Theoretical Computer Science

Aaron Sterling

5:00 PM

Hi Hsien-Chih

Hi everone

Hi Kaveh

Shall we start?

yes :)

Cool. I'll be around for a few more minutes.

Kaveh

5:07 PM

@Suresh, are you here?

@JoeFitzsimons It is nice that you and @AaronSterling are here

Let's start with the blog

Aaron Sterling

The blog!

Rebecca Chernoff

blog? where?! q:

Aaron Sterling

This is what I see: Joe and I would primarily encourage people to write posts, and make brief posts ourselves about updates on the site.

Example of brief post: these are the three questions with outstanding bounty this week. Bring us the answer, dead or alive.

3

Kaveh

@RebeccaChernoff here: cstheory.blogoverflow.com ;)

Rebecca Chernoff

lol

Hsien-Chih Chang 張顯之

5:12 PM

highlights of the weeks looks nice

Kaveh

@AaronSterling, do you think weekly posts + weekly updates is a reasonable schedule?

Aaron Sterling

It depends how many people sign up on meta to contribute. Just Joe and me doing that, no. Joe, me, and 4 other people to start, yes.

Kaveh

so we need people to sign up! we should set a system message linking to @Suresh's post, meta traffic is not high as you know :(

Joe Fitzsimons

@Kaveh, I would think certain posts are easy: like Aaron suggests "unanswered questions" and "hot questions" would be easy ones to generate a post for every week, so even with low numbers of contributers, that should be easy enough.

Aaron Sterling

I would like to plug for Theory B people. In part because some of the weightiest answers are in theory B. Also, I often don't understand those answers as well as I'd like, so more exposition would be fantastic.

Joe Fitzsimons

5:17 PM

However, more detailed posts require more work, and hence more contributors if we wish to maintain a steady stream of interesting material.

Suresh Venkat

sorry. will need to be in and out

we need to co-opt Neel or Dave :)

Hsien-Chih Chang 張顯之

At what levels should our blog be? That is, should the posts be a survey in one topic, a research idea, an introduction for the one who is new to an area?

Aaron Sterling

Andrej has produced some phenomenal answers too. As have others.

Suresh Venkat

@HsienChihChang張顯之 yes

:)

Aaron Sterling

I think there could be more than one tag. Say "general audience," "algorithms," "programming languages." So the intended audience could be more or less what the contributor (and editors) wanted that post to focus on. People could sort depending on tags.

Kaveh

5:19 PM

@AaronSterling and he is very good blogger also: math.andrej.com

Aaron Sterling

Thanks, Kaveh, I have never seen that.

Joe Fitzsimons

Maybe Neel, Dave and/or Andrej could be persuaded to write a post on an interesting Theory B topic that has come up on the site?

3

Rebecca Chernoff

keep in mind that not all of your posts have to be about questions on the site

Suresh Venkat

some could be about new and exciting results

2

Joe Fitzsimons

@Rebecca: Yes, I gathered that, and it would be nice to maybe highlight recent interesting preprints (ala the arxiv blog/this weeks finds). There is however a lot of interest in expanding our theory B side.

Rebecca Chernoff

5:22 PM

sure, was just saying (:

Stack Overflow Blog: Blog Overflow

Posted by Rebecca Chernoff on June 23rd, 2011

Every Stack Exchange site starts with a Q&A site, made up of three pieces that help bring the whole community together:

bicycles.stackexchange.com, the main Q&A site

meta.bicycles.stackexchange.com, questions about community and administrative matters

chat.bicycles.stackexchange.com, the third place for real-time collaborations

But wait, there’s more?

A couple months ago, the Super User community took it upon themselves to create a blog run by the community. This effort has been so successful that a couple other communities floated the idea amongst themselves. Internally, we recognize …

lists some ideas for topics

Suresh Venkat

yes and of course @LevReyzin is ready to do a 'this week in learning theory' post ;)

Aaron Sterling

On the other hand, we are sitting on a computational geometry goldmine. No reason not to make use of that too.

@Lev you are? That would be great!!!!

Suresh Venkat

;)

what comp. geom goldmine ?

Aaron Sterling

Multiple top people asking and answering questions.

Lev Reyzin

hi sorry im here

I don't think we get a post per week in learning

Suresh Venkat

5:24 PM

i was joking :)

Lev Reyzin

but i'd be happy to do this month in learning theory :)

Aaron Sterling

Right now we are talking about a post per week in the blog. So if four people do a post a month, and Joe and I add in shorter announcement-type posts, we have some real content.

Rebecca Chernoff

\o/

Joe Fitzsimons

Maybe that would be a good idea: This month in 1) Learning theory, 2) Complexity, 3) Comp Geometry, etc.

2

We could cycle through them at a rate of 1 a week.

Kaveh

@AaronSterling I think starting with biweekly schedule is safer start.

Joe Fitzsimons

5:26 PM

Or more frequently if the number of topics grows.

Suresh Venkat

queueing up drafts is always good

from my personal experience, having drafts lying around makes things easier

Joe Fitzsimons

@Kaveh: biweekly meaning twice a week or once every two weeks?

Suresh Venkat

as long as timeliness isn't an issue

Joe Fitzsimons

I'm in favour of the former. The more activity initially the more regular readers we'll have.

Kaveh

@JoeFitzsimons one weekly high-lights post + one other post every two week

Suresh Venkat

5:27 PM

I wanted to make sure we discuss one issue we didn't have time to discuss last time: @VVinay brought up the idea of expanding our scope downwards a bit

Joshua Herman

a blog for the tcs stackoverflow?

Rebecca Chernoff

@JoeFitzsimons this is so annoying. It really bothers me how ambiguous that is. ):

pet peeve hehe

Aaron Sterling

Yes, I was thinking of queuing drafts too. I don't intend to write posts if there aren't any. I am behind on my own blog schedule as it is. I am fine with any schedule literally, as long as there is the community participation to justify it.

Suresh Venkat

@VVinay did you want to elaborate ?

Joshua Herman

Maybe we could have the blog collaboratively written like using a etherpad or google docs?

V Vinay

5:29 PM

Thanks Suresh. As I was saying in the morning, if we are cstheory, we should have much of the "standard" cs material on our site so that search brings new people to the site

2

Joe Fitzsimons

@Joshua: I'm not sure how well that would work, as it would substantially extend the time taken to write a post.

John Moeller

Aside: My labmate just pointed out that @RebeccaChernoff 's last name is really apt for our community. :-)

Lev Reyzin

Collaborative writing is too slow... I remember the SIGACT article was a bit of a pain

Joshua Herman

Yea thats probably a bad idea. nvm @JoeFitzsimons

Rebecca Chernoff

@JohnMoeller o: ?

Suresh Venkat

5:30 PM

Chernoff bound: famous theorem in probability

Rebecca Chernoff

googles

Lev Reyzin

like this the author of an article can attach his/her name and then they can be blamed for anything that's wrong :)

Kaveh

@VVinay interesting point, the effect of non-research-level questions on our ranking on Google

Joe Fitzsimons

As regards bringing in more people via more basic material, I would suggest that this is really what the Tag Wikis are ideal for

Suresh Venkat

since we're not organized like a wiki, the only way to get to where @VVinay wants us to go would be to allow more basic questions

Joshua Herman

5:32 PM

Wait what if for the blog we could pick the best questions on the site

Like best question of the month

Suresh Venkat

@JoeFitzsimons how ?

Aaron Sterling

@VVinay: Do you mean repeating some of what is in eg Wikipedia? I could see this site as complementary to Wikipedia. A question like, "What does this Wiki definition mean to me in my semi-practical setting" seems like something that should be allowed here.

Suresh Venkat

@JoshuaHerman indeed. that could even be by voting

Joshua Herman

or even best question(s)

Rebecca Chernoff

6

Q: Top Question of the Week #23

Now we're into week 23 of the "Top Question" of Super User in the Super User blog. Please post and vote for your favorite question for this week. Please post any question that you feel is of worth and the reason why. Try not to promote your own questions or answers for publicity sake. We are l...

Joe Fitzsimons

5:33 PM

@SureshVenkat: Well, we could essentially turn them into an introductory article on the topic, couldn't we?

Kaveh

I am personally not in favor of extending the scope downwards

Suresh Venkat

which makes me wonder: I don't think we want to (or can, or should) supplant Wikipedia

John Moeller

@Kaveh Why?

V Vinay

@AaronSterling No. Our problem is not too different from stackoverflow vs a man page. Am I making sense?

Rebecca Chernoff

blog.superuser.com/category/qotw

Suresh Venkat

5:33 PM

@JoeFitzsimons maybe, if wikipedia doesn't address it

But I'm loathe to duplicate content

Aaron Sterling

Not to me, sorry. But I am slowwwwwwwww.

Rebecca Chernoff

note that blogging about a question on the site shouldn't just be rehashing the content on the site.

Aaron Sterling

Could you say it differently please?

Joshua Herman

I would like to keep the scope the same

Kaveh

@JohnMoeller I think it might drive expert users away. There will be too many basic questions. I don't like cstheory overridden by undergrad questions.

Joshua Herman

5:34 PM

yea maybe a discussion about the question

Joe Fitzsimons

@Suresh: Maybe. I'm not really in favour of lowering the standard, but I can see why we may want a standard bit of introductory material.

Rebecca Chernoff

delve deeper into it, explain how you arrived at an answer, things like that

Joshua Herman

yea

Joe Fitzsimons

With that, I need to run off. I'll try to follow this conversation from my phone.

Rebecca Chernoff

if it is something that can be tested/done/implemented, talk about results, etc

John Moeller

5:35 PM

@Kaveh I don't see why we can't accomodate both and let things settle out naturally

Aaron Sterling

The posts on my blog about questions that have gotten the most hits have been the ones that are "autobiographical." What I thought about the issue before the question, and how the question and answers affected my subesequent thinking.

Kaveh

@JoeFitzsimons thanks for joining us :)

John Moeller

@Kaveh To expand on my point a little, I don't see why we can't accomodate FAQ-type questions from undergrads. I think it's silly to exclude those. Great answers to simple questions can still be informative to everyone.

Kaveh

@JohnMoeller I think they check cstheory to find interesting questions, if say 80 percent of the questions are undergrad level questions that would drive them away.

John Moeller

@Kaveh How do you know?

Aaron Sterling

5:39 PM

This suggestion of JohnM has been debated to death on Meta. There is no way in the world I would support that change. VVinay sounded as though maybe he was saying something different.

Kaveh

@JohnMoeller we have discussed before on the meta, that was my impression. Also check Noam's blog post. I think Math.SE is a better place for such questions.

V Vinay

@AaronSterling Let me try ...wikipedia is usually factoid, but not much conceptual exposure to ideas happen. Can our site work with the important concepts/insights into cstheory?

Joshua Herman

As an undergrad I usually look for interesting questions on the cstheory site. I would look somewhere else for regular questions but that is my humble opinion

I guess if we made a community wiki that would be interesting for FAQ questions.

Suresh Venkat

so @VVinay you want to allow basic questions that address insight ?

V Vinay

yes. if that is possible.

Suresh Venkat

5:40 PM

the somewhat rigid Q/A format limits our ability to get into this I think

in other words, you're asking for discussion/insight

Aaron Sterling

@VVinay: DO you mean like Russell I's answer here? cstheory.stackexchange.com/questions/2434/…

Kaveh

It would be very very difficult to draw a line. We already deal with lots of misconception about what is on topic here, allowing such questions would not help. On the other hand, I have no problem with good insight questions, we have discussed this also previously on the meta somewhere IIRC.

Suresh Venkat

that's a good example

also this: cstheory.stackexchange.com/questions/2136/…

@RebeccaChernoff how do you add questions in that nice bold format ?

V Vinay

Yes, the padding example is a good one.

Rebecca Chernoff

for oneboxes, the link needs to be the only thing in the message

Suresh Venkat

5:42 PM

so then we do admit certain questions along those lines. the problem is that these questions are more likely to be asked by experienced users

(experienced in content, not site usage)

@RebeccaChernoff thanks

Aaron Sterling

Perhaps the mods could encourage questioners coming from a good place to modify their questions to be about "expert insight" instead of a specific technical answer to a low-level problem. I don't know whether that addresses VVinay's concern though.

2

Suresh Venkat

in fact many of the big-picture tagged questions might be of this nature

cstheory.stackexchange.com/questions/tagged/big-picture

V Vinay

Let me be a bit controversial here to make a point; padding is an "elementary" in complexity theory but an "equivalent" algorithms question, in my opinion, would have been shot down :-)

Suresh Venkat

hmm

Kaveh

we have 15min, let's leave this topic and move to the next one (we can continue to discuss this on the meta)

Suresh Venkat

5:44 PM

interesting point

I won't deny that we seem to have double standards based on area expertise

but I agree with Kaveh: let's start a meta discussion on this. @VVinay ?

V Vinay

sure

Lev Reyzin

unfortunately, i have to run, but ill check the transcript to see what gets decided on about the blog and other things. - Lev

Kaveh

We have discussed the downvoting without comments in the previous meeting.

About the involvement of users in managing the site

Aaron Sterling

Also, I see no reason to limit the scope of the blog to research-level TCS. Exposition of "more basic" topics seems like a great community service.

4

Kaveh

is it only me or do you also feel that there is a drop in the involvement by high rep users?

2

Suresh Venkat

5:47 PM

and might help newer folks get into the swing of things by providing a forum

Joshua Herman

Goodbye I might pop in later.

Suresh Venkat

@Kaveh kind of. but these things move in phases I've noticed

Kaveh

@AaronSterling and off-course we can have posts starting argumentative discussions which are not suitable as a QA question :)

Aaron Sterling

@Kaveh: I wanted 3,000 rep to be able to vote to close. Once I got there, I stopped being as interested in trying to answer questions. I think a lot of the "old hands" were pushing to get out of beta, and once we accomplished that, there were other things to focus on.

Kaveh

when I checked the users with +1000, almost all of them (with a exception of a few recognizable names :) check the site at least weekly

some of them used to be much more involved in managing the site

Suresh Venkat

5:50 PM

I think the site gets enough managing, to be honest

people roughly have a sense now of what goes and what doesn't, and I've been finding that questions get closed even before I have a chance to see them

Kaveh

@SureshVenkat yes, but I feel that we (in particular I) might have done something that has discouraged them from participating

Aaron Sterling

@Kaveh: There is much less to manage now. We have pretty much all the policies we need.

@Kaveh: Don't be silly.

Suresh Venkat

I think all is well on the management front. there are other things we should work on of course

unless others feel otherwise :)

Kaveh

@SureshVenkat like promotion?

I was also thinking about this

Suresh Venkat

right, but we've already discussed this last time

Aaron Sterling

5:53 PM

We kinda skipped the downvoting thing. Is that ok?

Kaveh

so far we have mainly focused on classical places for cstheory (top US/EU+Israel universities)

@AaronSterling it doesn't seem so

I think cstheory can be quite useful for people who are not as close to conferences as we are

e.g. India, China, Middle East

Hsien-Chih Chang 張顯之

Indeed!

Aaron Sterling

There are a fair number of questions from students in IITs, or so it seems to me.

Kaveh

it can be much more considering their number

the problem is I don't have any good ideas on how to promote cstheory in these places

Aaron Sterling

Heh. Maybe VVinay does. :-)

Kaveh

5:57 PM

(unless they read theory blogs)

V Vinay

There are a decent theory community in India, I can do the promotions on FSTTCS and IARCS.

2

Kaveh

Asian conferences can be a good place for promotion! :)

Suresh Venkat

IIT just produced the new maximum matching result that everyone's pondering

Hsien-Chih Chang 張顯之

Language is a barrier. Most of my labmates don't want to ask questions merely in English. But if there are enough interesting/useful questions, they will :)

Kaveh

we are running out of time, maybe we should continue this on meta

@HsienChihChang張顯之 reminds French researchers ;)

V Vinay

5:59 PM

@SureshVenkat pointer? seem to have missed the paper ...

Kaveh

I guess we have discussed the main topics in the agenda for the meeting

Aaron Sterling

Okay, everyone, don't leave Joe and me stranded with the blog, please! And I am excited by the suggestion to publicize at FSTTCS etc. I am signing off, bye everybody.

Kaveh

@AaronSterling bye Aaron and thanks

If you have any other issues or topic this is the time to bring it up :)

So I guess that is it

thanks @everybody

V Vinay

Kaveh, close the session??

thank you all ... bye

Hsien-Chih Chang 張顯之

thank you everybody!

Suresh Venkat

6:11 PM