StackOverflow unwinding?

As an aside, I have to say spolsky that SO has made a real, measurable impact on my life. The number of times I type my query site:stackoverflow.com into google is higher than any other site.

Humans are inherently social, and the amount of effort SO staff and moderators have to go to trying to stamp that out is a sign that SO really is a social network.

I always think it's a bad sign when product designers think they know much better than their users what they want. I think that's what killed Friendster. I think that's why Google Plus is kinda sad and sterile. If you're going to go to all the trouble of making a product for people, I'd rather you listened to them from time to time.

If it means "overall, people lose interest and stop visiting to ask and answer questions," I don't see that. I'm not visiting to fill in all the gaps in SO's content, and feeling disappointed when there aren't many gaps to fill. I'm running into questions, Googling, and hoping to find my question already answered.

If I find it and it's not answered, I'll do research and try to answer it myself. If it doesn't exist, I'll ask. If I find the answer before anyone else does, I may answer it myself.

I'm not going to stop doing that, and I don't see why anyone else will, either. What the overall trends are in the site really has no bearing on my individual participation, so it never feels "stagnant" to me.

Maybe you're imagining users who show up and surf for questions to answer, and those people are getting bored? That's just not the way I use the site, though.

The hard part in my experience when going to try and answer some questions is it seems like there are more and more questions by inexperienced coders expecting people to write basic code for them, or not formulating the question in a way that will be useful to others in the future - instead they're pasting in their code and asking, "why isn't this working". Instead of, "I would expect this framework to behave as follows but it's not" it's "Here's 50 lines of code, why isn't it working?" The only way to answer such questions is to get their code up and running somewhere, perhaps on a jsfiddle, figure out what's wrong (often something as trivial as a syntax error), and then tell them. Doesn't feel very gratifying to do one off debugging on behalf of the person asking especially if the answer is unlikely to be useful to others in the future.

SO is pretty rigid about what they want. That's a good thing - their servers, their rules. Firm guidelines allow users to find either post useful questions, or find some other site.

Except, that's not what happens. Someone asks a question that's a lousy fit for the SE format; and people like it, and upvote it and answer it and the answers get upvotes. Then some mod comes along and closes it, and there's a bunch of negative feeling.

SO is huge, and needs constant tending to keep it useful. Cool urls don't change, but when Mark Pilgrim (who has a gajillion urls over SE) leaves the Internet there's a gajillion urls that need to be found and changed. (No, I have no idea why a site run by programmers, for programmers, can't kludge together a script to search for and replace urls.) That's one small example. There are plenty of other examples.

Then there's the "problem" of duplicate answers. Ideally, there would be one great question, with a bunch of answers, with the best answer being upvoted most and the answer most fitting being accepted. But very few people searches before asking, and all the tags are not useful, so there are a lot of duplicates. That could be okay, except the dupes are usually worse quality than one great question. And the answers lose quality over time - there's only so many times that someone is going to write a great answer.

Then the gamification is broken on some of the sites. You start off almost unable to make any contribution, and people with low scores do not get the upvotes. (This guy with a million votes must know what he's talking about, UPVOTE.) In theory you could concentrate on an obscure segment. In practice not much on SO is obscure because there are so many users.

The fragmentation is a bigger problem, I think. I want to ask a question about linux software, on my Ubuntu install. Already there are two different SE's I could ask on. There could be more depending on the software. This is a more vicious form of the duplicates problem.

The proliferation of SEs into ever more niche segments is okay, I think. They either sink or swim. Lousy for you if you've invested time and effort into something that sinks.

There is definitely a niche for a subjective version though. All those questions that peope love to ask, and answer. ("What's your favorite monspace font?" "What are your 3 most important programming pearls of wisdom?" "What book did you find most useful when you started, and what is most useful today?" etc etc. Obviously, versions of these questions exist in reddit and /prog/ and news:comp.* and etc etc.

FWIW I love the SE sites, and I love the fierce moderation. I've disagreed with it at times, and certainly with the cold way it gets implemented. But I'm glad SE exists. I just wish I knew a where I could ask different questions and have them seen by a huge crowd of clueful users.