This niching is completely wrongheaded. As I wrote in a comment on this blog post, the last thing we need is for ML people, NLP practitioners, and statisticians to communicate less. The fact that these groups attend different conferences is a BUG not a feature. My site was designed so that these adjacent fields can crosspolinate information.
As we’ve learned from StackOverflow, it’s better to pick a broad topic and have a lot of experts in one place, communicating and helping with each other’s problems.
As we’ve learned from StackOverflow, it’s better to pick a broad topic and have a lot of experts in one place, communicating and helping with each other’s problems
Yeh, solid point. All the current sites are highly active because of their broad domain overview.
I don't think that it's a foregone conclusion that Area51 generates niche sites, just because it generates niche site proposals. I for one believe that it's the big blockbuster stackoverflow-size sites that will actually make it to the end of the Area 51 process and will actually get created.
And, if I were a betting man, I wouldn't bet against StackOverflow's monthly 7 million users. In fact I would be betting on the statistics Stack Exchange site launching in the next couple of weeks and getting critical mass very very quickly.
Not without an existing community to build off of you won't. You'll fail miserably. The people in the Area51 proposals are generally meeting there for the first time — they didn't know each other already.
MetaOptimize is building on top of an existing well-targeted community like MathOverflow did. Really, just like you did originally with your and Jeff's blog communities. AskMetafilter is successful because it was an outlet for an existing community — it wasn't attempt to start a new one.
Area51 is scavenging small clumps of people that happen to overlap with your existing main sites. That your sites are so QA-focused makes it very difficult for an actual community to grow within it — there's no outlet for chat and friend-making, for fucks sake even your moderation site is the same QA format! At least you've integrated it a bit more so that Meta isn't a completely separate community with it's own account system…
>if Area 51 had their druthers, they would have fragmented the community into five niches
I agree. I would see these new stackexchange clones every other day at /r/machinelearning. It would have hurt the ML community if all of the sites started competing for eyeballs.
Anyway thanks a lot for the initiative. I wish you all the best.
> I do believe that the stackoverflow people have (much) more experience in handling such websites then Joseph. I can very easily trust them to do regular database backups
Except, you know, for that time when the stack overflow guys lost everything on their blog due to "catastrophic data loss" and also subsequently found out that their backups were toast, too. :P
I kid, I kid...
Just use the site that is currently operational, this isn't really a life-altering decision. (Right now I am agonizing if I should submit this comment to hacker news or reddit.)
EDIT: On an unrelated note, my IP appears to be banned from StackOverflow.com. I have never used/abused the site (I don't even have an account there) outside of browsing a few times. It has been banned for at least half a year. Very annoying. Even the big guys get stuff wrong sometimes.
A while back I was in Guatemala and kept getting an HTTP 403. mailed them and it turns out they blocked the entire country because they were getting hit "1,000 times a day". After a couple more emails they unblocked it again.
I believe the data loss was on Jeff Atwood's personal blog codinghorror.com, not Stack Overflow, and it was due to outsourcing the hosting/backup to an (incompetent) third party.
Having listened to Joel and Jeff go on about the new StackOverflow ideas on their podcast, as well as in blog posts, it seems they're being utopian with the whole "people don't care about money or control, they just want to help" bit.
Running your own site and having control over how it works is important for a lot of people. If it weren't, geeks wouldn't self host WordPress for their personal blogs - they'd use Blogger or WordPress.com.
I'm glad that pg developed HN as a new, separate thing instead of rolling it out on Ning or whatever system was flavor of the day 1300 days ago. Likewise, it seems good that someone has a deep enough passion to set up MetaOptimize as a separate site, because they might actually bother to keep it up to date whereas SO might become the next Ning or Blogger over time.
No matter how bouncy SO's castle may be, true geeks like to build their own castles, even if they're not quite as shiny.
I'd be rather offended if I designed Stack Overflow. I've had my work carefully copied in the past, and even aside from the economics, it's not a good feeling.
This site needs to step back and decide which of the design decisions they made were because "we're duplicating stack overflow" and which are simply the straightforward way to accomplish whatever. It has the exact same layout, the same format, the visual similarity. Even the footer. Come on... at least design your own footer.
A large measure of StackOverflow's success-- perhaps the most significant factor-- was the immediate traction that came from readers of Joel and Jeff's blogs. SO had a ready-made audience, due to the notability of its sponsors.
It's not at all clear to me that this is scalable via the Area51 model.
Clearly, it takes more than 360 people to make a site successful. The question is: who is in the best position to attract the attention of the prospective users quickly and effectively? Time will tell, I suppose.
I think you are mixing up the fact that IT has a large userbase, while ML, NLP, etc; are more niche (relatively speaking). 360 users IS immediate traction for this area.
> Clearly, it takes more than 360 people to make a site successful.
Successful for it's founders, yes. But for me, the reader, I'm quite happy to have only the few thousand most-active researchers and professors and their students of Machine Learning, NLP, etc; post on metaoptimize.com. Otherwise, I might as well go to Yahoo Answers for generic, watered-down responses.
Yeah, I dunno if this area is pretty going to be successful for the SO guys, it makes a lot more sense for the community to run their own not for profit system.
One of the problems with the area51 process is that most of the people participating are from the current SO sites, meaning that it is always going to be very biased towards tech based setups without looking at what could also be mainstream successful.
The question is: who is in the best position to attract the attention of the prospective users quickly and effectively?
I'm willing to bet bravura at this point. Simply because he has the traction way way before the Stackexchange versions launch plus he has all the communities together. All it needs is a little momentum and Area 51 sites will struggle to catch it.
UPDATE: After email exchanges with me, Tal (the author of the post) updated it to include a more balanced discussion, and posted some quotes by me explaining my side.
Please also do not overlay a grey "JavaScript for Mobile Safari is currently turned off" layer that you cannot remove without Javascript. The site seems to render readable in the background. I am using Opera and only enable Javascript if necessary for crucial things.
c'mon dude, SO is pretty amazing. Poor answers are usually pushed to the bottom (though I have seen a fair few basic questions on SO with odd answers that no-one's bothered correcting).
The equivalent of php is dodgy bulletin boards that still doggedly come top of some google searches. And boy do a lot of them have the entirely wrong solution.
I'm starting to get to the point now where I ask SO first and then google, apart from Html questions.
machine learning (http://area51.stackexchange.com/proposals/7607/machine-learn...)
NLP (http://area51.stackexchange.com/proposals/2761/natural-langu...)
statistical analysis (http://area51.stackexchange.com/proposals/33/statistical-ana...)
data mining (http://area51.stackexchange.com/proposals/4229/data-mining)
AI (http://area51.stackexchange.com/proposals/6607/artificial-in...)
This niching is completely wrongheaded. As I wrote in a comment on this blog post, the last thing we need is for ML people, NLP practitioners, and statisticians to communicate less. The fact that these groups attend different conferences is a BUG not a feature. My site was designed so that these adjacent fields can crosspolinate information.
As we’ve learned from StackOverflow, it’s better to pick a broad topic and have a lot of experts in one place, communicating and helping with each other’s problems.