What's Wrong with Stack Overflow: Social Over Substance | Hacker News Reader

What's Wrong with Stack Overflow: Social Over Substance

stackoverflow.com

9 pointsJabavuAdams28 comments

I don't get your complaint.

You posted a trivial question you knew the answer to, and then posted your trivial solution, at which point you were told that it was preferred if people asked genuine questions they didn't know the answer to.

If anything, people knocked you for not providing enough substance-- and I don't see anything "social" about the interaction at all.

There's a lot of things wrong with StackOverflow-- but you're not offering much of a critique.

What's trivial to one is not trivial to all.

The reason I posted the question (and answer) is that I saw someone else was confused about these operators in a different Stack Overflow question.

The "social" complaint was because the first commenter down-voted me because he though I was trying to get a badge, not because he thought there was something incorrect about the information (maybe he just didn't say so explicitly).

A lot of people coming to Lisp are new, and need small little code snippets.

Although my answer may seem trivial, it had a few carefully-thought-out properties: 1) A very specific title. 2) A specific answer, with alternatives, and references to the language spec. Many Lisp newbs won't know to go to the Hyperspec.

So, maybe I just posted this in the wrong place. What is the right place for standard techniques or snippets that can improve the code-quality of a cut-and-paste beginner?

I think it's trivial to exactly 100% of people that "Ask A Question" means "Ask A Question" and not "Cut-and-paste standard techniques or snippets that can improve the code-quality of a cut-and-paste beginner".

The reason I posted the question (and answer) is that I saw someone else was confused about these operators in a different Stack Overflow question.

Then that was the right place to post your answer.

It's generally considered a good idea to lurk a little while in an online community, to see the local standards. If you had done so, you'd have seen that people don't post "standard techinques or code snippets" to StackOverflow-- they ask questions they don't know the answer to, and try to answer other's questions.

If you want to help folks on SO with Lisp, you can try to answer some of the genuine questions that have been asked.

It's generally considered a good idea to lurk a little while in an online community, to see the local standards.

True. I did check the FAQ, at least, and concluded that this was a legitimate post.

I think there's an unanswered need, here. It's a problem of search. You want the first search result in google to be a small, rock-solid code snippet.

Small, because small is simple, and small can be perfect. Longer code samples introduce too many degrees of freedom. Small samples can be unambiguously correct (for a specific purpose). This doesn't mean they're trivial.

Anyhoo. If this isn't what Stack Overflow is, someone should make it...

I agree with you. You saw a question that should have been asked but wasn't. You phased the question in a specific way to give a specific answer.

It does not make sense to bury this information in tons of other posts that would not come up in a google search when searching for this problem.

I don't get the nontrivial argument. Is everything in the documentation nontrivial? Where is the ideal place for a user looking for the answer to this question?

I think even if you thought the question should have been asked... you should have waited a day (or longer) and given someone else a chance to answer the question. Then you could have accepted their answer, or written your own if it wasn't good enough.

It does very much look like badge/rep whoring.

I can see that, but doesn't that point out an interesting flaw in this social site achievement model?

Specifically:

1) There is a question that is trivial to some, but non-trivial to others.

2) I have a carefully-crafted answer to that question (trivial though it may be to some).

3) I need to wait to provide my careful answer, because if I do it too soon, I'll be seen as self-serving.

This is inefficient!

Storage is cheap, so what does it matter to anyone if there is a concise, correct answer that they find trivial? If something is not useful to me, but useful to others, it doesn't harm me ...

... unless it adds noise to the signal, i.e. corrupts search terms.

However, I explicitly chose to improve the search terms w.r.t. this topic.

This is very interesting for me...

It's really more that it seems to be self serving, because you weren't really asking a real question. The StackExchange model isn't really a Wiki, or a blog, or whatever. If you want to write user friendly documentation there are lots of ways to outlet that. SE is about people who have real problems that are really searching for some help, IMO, or maybe the just want to learn something they didn't know about (e.g. what's your favorite editor might get answers that included editor's and reasons you'd never heard of).

I do think this adds noise to signal, because what people who answer questions are there for is to help people who need it. You didn't, and thus it seems more self serving.

I do think this adds noise to signal, because what people who answer questions are there for is to help people who need it. You didn't, and thus it seems more self serving.

This is interesting. When I said it didn't add noise, I was specifically thinking of question-askers. I didn't consider question answerers.

On the other hand, with a very specific title, people who are just browsing should be able to make a snap decision about whether to read further. Also, it takes much more effort to compose a good answer, than to quickly validate an existing answer.

Consider, that once an answer is there, it's there for all time. We don't know how many people searched for this in the past, but gave up due to poor results. We don't know how many people this will help in the future.

Why wait until they ask the question? When I was learning Lisp, I just used "(+ foo 1)" to increment a number. It was a while before I realized that I could just "(incf foo)". So, in that sense it's a genuine question on my part, just time-delayed from a past self.

Part of good teaching is anticipating what questions will be asked, and answering them before-hand.

What bugs me here is that I've created something. I've organized information -- decreased entropy in the universe. I did all that, but now other people want to bury that information that took energy to create, and that could be useful for all future time.

I'm perfectly okay with someone saying: "this is trivial to me". What I don't like is that someone can say: "this is universally trivial, for all users", let's bury it.

Corrupting search terms can be seen as adding entropy, while improving the likelihood that information will be found is a public service. Picking good search terms is key to the long-term survival of an idea.

sigh I usually only click on other peoples questions if I think I know how to answer it. That's like 70% there's another 20% which might be curiosity, and maybe 5% other reasons. (This is all not including Unix SE where I read stuff to see if it needs retagging or editing or some other form of moderation.)

So imagine if everyone did what you did and now 50% of question askers really don't need an answer, they're just trying to add yet another source of "documentation*. If the noise ratio was that high I wouldn't bother.

'Social over substance' here most likely means that people on Stackoverflow, esp the ones who have reps in thousands tend to treat simple questions (also fundamental ones) as 'not enough in substance'. Which I think is wrong. Substance is knowledge and transfer of that, not who can ask a star-studded question.

SO has a badge which says excellent question and I think those with awesome reps (who generally are ace in their fields) tend to feel bad about such 'naive' (according to them) questions.

Reminds me of that saying "no good deed goes unpunished".

As a user of SO, I always prefer to see an answer already there than asking it. Even though SO users are super fast at answering my questions (usually get answers < 1h), there is something to be said about the answer already being there.

Having said that, the "hive" at SO has already developed a set of informal rules, and there are plenty of people who seem to make a point of making meta-commentary on questions.

While this is standard fare for social sites and the interwebs in general, at least the signal to noise ratio is an order of magnitude better than most other social sites.

I don't get it. Given your self-proclaimed skillset (below) you are trying to tell us you genuinely don't know how to increment a number in Lisp?

    Expertise:

    Computer, Video, and Mobile Games (PC/Mac/Consoles/iPhone/Android)
    Real-time physics simulation
    Game-logic and character AI
    3DS Max scripts and plug-ins
    Unity game engine
    OpenGL/OpenGL ES/Cg Shaders
    C/C++/Objective-C/Java/Python/Lua/Common Lisp

You're right, I knew the answer. I wanted to create a canonical answer that others could find, since I could not find one, and since I have evidence that some other programmers don't know this.

So why not let them ask the question themselves? I understand that's part of the premise of StackOverflow.

Chicken-and-egg. As a senior dev. I often see junior devs who don't know what questions to ask. They do a quick google search, and if it doesn't come up in the first few hits, they're lost.

What is wrong with the terms google came up with.

I got this as the first term by searching how to increment a number with Common Lisp. http://www.cs.sfu.ca/CC/310/pwfong/Lisp/4/tutorial4.html

This seems like a likely search term for a "noob".

It isn't a matter of whether he genuinely doesn't know it or not. Joel Spolsky knows how to move the turtle in LOGO, but it is a valid question: http://stackoverflow.com/questions/1003841/how-do-i-move-the...

The goal of Stackoverflow is to become a wiki for programming questions. If you look in the faq, it clearly states:

> It's also perfectly fine to ask and answer your own question, as long as you pretend you're on Jeopardy: phrase it in the form of a question.

I get that you can "legally" ask your own questions, or edit other people's posts, etc but there has to be some common sense. I think it's disingenuous to post a question on how to increment a number (in Lisp, which is not something that total newbies fall into in the first place), have the question tagged as badge-bait, and then post a rather inflammatory headline "What's Wrong with Stack Overflow: Social Over Substance".

Well that was not my intent.

Lisp, which happens to be one of the languages that total newbies don't fall into in the first place

Wrong! They hear from their cool but slightly wacky friend that Lisp is the new old hot sh*t. Or, maybe they read Paul Graham, or Joel Spolsky.

They go to try it ...

I've answered my own questions many times on SO site's... but I don't think that you should ask a question you know the answer to when you ask.

I do not think SO is a wiki for programming question. I feel it's a place for people to go for help when they have a problem.

Given I occasionally ask questions I could have looked up, so it helps other people, and occasionally I end up answering my own question (by looking it up), after I've given others a chance to provide an answer.

It just feels like abusing the system by asking a question you know the answer to. That will stay unanswered for 2 days (since that's how long it takes before you can accept your own answer). At the very least he should have given other people a chance to answer, instead of rep whoring.

(unix.stackexchange mod here)

See, this is why I chose the HN title.

As soon as a site adds rep, there come the accusations of rep whoring. I don't care about my SO rep, as long as it's not libellous or something. I don't need a job, I don't want a job. I want (in this specific case) to teach.

What if people don't know they have a problem?

If SO becomes a standard destination (and it appears to be from watching the devs around me), then what is not on SO / accessible to Google might as well not exist.

Re: the letting others answer. Let them edit my post instead, if they think there's a problem with it.

Idk about you but my blogs return quite quickly on google if you search for something I wrote about. This seems like a more appropriate outlet for this kind of thing. Of course if the dev's around you have forgotten how to search... well I doubt they are worth what they're paid.

You need enough rep to edit a post, but really the question isn't the problem. It's the fact that you didn't actually need an answer.

I'd much rather have a SO question as a search result than a blog post.

Blogs are often littered with other information, have weird color schemes, don't always have proper code display, aren't editable by somebody who finds an error, sometimes have spam etc etc

I'd much rather have excellent documentation, and less crappy defaults so I don't have to ask so many questions, or do as much search, or blog as many "how to's", but we can't all get our wishes.

I think it is a wiki for programming questions, and Joel and Jeff repeatedly said so on the SO podcast. The fact that you think it is somehow "unfair" to get reputation that easy illustrates Jabavus point: SO has become less about being information source and more about being a game.

Not only is it OK to answer your own question on SO, there's a badge you can only get by doing so: "Self-Learner: Answered your own question with score of 3 or more."

Here's the tip I shared to get my Self-Learner badge:

http://stackoverflow.com/questions/1107551/css-strikethrough...

Reply on news.ycombinator.com