Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

What does Stack Overflow want to be when it grows up?

blog.codinghorror.com

Post reply on HN
Print view View original

61–70 of 216 posts

Re: What does Stack Overflow want to be when it grows up?

#61

Post by tokyodude »

Flimm wrote: post #42
gonational wrote: post #20

I’m gonna catch some flak for this... Listening to Jeff try to describe what stack overflow is and the “fundamental misunderstanding” by all of us millions of unwashed masses is amazing. It’s also a bit like listening to a muskox describe why it decided to grow long hair after so much time in the snow; IOW, a complete and total lack of understanding of emergence and a rewriting of history at the behest of Jeff’s enor…

I don't think you are giving Stack Overflow's leaders enough credit. There were many Q&A sites before it, but Stack Overflow was much better for the lurkers, and all of that content is licensed under Creative Commons, similar to Wikipedia. Also, Jeff is not trying to take credit for the users' hard work, I'm a contributor and I don't feel that way, especially after reading this part: > I am honored and humbled by the…

> and all of that content is licensed under Creative Commons

it's CA-BY-SA IIRC. In other words if you copied and pasted a snippet from SO you now need to open source your project (the SA part of CC-BY-SA). I'm surprised someone hasn't written some ransomware that scans SO for snippets and compares them to github and starts sending out the "pay up" notices.

Re: What does Stack Overflow want to be when it grows up?

#62

Post by irrational »

jrochkind1 wrote: post #26

One challenge not discussed here is when an accepted and highly voted answer _used_ to be a good answer, but no longer is in later versions of whatever software/platform/library is discussed. This is also something that gets worse over time, like the uniqueness problem. I think it's worse than the uniqueness problem, multiple duplicate _good_ questions/answers don't actually harm, misguide, or delay anyone looking fo…

Recently I have really started noticing this exact problem. Every answer (usually from closer to 2010) I have looked at is recommending using jQuery, even though there are perfectly fine ways of doing it without jQuery in 2018. I've taken to looking the dates of answers and ignoring any that date from more than 2-3 years in the past.

Re: What does Stack Overflow want to be when it grows up?

#63

Post by CodeCube »

gonational wrote: post #20

I’m gonna catch some flak for this... Listening to Jeff try to describe what stack overflow is and the “fundamental misunderstanding” by all of us millions of unwashed masses is amazing. It’s also a bit like listening to a muskox describe why it decided to grow long hair after so much time in the snow; IOW, a complete and total lack of understanding of emergence and a rewriting of history at the behest of Jeff’s enor…

> Stack overflow is and always was a question and answer site. To attempt to take credit for the following value of search results containing stack overflow questions that have already been answered that sound like what you’re asking is to take credit for search engines and to take credit for your users’ hard work.

I don't know man ... while I can kind of agree with the general "thing" you're trying to convey here (that of SO hard-headedly blaming users for "doing it wrong"), this particular point is one that I'll have to defend him/them on -- SERP-as-an-interface was a primary consideration from day one. I listened to all the podcasts they did during development, and they were pretty explicit about it in all the blog posts. They knew that most people would arrive at a previously asked question via search engine, and they optimized the experience for that reality. That's why they wanted each question and answer page to be an "artifact", something that grows over time, gets edited, improved ... precisely because they thought of this ahead of time ;)

Re: What does Stack Overflow want to be when it grows up?

#64

Post by Insanity »

skohan wrote: post #46

I feel like Stack Overflow has gotten less useful to me as a programmer over the years, and these days I more often find the answers I need on a forum geared toward whatever technology I have a question about. I have a hard time telling whether this reflects a change in developer culture online over the years, or a change in the types of questions I tend to ask as I have become a more experienced programmer.

I feel the same way.

Questions I have are more specialized to the domain I'm working in rather than general programming questions.

No longer of the kind "X in language Y". Most of the questions about the system that I'd have nowadays can also be answered by colleagues over slack with the domain knowledge.

Re: What does Stack Overflow want to be when it grows up?

#65

Post by grey-area »

k__ wrote: post #5

When I wrote my bachelor thesis about democratic methods in online communities, I found out that all these crowd-sourced sites had one major problem. When they start, senior players in their field of work laugh at them because they don't have much content. So they try to get as much as possible, trying to get as many new creators on board as they can. Later, when they have grown enough, the senior players in the fiel…

This is true, there is definitely a lifecycle to these sites (which even perhaps applies to forums too), it can be seen in all large online communities (and perhaps all communities?).

There's another even more fundamental problem though which emerges at every stage. The people who want to participate the most at any stage are not the most qualified, and the people who are qualified to answer are busy doing instead of answering. It's not just that inevitably you need to kick out bad content, it's that the people doing the kicking out don't know enough to judge what is good and what is not.

So by the time you get to the recruiting moderator stage most of the people interested in becoming gatekeepers/moderators are not at all the people you actually want to be gatekeepers (those people are generally too busy or not very good at such things), and your moderators are a self-selecting group who are more interested in form and adherence to rules than content.

Re: What does Stack Overflow want to be when it grows up?

#66

Post by ravenstine »

irrational wrote: post #62
jrochkind1 wrote: post #26

One challenge not discussed here is when an accepted and highly voted answer _used_ to be a good answer, but no longer is in later versions of whatever software/platform/library is discussed. This is also something that gets worse over time, like the uniqueness problem. I think it's worse than the uniqueness problem, multiple duplicate _good_ questions/answers don't actually harm, misguide, or delay anyone looking fo…

Recently I have really started noticing this exact problem. Every answer (usually from closer to 2010) I have looked at is recommending using jQuery, even though there are perfectly fine ways of doing it without jQuery in 2018. I've taken to looking the dates of answers and ignoring any that date from more than 2-3 years in the past.

LOL Half my SO searches include "-jquery".

Re: What does Stack Overflow want to be when it grows up?

#67

Post by coldtea »

kauffj wrote: post #57

StackOverflow being a wiki is at substantial odds with the way answers are selected. If StackOverflow is a wiki, then the best answer to a question should be determined by the community, not the asker, and needs to be able to evolve and change over time. However, StackOverflow simultaneously insists questions can't be duplicated while making answer selection immutable and controlled by the asker. This results in a su…

>If StackOverflow is a wiki, then the best answer to a question should be determined by the community, not the asker.

You're thinking of wikipedia. A wiki doesn't determine what the "best content" in it is. It's just a collaborative way to edit text together.

Re: What does Stack Overflow want to be when it grows up?

#68

Post by EpicEng »

sureaboutthis wrote: post #27
PudgePacket wrote: post #4

I've used stackoverflow for years and have never heard it put like that. "Stack Overflow ultimately has much more in common with Wikipedia than a discussion forum. By this I mean questions and answers on Stack Overflow are not primarily judged by their usefulness to a specific individual, but by how many other programmers that question or answer can potentially help over time ." Personally I would have been much more…

A version of that is published in the Help Center and The Tour--both which everyone is directed to but nobody reads--and then everyone argues with us about.

If no one reads it then perhaps they need a better mechanism to explain this stuff.

Re: What does Stack Overflow want to be when it grows up?

#69

Post by umvi »

cdoxsey wrote: post #35
umvi wrote: post #10

Earlier quoted context omitted.

The only questions I've seen constantly outdated are web dev related. More stable technologies like C/C++/python seem to have up to date answers most of the time (purely anecdotal).

Lately I've been working with c++ and, at least in my experience, as a developer who's native tongue isn't c++, many of the answers have failed to keep up with the radical changes in the language in the last 10 years. Post-17 C++ is completely different from pre-11 C++. That said there's usually an answer further down the page with a more up to date solution, it's just not the accepted or most popular one.

True, but consider your user base. I, for one, maintain legacy C/C++98 code. C++17 answers are not helpful to me because the legacy system does not have a new enough compiler, sysroot, etc for me to take advantage of them. It's not like web development where users' browsers auto-update and you can start immediately using the newest versions of everything. Only our newest projects are using modern C++ (14).

I would wager the majority of C++ users "in the field" are using 11 and below and haven't had a good opportunity to switch to 14 or 17 yet. As such, answers that seem outdated to you may be a good fit for the majority. It's sort of like how any native answers automatically float higher than boost answers - why is this? Because most c++ users are not using boost and have no interest in using boost.

Re: What does Stack Overflow want to be when it grows up?

#70

Post by badosu »

0x445442 wrote: post #39

I think the need for, or at least the level of reliance on, a site like Stack Overflow is an example of how utterly messed up the "profession" is compared to other engineering and non-engineering professions.

That's not a fair comparison, the level of diversity in ways to solve or diagnose a problem are much bigger in the software engineering world.

Where in a typical engineering profession you would have a somewhat defined checklist on how to diagnose an issue and a proven method to fix it, in software engineering you have a million possibilities: "- Did you check x? - Was this nth specific step for this specific framework followed? etc.."

It's messy because the problem space can't be as easily enumerated.

Post reply on HN