very strange term for fpqc.
While I have quite an appreciation for fpqc and all, I think that calling him "very influential and renowned" is a little bit of a stretch...
@Mariano 'renowned' -> '(in)famous'?
I second Gerhard's remarks - if we were part of SE2.0 but a casual visitor (eg, mathematician of the highest rank who popped in now and again) couldn't tell the difference between the current platform and the new one, then that would go a long way to smooth the path to a transition.
@Daniel: If we migrate, I promise to continue to strut around like I own the place, at least to the same extent I do now.
While I have many of the same concerns that people have expressed here, the more serious thought I give to it, the more I feel in favor of migration. The main reason is that among those concerns, the most real ones are things we'll have to deal with anyway. I feel the potential cons of migration have been established, so I'll try to put forward the case in favor.
(I have to board a flight now. More later if I remember any other points.)
To address peoples' fears that SE Inc will interfere too much or even merge us wtih math.SE, it is worth pointing out that a very large percentage of the high quality answers on math.SE come from MO regulars. Seriously upsetting the MO community would result in a huge hit to the usefulness of math.SE, should those MO regulars stop contributing, and I am sure that nobody wants that.
As for the concern that we will be flooded by homework questions, it is worth pointing out that the possibility to migrate questions within the network will in fact keep the MO site cleaner. At the moment, questions get closed, but still take up real estate on the starting page. If they could be migrated, they would disappear from MO altogether, which is much more sensible. It is also a less radical reaction than closing, so people will have less inhibitions to actually do it. Since we have plenty of high rep users, there should be no problem with a starting page that is cluttered with junk.
Whew! Take a holiday for a couple of weeks and return to find that the World's Gone Mad.
Since memories are so short, let me start by introducing myself: I've been on MO almost since it's inception, I've been on meta since before it's (official) start, indeed the suggestion to use this forum software was mine. I've also been active on some SE2.0 sites, most notably tex-sx where I was a "pro-tem" moderator before stepping down at our moderator elections (and, in light of the above, I will make it clear that my reasons for stepping down were completely positive). I've tried a couple of the other SE2.0 sites but I'd only rate my participation as "significant" on two of them: TeX and photography. The reasons why my forays into other sites were not successful are all to do with the local community and nothing to do with the SE overlords.
I can see reasons both for and against migrating. At the moment, my instinct is against. Here are my main reasons:
Joel says above about the governance: "It's unusually democratic at many levels.". I don't want a democracy. One of the great things about MO is that it is run by mathematicians. It wouldn't have to be Anton, (but I pity the poor person who tries to follow in his footsteps!) but it would have to be someone who could quickly gain the respect of the already-existing mathematical community. I would worry that making the site governance truly "democratic" would mean that it could (gradually) lose its focus simply because there are far more people who think that mathematics is about finding "x" than there are professional mathematicians. There is a strong subcommunity of SO who think that sites should be defined by topic and not community and I have seen people meddling where they should not. Joel talks in places about viewing SE as a "university" but then he's also the one who proposed to the cs site that they expand downwards, so I don't see that he understands his own analogy. To make one thing clear: I don't see that anyone would seriously suggest merging MO and MSE, but I do see that it might be possible to get to the stage at which it is a serious proposition by which time it will be too late to reverse the process that got us there. So I would want a system that was less democratic than "one user, one vote". Maybe ones votes scaled (logarithmically?) with ones reputation. Yes I'm an elitist. But as someone recently said in the UK: "So a disproportionately large number of people in government have been to Oxford or Cambridge. Do you really have an issue with the fact that the people governing us are more highly educated than the average person?". Of course, I would want the system to be transparent, but not democratic.
The other big thing for me is that we would lose meta. I find the format of the meta sites constantly frustrating and a great example of fitting a square peg in to a round hole. There have been many great discussions on meta which would not have been possible in the, frankly bizarre, Q&A format that the SE2.0 meta sites have. One reason why I am so confident that MO is safe under Anton's benevolent dictatorship is that through the many discussions we've had over the course of the lifetime of meta.MO then I've been able to see a bit of how he thinks. He's even been able to convince me that he's been right. Once or twice. There is no equivalent to meta.MO on the SE2.0 network. "Chat" is too noisy and messages are limited in length (unless you know how to get round that). I notice that they are now introducing site-specific blogs! I gave up reading maths blogs because MO came on the scene and it was far more useful to me. But again, blogs aren't suited to the type of discussion that we've had here. Discussions are the grease of the community. But in SE, it is really, really hard to engage in a decent discussion.
(ctd next post)
[1] I admit I'm a little inconsistent here! I recently "upgraded" from Ubuntu to Debian precisely because I was annoyed at Ubuntu changing too many things every 6 months.
(ctd from above)
If it were purely an issue of technology, I would say "migrate today". Yes, it's annoying that SE introduces new features without telling us[1], but it only takes a couple of "on the ball" users to spot these things and inform the rest - we even have a thread on the meta.TeX site for this purpose. Yes, the FAQ is a bit annoying, but again, who reads the FAQ? And there are ways to ensure that the "Unofficial FAQ" (as we on TeX-SX call it) gets a link from the main site. The migration of questions from other sites isn't an issue either: only mods can migrate questions arbitrarily. For mere users, there are only certain channels that are allowed and these channels have to be "opened up" by the overlords. So as long as SE agreed not to open a channel without getting the agreement of the MO mods (which, really, ought to be standard operating procedure on all SE sites), then I don't see a problem. We could have a channel from MSE, one from CS, and that's it.
In summary: the method of governance and the loss of discussions are my stumbling blocks.
Most of the discussion so far has focused on governance issues. I do share these concerns, but I am also worried about technical issues associated with migration to SE 2.0. Let me point out a few such issues:
Inability to edit the faq. I mentioned this one before, but it is sufficiently important to repeat it. A great part of the faq on SE 2.0 sites is fixed. Sites can only modify a small part of it. Various aspects of the fixed part are inappropriate for an academic site like MO. For example, the bit on self-promotion conflicts with how academics usually cite their own work.
Recalculation of reputation. Reputation is calculated in a slightly different way on SE 2.0 sites. For example, question upvotes only give 5 points instead of 10 points. I don't think good questions should get half the credit of good answers. Worse, upon migration, reputation may be recalculated to reflect this. Excellent question askers, like Joseph O'Rourke, will be seriously affected by such a recalculation.
Handling of images and other links. I recently learned that SE 2.0 systematically alters certain links for commercial purposes, which is just plain wrong for an academic site. Images are also handled differently on SE 2.0. I am worried that this will seriously affect many excellent questions and answers from the MO database.
Math rendering issues. This will hopefully improve soon. SE 2.0 currently has some speed issues with math rendering. As far as I can tell, they are not willing to implement selective disabling of math rendering like we currently do on MO.
Since this came up I believe already twice as a negative point, let me just say that I consider the potential change in the points-calculation as a positive one. Indeed, if I were to decide I would be for a quite different points computation, something like Q: 2 (up), -4 (down) A: 10 (up), -8 (down). However, I do not care much about the points in general.