I am excited to see how this works as social software.
I hate "karma" systems in discussion forums. As soon as you introduce karma points, your forum is now dual purposed, part for the original intent (say, discussing Hacker News) and part for playing the karma game.
I accept that karma points are at least a known working solution to spammers and trolls. But if it wasn't for that, I'd much rather read a conversation without the built in popularity contest.
At stackoverflow, they've turned that all on its head. When I read the reputation rules (see the unofficial FAQ) I thought, "this is all karma game". And then I thought a little more and realized that for the purpose of the site, that might work perfectly. If the guy with the question is the one who hands out the karma, the incentives of the guy and the karma gamer are nicely aligned. Very clever.
I saw someone talking about the closed beta, and he described it as "fun". Well, that sounded strange, but now I believe it. Jeff and Joel turned karma gaming into an actual game, with the incidental effect of generating good answers to technical questions.
Already annoyed by it and all it took was some random clicking around: http://is.gd/2EPh
Apparently the "right" answer to this language-agnostic question is one that provides a single Ruby (surprise) alternative. Since when do opinions have correct answers?
I would love to see great things from Stack Overflow, but I would be surprised if it turns out to provide better answers than can be found by googling and reading mailing list archives -- or at least better answers that are consistently marked as such (or not marked at all, depending on the situation). Introducing voting doesn't magically mean the "good" stuff floats to the top.
Edit: Oh, and people are already "trolling for reputation" as evidenced by a single question having 10 of the same exact answer. I thought this is what voting was going to "solve?" sigh
Usually the duplicate answers happen in unison. If it's an easy question, there are usually around 5 answers posted simultaneously. If this happens, you're usually supposed to delete your answer if you see a better one, but as you can imagine there are different criteria used to determine which answer is better. Ultimately this is solved by the voting and answer selection by the asker.
I'm not sure I'll be a regular user of Stack Overflow, but I do hope they replace Experts Exchange as the top search hits for all my random programming questions.
I don't find it terrible, but I think the question text on the main page could be larger than the Votes/Answers information. I mean, isn't the question more important the number of votes or answers? I'm not scanning the home page for the question with the most votes or answers.
What kind of effect will making the number of votes, answers, and views more prominent than the question have on the site? Will it encourage gaming the system / posting for points rather than to be helpful?
Interesting classification. But I wouldn't put the answer/view count in crap. Otherwise there is no way of judging/weighing the answers and its usefulness.
Those numbers tell you nothing about the actual usefulness of the answer since that's entirely subjective. They only tell you a little about the community there, which is only useful if that matters to you (i.e if you're there to play the game).
However from the point of view of someone who admittedly wastes too much time on distractions like YC, I'm glad their design sucks since if they copied YC outright they would be 4x as likely to suck me in. In the same screen space YC manages to show me 20 articles whereas Stack Overflow only manages to fit in 5.
In my opinion, the colors don't look good together, and the main page is way too busy. Tags have too much emphasis. (Reddit does the same thing, and it's ugly there too.)
Anyway, I am not the target demographic, so my opinion really doesn't matter.
I was amused by the contrast between (on the one hand) Joel's long list of instructions and "how-to" hints in his blog post, and (on the other hand) the seeming absence of instructions/guidance on the site itself. (The guidance is actually there if you dig around far enough, but it sure doesn't look that way to a casual user.)
I also find it interesting that you apparently need 15 reputation points before you can upvote, which means you have to successfully answer one or more questions. In practice, this means that a drive-by/first-time user can't use their expertise to choose the best answer among the existing set, but have to chime in with a (most likely redundant) answer first, which seems counter-productive.
The threshold is to prevent bots from upvoting stuff. It requires a minimal amount of participation that can easily be achieved by a simple answer or question.
This is a bit of a meta comment, but I find this sort of response to feedback to be counterproductive. Don't defend/justify the way the system works or why it works the way it does. Read between the lines and find a way to satisfy the customer.
Don't defend/justify the way the system works or why it works the way it does. Read between the lines and find a way to satisfy the customer.
That might be good advice if the customer contacts you one-on-one. But when the customer posts their complaint to a message board that's read by thousands of third parties, it's helpful to put up an explanation of why the system was designed the way it was. Otherwise I, who have never used the service and am just skimming this thread, might come away with the wrong impression.
It's important to avoid getting into a flamewar with your customers, but pointing out the rationale behind your design decisions isn't necessarily offensive. Yes, it's not what the complainer really wants to see, but it is what I, the silent reader of the thread, want to see. And people like me outnumber the original complainer N to 1.
You might outnumber them, but you may be less vital. A person that takes the time to write is a potential evangelist.
And you'd better be pretty confident about 'N'. Threads like this and readers like yourself may provide a short term boost or bust, but communicative visitors can broadcast the message off-line for months to come.
I think you'd agree that we should aim higher (much, much higher), than just avoiding getting into a flame war.
I am in no way affiliated with Stack Overflow - I've just been a user for the past few weeks so please don't lecture me about satisfying the "customer". During my time on the site I've seen the creators attempt to walk the delicate line between ease of access and attempting to curb the abuse that comes hand-in-hand with that. Personally, I think that they've made an incredibly open system that anyone can contribute to if they want.
You may have been confused about the scoring system using to calculate reputation. It's a little more fine-grained than a traditional up/down karma system. Since voting up gives you +10 rep, getting 15 rep is as easy as getting two people to vote up your submission.
1) The site builder added a certain effort treshhold to block out spammers.
2) The complaining user doesn't want to invest the effort to cross the treshold.
These statements are mutually exclusive. The only way to resolve this problem is to remove the need for a treshold. (i.e. remove spam from the internet)
The question I meant to ask was: "what would you do if there isn't a solution to the customers problem?" + explanation why this particular question doesn't have a solution.
Customer: "I would like it better if every time I opened up your web page a slice of buttered toast came out of my CD drive."
Response: "[APPRECIATION: Thanks for your feedback]. [EMPATHY: We understand your desire for a tasty breakfast], and [PLAUSIBLE CONNECTION: we see how a website about coffee should help you plan a more complete meal]. Unfortunately, I'm not aware of a way to do this right now. Just for future reference, [OPPORTUNITY FOR MORE INVOLVEMENT: if we were able to do this, would you like to help us test, and would you be opposed to margarine]? For now, what if [SOMETHING CLOSE TO A SOLUTION: we began building a complementary breads chart, like [example]]?
Agreed; that's why I didn't get very far with the site (and I was a member during the beta period). It seems that people answer questions very quickly and it feels like you're in a contest for the "points" - who wants to compete against others for mere points on a single site? Not me - so I stopped going there.
That said, I think it will be excellent for those asking the questions, but for those answering..? It'll just be the prima donnas and gold-hoarders playing the game.
I was amused by the contrast between (on the one hand) Joel's long list of instructions and "how-to" hints in his blog post, and (on the other hand) the seeming absence of instructions/guidance on the site itself.
The way I see it, the site is right now (mostly) good for two types of questions:
Really basic programming questions that could be answered by reading some docs.
Really domain-specific questions that are hard to find info on online. If you're working on something really interesting, but don't know enough about the domain to know where you should be looking or what you should be searching for, you can get really good guidance from the people on stack overflow who know what they're talking about.
It's also good for getting an idea of consensus about certain technologies.
Ugh... I know it's a site for programmers, but would it have broken the budget to hire a designer? I can't tell what anything is at first glance and it all looks jumbled to me.
Are you sure about that? I haven't asked a programming question (in public) for a number of years. I am too lazy to sit and wait for someone to tell me how to solve my problem when I could discover the solution on my own.
I assume that most people here are like that. (Otherwise they would be on Reddit instead ;)
It's also perfectly fine to ask and answer your own programming question, but pretend you're on Jeopardy: phrase it in the form of a question. </quote> http://stackoverflow.com/faq
If (if) you don't want to share you knowledge for the benefit of fellow programmers then that's another question.
True, my comment was a bit quick off the block. I didn't mean to imply that the average HN user and the average web-forum-programmer were anything similar.
Except, perhaps, aesthetically. Maybe I'm wrong here too, but a failed visual design for HN readers is probably a failed design for the average web-forum-programmer, too.
No, I meant lazy. Sitting around doing nothing would require a lot of mental effort and work. Thinking about programming would be relaxing and enjoyable.
That's a good point. I feel the same way, I just assumed everyone else enjoyed sitting around and doing nothing and there was something wrong for me. Glad I am not unusual. I have a lot of non-programming tasks lately, yet I find myself finding a small problem to hack around as "procrastination".
Yup. The reason for this is that the site usage has been ramped up with private invites etc. The makers are well-known.net pundits, so the people that have been following thier activity and signing up are too.
Now they're ready to let the general public in, so if you want your favourite environment/tool/language represented, get on in there. If you don't want to get in there, stop complaining.
Well, all of the founds are pretty heavy MS based; same with their audiences. I believe that joelonsoftware's .NET questions group was going to shutdown and forward all the users to here too.
I did find some decent Non-Windows questions in there, but there could be more.
Sure, my search results for answers to programming questions have turned up dubious results from time to time, but usually my answer is on the first page. Occasionally I'll have to refine my query and find the answer on my second attempt. With that in mind, I'm not sure how much better this new site is going to be.
The founders intend that eventually the answer you're looking for will show up as a stackoverflow.com question in the first page of Google results. They get more advertising revenue, the purported benefits to their users are:
* You don't have to pay or even register to use the site (unlike Experts Exchange).
* If an answer goes out of date (e.g. it only applies to the beta version) it can be updated.
* You can always add your own comments (not all blogs allow comments).
* If you manage to solve some tricky problem yourself, you can easily publish the information to help others (not everyone has their own blog).
Add .NET to the list of languages whose libraries don't work with my OpenID. (So far, only the Perl libraries seem to support delegation. Considering how widely-documented delegation is, this really surprises me. It's a conspiracy, I tell ya.)
Very interesting. I ( http://jrock.us ) seem to have an X-XRDS-Location http-equiv header, whereas you don't. I'm not sure why extra information would prevent it from working with some sites (and not others).
I had a quick play with the design using Greasemonkey, and I think it's much more readable by increasing the page-width and standardising the font-sizes. The script is a little slow but shows that with only a bit of work the site's design could be improved dramatically. I'd have liked to expand the banner at the top, but sadly it's a table so a pain to work with.
Somehow tags don't seem to be the best way to organize this type of question asking and information gathering. They're certainly helpful for searching and gathering semantic data, but I'd find it much more useful to see the site have "sections" for different programming languages a la C++, Perl, PHP, Python, Ruby, etc.
In this specific case, the tagging works out a lot like "sections" anyhow - I do believe that most of the questions are language-specific, and that if you go browse the tags, most of the ones with a lot of questions are language tags.
The tags-as-sections thing breaks down when you get a little more abstract than "x language", as the tags used will be ambiguous (people can retag stuff though, so that's something of a moderation system).
The difficulty comes in having to browse through the hundred popular tags to find one specific language tag where you want to ask a question or provide help to people.
Granted, it's not that hard to do, but it's not optimal from a usability perspective.
For instance, PHP wasn't listed in the popular tags even though it's one of the top ten most popular programming languages (http://www.langpop.com/). To find this tag, I had to click Tags, search for PHP, and click PHP.
That is true, although for asking questions you can just tag it as the language.
Php probably wasn't listed in the popular tags due partially to the windows/.NET tendency of the site, and the tendency of the people participating in the beta to be the type that doesn't like php.
I like how you took an idea that was semi-out there, molded it, made it free, and tossed it out there. genius. I think the best thing you guys did with stack overflow was to give the community so much potential power. bravo.
Voting is good. for most part, people are just to lazy to give a comment(me,too!). So, web 2.0 sites developed such voting and poking things to get us out of typing.
I hate "karma" systems in discussion forums. As soon as you introduce karma points, your forum is now dual purposed, part for the original intent (say, discussing Hacker News) and part for playing the karma game.
I accept that karma points are at least a known working solution to spammers and trolls. But if it wasn't for that, I'd much rather read a conversation without the built in popularity contest.
At stackoverflow, they've turned that all on its head. When I read the reputation rules (see the unofficial FAQ) I thought, "this is all karma game". And then I thought a little more and realized that for the purpose of the site, that might work perfectly. If the guy with the question is the one who hands out the karma, the incentives of the guy and the karma gamer are nicely aligned. Very clever.
I saw someone talking about the closed beta, and he described it as "fun". Well, that sounded strange, but now I believe it. Jeff and Joel turned karma gaming into an actual game, with the incidental effect of generating good answers to technical questions.