« first day (1 day earlier)   

12:30 AM
I may be around for some time during this chat, but it will not affect my votes because I have already cast the three votes.
 
 
18 hours later…
6:34 PM
We'll get started in about half an hour.
 
7:00 PM
Hi. (:
 
I'm here, on my Droid, all 21st century-like.
 
Greetings.
 
hi everyone
 
hey./
 
7:01 PM
hi
 
Let's get started!
 
Suresh joined. It's now a formal meeting :)
 
Welcome to the CSTheory Town Hall Chat
 
oh boy
 
looks like most of the people here are the candidates :)
 
7:02 PM
except for ross
 
I'm just doing civil duty.
 
yeah, so let's not ask too hard questions ;)
 
i guess there will be a transcript left afterwards
 
We're here to get to know the candidates and discuss any questions people may have that might help with voting.
 
I saw Jukka I thought.
 
7:02 PM
Just lurking...
 
The candidates I see here are: @DaveClarke, @LevReyzin, @SureshVenkat, @SadeqDousti, and @AaronSterling. @Kaveh was able to join partway through.
 
correct
missing: Kaveh, Neel (who said he couldn't make it) and ....
 
ah yes
so now what ?
 
Anyone have any questions to get us started?
 
7:04 PM
Hsieh chih who said the time was hard for him.
 
I vote that the four users here ask questions
 
I agree
 
I've got questions to fill in as needed as well. (:
 
7:05 PM
Should we be worried about this situation? Happy about this?
 
Can you elaborate?
 
the number you mean ?
 
Can you clarify your question for the transcript please?
 
119 vs 61 and 5 ?
 
The numbers are funny. We have got > 100 unanswered questions. For example, tex.se and photo.se have exactly 0 unanswered questions.
 
7:06 PM
I think there's little open problems on Photography / TeX w.r.t TCS
 
the polynomial hirsch conjecture is harder than latex questions, i think
 
Wimp!
 
prove me wrong :)
 
MO has 1865/15180
we have 119/1255
 
Oh, I should be more formal, I suppose. Sorry.
 
7:08 PM
This partly could be reflected by the relatively youth of our community, and perhaps that our questions can be more open than TeX and photography ones.
 
if we view the two communities as comparable
then we are not off track in a relative sense
 
you mean with MO? I agree
 
Also, and maybe Rebecca can help, I'm not sure what the defn of unanswered is
 
So all candidates think that this is ok, this is the way that things should be, and the difference between us and tex/photo is just that we are comparing apples and oranges?
 
@LevReyzin yes with MO
not quite
 
7:09 PM
Questions without an answer marked as accepted or upvoted.
 
I think that we need to educate people more about accepting answers that are appropriate
and doing more upvoting
that would reduce the numbers a bit
 
I totally agree
 
I think it's fine to have unanswered questions. It's actually good that people aren't asking only easily-solvable questions.
 
Certainly one way of helping old questions find answers is by putting them back on the front page.
 
beyond that I suspect it's a function of being a research community
 
7:10 PM
There are many acceptable answers which are not marked accepted
 
with an emphasis on research-level questions
 
How would you handle a question raised on meta regarding your moderation actions?
 
criticism, u mean?
 
I think I would try to leave it to the other moderators -- they'd be less biased. In general I wouldn't treat myself any different than another user, to the best of my ability.
 
Yes, someone unhappy.
 
7:11 PM
I think we should produce more unanswered questions, at least in the short term. We had a lot of big list questions in beta. More research questions from new areas means more things the regulars don't know.
 
If I agree, I'll do what is expected from that user;
if not, I ask people to vote, and will respect the result
 
(Thinking of someone who disagrees that you closed a question for example)
 
I'm with Sadeq on this one
 
I mean in general, I think people can express unhappiness with moderators on meta. And I'd try to incorporate suggestions that people like.
 
Point the to appropriate part of the FAQ and moderator guidelines that guided the particular moderation actions.
 
7:12 PM
I don't want a repeat of the meta.math fiascos
one thing that I think is helpful is to enshrine policies
when issues come up
then, as Dave points out, we just point to the policy
it's no longer personal
 
Agree about making formal policies.
 
Be humble, listen, patience.
 
I'm not sure I agree. There's always room for interpretation and judgment. Even if we just point to policies, people might be unhappy.
 
I don't think it's possible to ensure that people are happy
 
"was this really a homework question" etc.
 
7:14 PM
only that they are treated fairly
 
I think in general, people have been very happy @ meta.cstheory.
 
I've seen some major insulting of me on SO :)
over some homework questions
 
I mean, surprisingly happy if we keep in mind that we are on Internet...
 
aha: Kaveh is here
 
Insults never really help anyone. Avoiding getting into arguments is certainly wise.
 
7:15 PM
Hi Kaveh
 
How about a question that another mod closed that you feel shouldn't have been? How should that be handled?
 
Really? Good thing you closed their questions then. We don't need such immature people here.
 
Welcome @Kaveh.
 
I'm inclined towards transparency: open up a meta post
and discuss it there
 
One probably shouldn't overrule other moderator's actions, even if we don't agree with them.
 
7:16 PM
I might just contact the other mod to see if we actually disagree. If there is disagreement afterward, then I'd take it to meta.
 
At some point in time (meta.cstheory.stackexchange.com/q/378/873), I suggested that mods do not close a question all by themselves.
 
yeah moderator fights are no good :)
 
I suggest to alot some time so that non-mods can vote on closing
this will be more democratic
And people insult less :)
 
Sometimes leaving a question open can be harmful to the community though. Where/how do you draw the line?
 
I've closed posts unilaterally for two reasons: (1) when I felt it was so obviously off topic that I felt that it was merely clutttering up the front page and (2) when there were 3 close votes or more and it was only because we didn't have enough 3000+ voters to do the closing
 
7:18 PM
unless it's blatant spam or insulting, i like the idea. if a mod is the only one who thinks a question should be closed, there's something wrong.
 
Actually, that's good. Let the community decide more on closing.
 
Agree with Dave. Also a bad idea to disagree in comments to the original question. Discuss with mods in private, or take it to meta if disagreement cannot be resolved.
 
I'm hoping as time goes on, the need for (2) goes away
 
It would be nice if a mod could cast non-binding close votes.
 
Junk questions do seem to fairly quickly get downvoted. Even on Sunday morning.
 
7:19 PM
Note that as a moderator, there can be a private chat room that only the mods can see. This can be used for discussions such as this, also dealing with spam, etc. Most things we do want out in the open, but there are some situations that are for mods only.
 
@DaveClarke I'
ve been noticing that
So here's my current position (and people can tell me what they thing): I want to make sure that if (say) Christos Papadimitriou or Moshe Vardi comes to scan the site, they aren't turned off
 
Having an odd number of mods, the mod-team can always resolve issues by voting among themselves
 
Hi everyone
 
100% agree
 
and hence am willing to close unilaterally if the questions are crazy
 
7:21 PM
I agree with @Sur
 
But I have a somewhat politically-incorrect question: recently there has been a lot of unhappiness @ meta.cstheory regarding the stackoverflow.com maintainers/admins. What's your take on this? Are you loyal to SE?-)
 
sorry @Suresh
 
What about off-topic questions that get answers and upvotes, signaling that the community likes the question?
 
I have to say that I'm not entirely excited about how SE manages things
 
I mean, not everyone has been too happy with the oddities of the voting platform, for example. And the way SE people have reacted.
So on who's side are you?-)
 
7:22 PM
there's a certain unwillingness to adjust to the needs of different communities. We are not the SO community: we are smaller, more well behaved :), and less likely to cause ruckus
 
Whether a question is off-topic or not seems to be always debatable, so if the community likes the question, it should stay. We are all not capable of judging questions in each others' areas.
 
I'm on cstheory side
 
@Jukka: no. They have not been managing community concerns well. But I don't see how that affects the next few days.
 
@Jukka
 
Regarding off topic questions, I'm a little more hardline. In fact that's probably the only issue I'm hardline on
 
7:23 PM
Overall, it would be nice if they were more responsive. On the other hand, it's not so terrible under SE either. We can see how it goes, but I am not "loyal" in that I am open to change if our needs aren't met.
 
How does it differ? Can mods do anything against SE overloards?
 
specifically, I think it's critical that we man the battle stations regarding level of questions.
being research level keeps a lot of stuff out that I am personally not interested in answering
 
We're on your side too. (: But I'm not sure there's a need to get into that during this discussion.
 
and that's the one issue I've had arguments on meta about
 
I found the voting/nomination problems a little unfortunate. I could see both sides, so decided to not say anything. The process seemed to model real world processes, so from that perspective they seemed reasonable.
 
7:24 PM
I agree on off-topic question. I also feel that the quality of questions has dropped recently to some extent
 
Usually in real world processes the candidates try to keep each other out. I like that here the candidates encouraged each other :)
 
that's because it's beginning of semester ;)
 
We need to close these questions to keep the level of the site
 
Relating to quality, what can you do to keep the quality on the site up, which will help attracting professional users and users that will stick around?
 
Yes, but we also don't close the questions which are not research level
 
7:26 PM
@Sadeq: Mods can do something. We can always leave the SE platform and set up our own Q&A site. And then I would really like to know if our moderators would support the transition to a new platform or be against it. :)
 
I find that giving a newbie some hints in the comment, even though the question should be closed is good for trying to avoid any animosity.
 
e.g. a question which can be answered easily by Googling or checking Wiki or looking it up in an introduction book to the area (undergrad) should get closed
 
I think Suresh said it best. We want it to be the case that when top theorists come here, they want to stay.
 
Even if such hints are a link to wikipedia.
 
I think giving hints in comments is fine
 
7:27 PM
@Jukka: I remember a game where it said: "Dissertation's not an option" :)
 
but not answers: that only 'feeds the beast'
 
Dave, I am not sure I agree. Stating the way to find the answer or redirecting to math.SE should suffice
 
@Kaveh I think that's what Dave meant by hints ?
 
So what candidates are going to do with non-research-level questions? To close or not to close?
 
I'm all against answering homeworks. That spoils the purpose of thinking.
 
7:28 PM
I think it is better to give some sort of hint, than merely to redirect (or say "use google")
 
Yes - we dont want incentive for HW to be asked here
 
@OleksandrBondarenko I think it depends on the question: some non-RL questions are about modelling, which I think SHOULD be allowed
 
@Jukka: it is a cost benefit analysis. If it gets to the point that someone researches exactly what it would take to move, we could discuss it seriously. I hope Rebecca et al got the message. If a handful of active users got as pissed as I did over how Rebecca blew Lev off, then, yeah, I could see us moving. Hopefully unnecessary.
 
@Suresh What do you mean by non-RL modeling questions?
 
Yes, when I say research level I include all those exceptions we have already discussed on meta :)
 
7:30 PM
like questions of the form "I have this real life problem and I'm trying to model it theoretically"
 
sounds like research to me
possibly
 
If a question doesn't belong, leaving it open with a comment linking to google seems like a bad. We want to be the destination (for questions that belong), there are enough sites out there that send you on goose chases across the internet, imho.
 
I see what you mean now
 
By the way, should we encourage or discourage questions that ask for HW problems (not HW solutions)?
 
but I think Dave said that we should close it after posting a link ?
 
7:30 PM
Some of the potential HW questions, or at least the questions that are not at a research level, are from potential future users.
 
Say someone comes on the site and asks a question that is below your standards, and you're rude about it and close the question. What will happen in a few years when that person is a cs-reasearcher? Will he or she come back?
 
@JukkaSuomela: what do you mean?
 
That's a good question, @Jukka, and I don't think we've discussed it as a community. I'd be hesitant to make a policy on my own.
 
Indeed, close after posting a link. In effect saying, we want you here, but this is not the right kind of question.
 
What Dave said :)
@Jukka I ask those questions of colleagues all the time
 
7:32 PM
We close them politely. It is not rude to state that this question is not appropriate for this site.
 
a priori I wouldn't have a problem
 
@Banang the point is that we'll try not to be rude. There's never reason for that.
 
@Lev,
 
@Banang: we seem to be thinking along the same lines. We don't want to discourage anyone.
 
Do you think the distribution of topics popular in this SE is healthy and if not what do you think can be done to spur changes?
 
7:32 PM
@Sadeq: "I need to have some HW problems for my students for my course on XYZ, could you suggest something?"
 
@Sadeq: Which certainly is research-level, but it is also a bit like "could you do my work for me".
 
@RossSnider it's not ideal. Too little theory B, too much complexity theory for one thing
 
@Ross we can encourage a wider participation from all areas of theory, like theory B, etc. But in the end it's not the moderators who control who asks what on this site.
 
Perhaps the complexity people could be quite for a month or so.
 
7:34 PM
@LevReyzin, and I think that's good. But the problems you see over at SO for example, is that a lot of people treat neewbies quite poorly. I think that's an attitude that develops and grows from closing too quickly, or not being willing to redirect and aid people asking questions that might not belong...
 
I won't close that as a mod, but I certainly discourage them if phrased as such
 
I know Neel is talking to Phil Wadler over at POPL trying to help with outreach to theory B
 
@Jukka, I think we should distinguish between "not appropriate" and "we don't like it"
@DaveL not a good idea! :)
 
@DaveClarke ;)
 
Having Phil Wadler on board would be a great drawcard for his fan base. (I'm here already.)
 
7:35 PM
@Banang I agree it's something to be aware of -- I guess power corrupts. I would hope we follow in @Suresh's example of helping new users and treating them well.
But that's why I would have liked there to be periodic elections to oversee moderators over time.
 
@Lev yes
moderators get old and crotchety
 
@Banand: we can have a standard post on meta that explains the situation politely
 
ahem
To get back to Ross's question, I'm not sure what else we can as mods
except encourage people to ask questions.
 
How about questions when asker doesn't state her motivation, even after request?
 
yes, that's a problem
 
7:37 PM
Hopefully with time, the distribution of questions will be a little more balanced.
 
when OPs don't reply to requests in comments
 
The best way to get more questions in an area is attracting more researchers in that area
The reason for having a large number of complexity and algorithms questions might be having more users familier with them
 
we currently have an informal 24-48 hr period to let the OP respond
 
If a question is not appropriately modified after a few days, then the question opens itself to.
 
@Kaveh Makes sense, but is kind of chicken-egg.
 
7:38 PM
To get any researchers in that area...
 
The best way to get more questions in an area is attracting more researchers in that area
 
to closure..
 
@Oleksandr It depends on whether the question can stand on its own. for example, some questions are suspected as HW until the questioner can present a motivation. In those cases, if the OP never gives more info, even when asked, then we can close.
 
@Oleksandr those questions tend to die off on their own. No one enjoys talking to a wall.
 
@Ross, yes as it is always
 
7:39 PM
but should we close zombie questions like this ?
 
Should we then close them or leave?
 
@Aaron, some users like to answer questions, even when the rest are waiting for motivation.
 
or should we just leave them
 
but things does not happen suddenly
 
What is something concrete that you think CSTheory can do to improve/grow itself?
 
7:40 PM
Don't forget, there's a "re-open" option. If a zombie question is closed for any reason, all bets aren't off yet
 
@Oleksander: it depends one the question, but I think closure of such questions is good for quality. If the asker does not care about the question, then why should we?
 
Affiliate with SIGACT!
 
working on it :(
robert cartaino is not replying to my emails
 
Affiliate with SIGPLAN, too. And ETACS.
 
FCRC?
 
7:41 PM
I think the ultimate goal is having more experts and researchers
 
advertise at FCRC for sure
 
Anyone going to advertise us @ FCRC 2011?
 
@rebecca publicity steps that I and others have outlined.
 
@Jukka good idea, and none so far
I won't be there
 
@SureshVenkat You've not been forgotten. (:
 
7:42 PM
SIGACT will help as other ways of promotion do, but the real promotion for attracting researchers is becoming an unavoidable resource
 
@Kaveh exactly
 
I suggest professors encourage their students to ask some of their RL problems here
 
@kaveh Well said!
 
for them. If we start being cited in papers then that would attract large number of researchers
 
but since we were asked for concrete steps, here's mine: "Establish formal connections with existing institutions in the theoryCS land
 
7:43 PM
who would like to come and read and post/answer questions on cstheory.
 
I have been emailed by people in my field who've stumbled CSTheory across answers to questions they were interested in via Google. (If only they'd joined up to.)
 
at least some professors are warning their students not to waste time here :)
 
:)
@Suresh: You don't, ha?
 
@Suresh: What? Maybe that's why mine students have failed to make an appearence!
 
Would it make any sense to have something like "mini-polymath questions"? Questions that would aim at solving an open research problem and which would lead to a published paper?
 
7:44 PM
@Jukka my feeling is that the SE format doesn't lend itself to this very well
 
Could it work? Would anyone want to see anything like that? Abuse of the platform?
I just thought that it might be a nifty way to promote us.
 
having said that, consider this meta-thread: meta.cstheory.stackexchange.com/questions/784/…
 
Gil tried that (I think) and his is the top voted open question, so we could try it. I don't think it's an abuse of platform to ask open questions. @Jukka you've solved some open ones if I remember correctly :)
 
@jukka yes, and I think Gil Kali tried asking a couple
 
Should we impose upper bounds on answers for easy questions?
 
7:45 PM
upper bounds?
 
@OleksandrBondarenko I don't know what that means
 
You mean a limit? I'd say no. The best answer should make its way to the top.
 
It just occurred to me that if we could give a talk (e.g., @ FCRC) on a result that was co-authored by cstheory.se, it would be a nifty way to promote us...
 
@jukka, I agree with Suresh. I think it is OK if someone does that but personally think it is difficult to manage a polymath on SE sites, though I should mention I don't have any exprience regarding it
 
@JukkaSuomela Nice idea
 
7:46 PM
we may want to ask Tim Gowers or Gil Kalai about their exprience
 
How do you think your asking/answering abilities and time will be affected by being a mod?
 
Yes, I mean a limit.
 
@JukkaSuomela Sounds like a good thread for meta. (:
 
I'm running a mini polymath right now in my research group
 
Many such things could spontaneously arise.
 
7:47 PM
it takes a software platform that is more interactive and threaded than SE
 
(Aside: Suresh you have me really curious)
(Aside: I read the tease of a blog post)
 
yeah :)
 
@Rebecca I could see it going both ways. As a mod, I'd spend more time on the site, so maybe answer more questions. On the other hand, I'd have more work as a mod. We'll have to wait and see what happens when the new mods arrive, I think.
 
I'm not going to say anything more here ;)
I think I'd get more time ;)
because I'd have more mods to help
although I should say that all the people here do mod level help already
 
I think many of us were doing (half)-mod jobs during last weeks of beta
 
7:49 PM
my main worry is that once this election is over, all the folks who've been helping out will stop "because there are mods"
 
I don't expect it to be much different from it
 
I won't stop helping. It's in my blood since I was a playground policeman in kindergarden.
 
@Suresh: I don't think it will happen
 
I think we'll all try to help either way, @Suresh.
 
ha ha
 
7:49 PM
@DaveClarke lol
 
We've got about 10 minutes left before I need to scurry over to the SO town hall.
 
It's too soon to suggest, but as a future plan: Can we have a Q/A book on "Gems of CSTheory"?
 
Any other questions? (:
 
@sadeq why not? if blows can be books...
 
7:52 PM
Joe ORourke was suggesting that for the 'gems' big list
 
mods can organize the book. It'll make an excellent promotion
 
i thought it was an excellent idea
 
I think it is too early
 
My smart phone corrected blogs to blows. Oops.
 
but I'm not going to do it :) - need tenure first
 
7:53 PM
@Kaveh: me too. I was suggestion it as plan
 
it would be nice, but we need more gems
 
Yes, something similar to Dick Lipton's book from cstheory questions would be nice
 
first let the chickens hatch, then start counting...
 
actually the post I'm referring to is about gems of cstheory in general, and has a nice long list
and is already ready for book-ification
but maybe this is another meta thread
@Rebecca can you ask Robert to email me :) ?
ACM is wondering what's going on
 
Maybe when there are 5,000 Q/As, a book can be made
 
7:55 PM
hmm.
another 1.5 years to go
 
Enough time for tenure.
;-)
 
Yep. But its good to have a plan in mind
 
we're hoping for accelerating returns. maybe not 1.5 yrs.
 
@SureshVenkat, yeah, I'll talk to him. I know he's been passing me stuff because his inbox is overloaded. (:
 
A book only needs 100 great questions, however long that takes.
 
7:56 PM
when you become a mod you'll get to see the SOOPER-SIKRIT google analytics
 
so all we need is Noga Alon to join, and we'll be done in a day :)
 
that tells us the rate of questions and answers
@Lev YEAH !
 
Agree with Aaron.
 
I'll nominate the one about flipping quantifiers
 
@Lev :)
 
7:57 PM
yeah that number sounds about right (anchoring effect...)
 
Before time is up: I wanted to suggest @Suresh to teach a few of his trciks to the new mods :P
 
always count to 10 before replying
and then count to 10 again
 
lol
 
:)
 
7:58 PM
and before doing anything, see if doing nothing will help :)
the more we train the community to do modding, the less we have to do BWA HA HA
 
and abracadabra
 
Absolutely. Community moderaton ftw. Mods are human exception handlers! blog.stackoverflow.com/2009/05/a-theory-of-moderation
 
Alright everybody. The SO town hall is starting now, so I've got to run.
Thanks everyone for participating!
 
thanks @Rebecca
 
8:00 PM
thanks Rebecca
 
Thank you Rebecca.
 
I'll be updating my meta post with a summary, but it'll be later today.
 
Thank you
 
Bye everybody.
 
Thanks for the chat. Back to Deadwood for me.
 
8:00 PM
Good luck to the candidates. (:
 
Thanks for ur time
 
(: :)
 
bye everyone
 
bye all
 
Ta Ta...
 
8:01 PM
bye
 

« first day (1 day earlier)