Five Years of Stackoverflow.com
stum.de
stum.de
In a way I suppose that's a testament to the fact that the site still works, and works well. They solved most of the problem they set out to solve originally, which is a major achievement. Nevertheless, I still have a nagging feeling that some further iterative improvements could turn stackoverflow into something that would be even more useful, and would be more inviting for people to continue to participate in actively even without having expert knowledge in some tiny sub-category or a desire to constantly tidy everything up day-in day-out.
Edit: to clarify I mean that it's hard to visit stackoverflow on a regular basis instead of just using it as an answer generating machine. Which I think is something not unique to myself and probably hurts the overall quality of the site considerably, notwithstanding the many experts who use the site heavily. Specifically I find it hard to peruse interesting topics to comment on, since the site is now very aggressively tightly focused. Even worse I find it difficult to use the site as a learning tool. Often times when I want to learn something new I'll get it in my head to go to stackoverflow first, but that's never worked very well for me. It's just too difficult to sort or filter all the data into something that is actually useful.
Overall I think the biggest problem is that the people behind the site(s) have decided they've done their jobs, that what they've presented is good enough and at most needs only minor tweaking. I think that's a huge mistake personally and I think the result is going to be that the stackexchange sites end up as a footnote in history after having been rapidly eclipsed by "the next big thing" in a similar vein.
I can understand why it appears this way from outside, but we've never been moving faster than we are now.
I have been on board for about six months, and in that time we have rolled out lots of usability and layout improvements: streamlined sign-up process, a new anonymous homepage, a consolidated and curated help center, completely rewritten search, new badges page, upcoming new privileges page...
In the next couple of months, we will be finishing sitewide SSL, redesigns of our more high-traffic pages, and a complete re-factor around closing to make that process more clear.
Stay tuned, because I'm excited about where we're headed.
You already rely on the users to vote up the good answers and vote down the bad ones - why don't you trust them to continue doing this, rather than shutting down useful and interesting topics?
http://meta.stackoverflow.com/questions/171732/help-us-make-...
I think what GP was commenting on, which I agree with, is that SO doesn't appear to evolve its core mission from question-answers to something else -- perhaps a worthy HN/Slashot/Reddit replacement with discussions? Or expert-based code-reviews? Or documenting best-of-breed frameworks/platforms/service providers/etc/etc/etc (like GP posted)? I don't necessarily believe that SO should change its mandate, and I certainly don't know what would be a worthy change, but without an appearance of change then I can't help but fear SO will be relegated to the top google search result for me instead of a regular destination every day (like HN is).
They've achieved a lot, but if you look at that diagram of what stackoverflow is supposed to be you'll see it's still not that. It's very much not wiki-like, and it's very much not blog-like. And part of the reason to support those aspects was to provide precisely the experiences that I've found lacking, such as being able to use the site as a general education resource (like wikipedia).
How long till it follows expert sexchange.com's model?
And did you really think the ads on the site and the http://careers.stackoverflow.com/ job listings are free?
Some of the other sites that SO spawned are useful too e.g. I like reading the English language stack exchange. There is definitely room for both Quora and the Stack Exchange sites on the Internet (Quora more subjective and discussion based, whereas Stack Exchange sites I use for more definitive / right or wrong answers).
It is not designed, and will never be a place to foster discussions, or encourage exploration. I'm NOT complaining about SO, this is just how it is designed.
This kind of content, for example, is not something SO would create:
http://www.techempower.com/blog/2013/05/17/frameworks-round-...
https://github.com/skx/static-site-generators
Since we don't have a SO for discussions we still rely on information from individual research and/or the modern-day newsgroup (reddit, forums). I wish there was a place to fully explore nebulous unanswerable questions like "Which web UI framework is fastest to develop in?" or "What is the best way to start a 2d platform game today (in 2013)?"
This isn't an easy problem to solve, the questions above aren't really there to get an answer but to start a discussion. For both questions I want to know what is popular (TODAY, not in 2006), active, has been tested, and personal experience. I want to see a collection of people submitting their own takes until a top-five is produced where further study can be made.
http://www.slant.co/topics/217/~best-css-preprocessor
http://www.slant.co/topics/95/~what-is-the-best-search-engin...
What I mean by that is, rehashing previous discussions is verboten, shooting the shit rarely happens, and nebulous questions are never explored (even though some of the better questions linked to on SO are often closed topics!).
As a result, StackOverflow is an authoritative source on a great many things. But the atmosphere of "ask any question" is absent. You're supposed to know what to look for instead of honestly throwing your query into the void trusting that someone will catch you.
In other words, your question has to be novel in a specific sense. Which is great!
But it doesn't mean it's an encouraging environment.
Edit: I mean, witness this comment, posted at the same time as me:
https://news.ycombinator.com/item?id=5919768
I dunno, I have never felt motivated to "hang out" at SO.
I would like to see this happen. Not to harm Google, but rather to make search a problem that doesn't require billions to tackle.
A threat can only come if the searching audience moves elsewhere to search for [something]; and if [something] is very popular among the searching audience; and if [something] is also very important to individual searcher.
Because the reason for moving away can only be if [something] is not available from Google there aren't that many threats. I think last time there was a glimpse of such thing happening was in the summer of 2006, but there are other opinions.
I think the big problem I have with quickly-closed questions is that they get just as thoroughly indexed on Google as answered questions. This becomes a particular problem with niche topics; last summer, I was tasked with some Salesforce development and the number of unanswered, closed APEX questions I found was astonishing.
[1] http://stackoverflow.com/questions/17237166/sql-query-query-...
Also, closed questions can be edited. That's the main point of closing a question: to send a signal to the OP that the question needs to be edited and improved.
If the OP had provided a snippet of SQL or suggested a theory on how this might be solved, even if it was completely wrong, then the question would have remained open, gotten a few answers and perhaps even upvoted.
(not that there aren't asshole mods, but that's beside the point)
Anyway, hooray for SO (and hooray for Wikipedia), pretty much the rest of the web is composed of anything-goes sites and I love that they actively try to craft a particular experience.
Some people don't like that, and they kind of have a point, but Stack Overflow has a mission and they stick to it.
I love it for the fact I very often find answers there, but almost every time I've asked a question, the responses or lack thereof have been pretty brutal.
But there are those nagging questions where you pondered two days and researched and then pour everything you know into a long, elaborate question, asking for help. And then, nothing. 16 views, no votes, no answers. Can be very frustrating.
If you have mostly questions of the latter kind then there's a bit of luck involved whether the right person sees your question or whether it just goes unnoticed and dies.
That's exactly what I'm talking about. Even worse when a moderator comes along and closes it, as if it would have cost him something to just leave it open in case some kind soul came along and responded.
In such cases, a complete lack of attention is almost better than getting the responses that incorrectly suggest that the wrong question is being asked, or a moderator rewriting the question to ask something completely irrelevant, or comments that give blatantly wrong information.
For any moderately complex question regarding open source software, for instance, you're generally better off asking on the project's mailing list or IRC channel, or even contacting the main developer(s) directly.
When dealing with commercial software, you're often better off using whatever support the provider offers, even if it comes at a monetary cost. Some of the more established online communities are also helpful here, especially for older products.
Unless the Stack Overflow question/answer page shows up in a search engine's search results, I've not found Stack Overflow to be worth the time and effort to use it.
The site has some flaws of course, as others have pointed out, but it's leaps and bounds above every other Q&A site in their segment.
I'm interested in fragmentation. Reddit has it, and Stack Overflow has it too. I have no idea how they'll fix it.
If you want to ask a question you need to ask the right SE site. Some are obvious, but others have overlapping sites and it's confusing for the target audience.
What's worse is that some of these sites have made agreements with each other to funnel the "wrong questions" to the right site - that's great, but a bit confusing; while others have just decided independently that some particular style of question is "off topic" and they'll shunt you off to another site, and that other site will close your question.
I love SE, but I understand the frustration that some people have with it.
They have the Main sites, and the Meta sites, and the chat stuff. I think they could also have a branch off "discussion site". Any questions deemed as wrong questions go to that discussion site. You re-brand it, so as not to dilute the strict question only format of the main sites.
And, for all the "META IS DEATH" (true) sentiment there's a heck of a lot of meta going on over there.
But I do like SE.
We see this at Stack Overflow, we see it at reddit, and even at Wikipedia to some extent.
A relatively small number of users end up as moderators of one form or another, and can exert disproportionately more control over other, non-moderator users.
While this may not necessarily be a bad thing in terms of maintaining a certain minimum standard of quality, it does take responsible moderators for this to happen.
More often than not, we see people with very little power or influence in the real world using their moderation abilities on some online forum to try to compensate. They use their power just for the sake of using it, and perhaps for the emotional high it may bring them, rather than out of a sense of bettering the community.
I think that this kind of behavior is far more harmful than the fragmentation itself.
And then there is the other problem where edit reviewers have no clue of what they are rejecting and instead alienating the person who suggested the edit (rightfully). But that's a completely different point, admittedly.
Still, I don't think less-than-six-character edits would be a particular problem to handle as the edit review queue is often enough empty or very short.
But agreed, some limitations are a bit stupid and annoying at times. E.g. you downvote an answer based on the belief that it is wrong, further discussion reveals that it is in fact correct, but since more than five minutes have passed you cannot remove the downvote or turn it into an upvote unless the question is edited again. Which leaves to bogus edits just for being able to change a vote again.
Or award the badge only for substantial edits while still allowing fixing typos (and sometimes in code a single-character diff can make a huge difference).
its great when learning a new api or whatever, but i've had 4 out of 5 questions (in 4 years i need to ask 5 questions) closed as off topic, i tagged them with the tags from their own site also. example: config file management question in linux, automation, automated-install closed as off topic.
it's just a big karma jerk-off