24

Well I just saw a second Meta question expressing strong ill-will for this site after first witnessing this, seemingly well-trodden, battlefield. Given that I am already weary of it, I can only assume that others are much more so, but bear with me.

The problem

Breaking this down into as simple a a description as possible:

A substantial number of users are arriving here with a different/false understanding of what this site is for; this results in a lot of closed questions and consequent animosity.

There are two parts to this:

  1. People wanting this site to be something it's not; i.e disagreeing with / not understanding the site's objective
  2. People disagree with the site' enforcement; i.e. thinking that valid questions are being downvoted.

I would like to discount the second option as a valid grounds for debate since it seems to have been quite categorically answered by MichaelT's answer to a previous Meta question. Even if there is the odd question that is within scope that is getting voted down, this is a smaller problem compared to item 1.

There have been comments about poor communication when down-voting and some quite bitter words about specific users; I don't care about any of that because I think that it is all a symptom, not the cause of this phenomenon.

How can we better communicate the site's purpose?

Below are all current ideas as to what could be done. If you have any other ideas either edit this question or post them in an answer.

CN | Change Name

Self-explanatory, Pogrammers.SE become OtherName.SE.

ST | Add Sub-title

Add a sub-title to the website's header to better describe the site's purpose (e.g. Programmers: Architecture, Planning, & Concepts).

SA | Make Scope More Accessible

Display the scope in a simply-worded form next to the 'ask a question' interface; this makes it more likely users will read it before posting.

BL | Add Blurb

Add a descriptive blurb that is displayed to users before they are able to post their first question. Make this focus-stealing; i.e. background is greyed-out and the user must take an action to continue (e.g. click 'got it').

CS | Change Scope

Change the scope to include some of the areas that are currently being down-voted as off-topic.

CE | Change Enforcement

Actively cultivate questions to improve them; narrow their focus if they are too broad, make them more abstract of they are too narrow (i.e. they are debugging specific code instead of asking about architecture/concepts), etc. This would mean slower closing of questions and the encouragement of users to learn about what is/is-not suitable for this site.

CH | Change Help Center

Create a series of Meta Q&As that are designed to help user improve their answers. This way people can post links to those discussions rather than forever trying to discuss the same things over and over

So... votes, thoughts, ideas, scathing indictments... let the games begin.

  • 20
    +1 for trying to be constructive and trying to understand the problem. Those should not be rare things, but they are. – Ixrec Mar 8 '16 at 11:35
  • My attempt at solving what I see to be the problem, or as you put it alleviate "symptoms" of the problem: meta.stackexchange.com/questions/276621/… – Viziionary Mar 8 '16 at 13:29
  • 2
    @Viziionary Interesting suggestion but I'm going to leave out any solution that involves major changes to the general mechanism of the Stack Exchange network because they are likely impractical for resolving this issue. However, the general point about changing how bad questions are handled (so that more emphasis is placed on improving rather than closing them) is listed under Change Enforcement. – Marvin Mar 8 '16 at 13:36
  • @PeterTòmasScott I think my solution would be harmless to implement and would enact your Change Enforcement bullet successfully. I can't think of any lesser way of successfully implementing the "Change Enforcement" option though. So other than that, I would be an advocate of Make Scope More Accessible. However, even then, the interpretation of scope and wording of questions will often leave room for improvement, where I think my suggestion is a must have regardless in order to solve the deeply rooted issue. – Viziionary Mar 8 '16 at 13:43
  • You claim that not understanding the scope is the main issue, well in terms of bad questions it is, but for the record I'm interested in questions which are on the border of the scope definition and need some improvement. Questions where the user makes an effort to meet the scope but misses the mark. You claim less off topic questions would mean less trigger happy people and more willingness to suggest improvement in questions (in your answer below), but I'm not convinced. I still think a few highly active stray users will be an issue still. – Viziionary Mar 8 '16 at 13:46
  • 7
    @Viziionary Regardless, I think that addressing the bad questions should be the priority. If that situation improves then everyone will be a bit happier to begin addressing other issues. – Marvin Mar 8 '16 at 13:53
  • 4
    I've never really understood why the site was called "Programmers.SE"... we're all programmers, aren't we? Pretty much the same over at SO... so from a newb's perspective, what's the difference? A better name (IMO) would be something like "SoftwareDesign.SE", which makes the site's purpose significantly more specific than "anything programming related". Not saying the name needs to change at this point, just that it was always a bit of a misnomer to me. – Mage Xy Mar 8 '16 at 18:35
  • 2
    @MageXy it was named "Not Programming Related" in the old days. That didn't work well. SE has consistently refused to entertain the possibility of renaming the site. – user40980 Mar 8 '16 at 21:04
  • Peter, do note that this site has already changed scope at least once. Maybe people are correctly identifying a problem with the current scope and rules? Are you weary of people complaining about it because you don't think there is any problem? – Andres F. Mar 9 '16 at 4:22
  • @AndresF. I think there is a problem, but it has the hallmarks of one that will stagnate by committee. I've added a conclusion of what I think should be done; if there is broad agreement then we will have a way forward, if not then I'll just wind my neck in and leave the politics to others. – Marvin Mar 9 '16 at 10:42
  • 1
    Personally I would favor making the scope more about "programmers" if this site is keeping the site name of "Programmers.SE". I think the (highly upvoted) answer here explains that I mean. I would want to keep our current quality standards though, and make sure we don't digress into the old "what should I name my cat" or "do you fart in a cubicle" ways (yes, those were both questions at one time). – Rachel Mar 9 '16 at 21:41
  • 2
    I think the biggest problem with this site for new users is that it is not really for new programmers. Unlike, let's say, SO, a new programmer may come there with quite a trivial but valid question and people help him. This site is more about design and architecture itself and to know what an architectural/design question is, to be able to distinguish it from a code-request, you need to have certain background in programming. I think newcomers don't really see the purpose of this site. – Andy Mar 11 '16 at 19:49
  • 2
    I invite everybody to have a look at my meta question concerning the outcome of this survey. – Doc Brown Mar 12 '16 at 9:36
  • @DocBrown Post an answer referencing your Meta question and Rachel's Meta question and I'll accept it as the answer, I think those two questions combined form the correct next steps from this one. – Marvin Mar 12 '16 at 20:51
  • 1
    @DocBrown Actually, I'll just add links to those questions in my conclusion post. – Marvin Mar 12 '16 at 21:10
32

Frankly, I think that you are off base. Or at least your "change enforcement" is too thinly scoped.

No amount of name changing or scope massaging or helpful blurbs will fix the community.

As it stands, we have a fairly insular group of regulars who view questions as something to vet - things to protect the site from. We reliably have purges where old, popular questions are closed. And the regulars do a great job at enforcing the rules to the letter.

That is a huge problem.

SE sites are here to help people. The rules exist to aid all of us in that goal. But I fear that many people have lost that perspective. The rules have become sport; the questions our game. The regulars fight over who can be the quickest to close, who the most rules-lawyery - rather than who can be the most helpful.

Yes, this site gets a lot of crap questions. And I'm not arguing that crap questions shouldn't be promptly and thoroughly razed. But there are a lot of useful, interesting questions that are closed because they're too close to an existing question, or a bit broad in their scope, or heaven forbid have differing valid expert opinion answers.

We spent too much time and effort pruning and not enough time and effort cultivating. Is it any surprise that things aren't growing?

It is entirely a people problem this site is facing, and things will not change without the people involved changing.

  • 11
    +1 for "SE sites are here to help people. The rules exist to aid all of us in that goal. But I fear that many people have lost that perspective. The rules have become sport; the questions our game." – Andres F. Mar 9 '16 at 4:31
  • 1
    I've posted another bit on meta about the questions that were closed out of the 50 newest ones the other day. Would it be possible to help identify which ones were incorrectly closed, and assist cultivating those that can be improved? To me, it seems like people are suggesting that half or so of our questions are ones that we are making mistakes in pruning on. And while people occasionally point to one here or there, I have difficulty believing that this is the general case. – user40980 Mar 9 '16 at 14:03
  • @MichaelT - looking at the list, 1, 2, maybe 21, 33, and 34. Though these are all rather borderline, even by my standards. I expect that half or more would still end up closed after an effort to work with the OP to clarify/improve things. To be clear, I don't think that it's anywhere near half of the daily closed questions are overzealous closures - probably closer to 5-10%. – Telastyn Mar 9 '16 at 14:26
  • 3
    @Telastyn 5% amounts to the two questions that Thomas pointed out. And yes, that does feel about right. And that leads to the question of how do we either get knowledgeable people on the subject to help the OP refine the question. When 90-95% of the closed questions are ones that you have to sort through to get those in front of eyeballs that care to fix them and can. You will note with #2 and #33 I've tried to get enough information from the OP to get something that can be answered... but haven't gotten anything back from the OP in either case. – user40980 Mar 9 '16 at 14:32
  • 4
    "I fear that many people have lost that perspective. The rules have become sport; the questions our game." Yes, this is exactly what's been going on here. It's a classic problem: when you want A, but A is hard to quantify so instead you measure (and reward) B which is highly correlated to A, you end up getting lots of B but a minimal amount of A. – Mason Wheeler Mar 9 '16 at 16:45
  • I entirely agree to this, and I admit, though I spend a lot of my time for answering and trying to leave at least helpful comment to a bad question, I do not always hit the right tone and I am not always so polite as I should be. I am not a patient guy. ... – Doc Brown Mar 11 '16 at 12:14
  • 5
    ... However, I do not think the problem is solved by closing less questions. IMHO MichaelT has proven, that there is objectively a high number of low quality questions which cannot be saved with reasonable effort. So IMHO we should continue to close them as frequent as we did in the past, but try to be "more polite" in that process. For example, when a newbie asks an off-topic questions, some close votes (and a hint what he did wrong) are enough, giving 5 additional downvotes does not help him more. That is what makes the site looking hostile. – Doc Brown Mar 11 '16 at 12:17
  • 3
    Another example: the wording of the "canned comments" for the prepared closing reasons might be improved to not look so hostile. Same is most probably true for the list of "canned comments" some of us (you know who I am talking of) might have prepared over the years to use as a standard respond to those questioners which ignore the site policies. – Doc Brown Mar 11 '16 at 12:24
  • 1
    Close votes can be hostile when they are misused. When someone asks a direct question, clearly stated which perhaps didn't fit on the site, "Unclear what you are asking" and "Too broad" are hostile votes. For instance here – Gort the Robot Mar 11 '16 at 23:37
  • This. So much this. – Lightness Races with Monica Mar 12 '16 at 13:47
  • @StevenBurnap - I think the Unclear and Too Broad questions should be addressed in a way to help the OP rework the question without a Close or down vote. At l;east give them some time. If anyone ever asks what book to read, class to take or language to learn, it just needs to be deleted without debate. – JeffO Mar 17 '16 at 18:46
  • It's also worth noting that even if a question doesn't get closed, getting one or two duplicate votes which are obviously totally spurious can be dispiriting to the asker. It also often hurts question quality because they make edits which should be unnecessary explaining why the questions aren't the same – Ben Aaronson Mar 18 '16 at 9:47
  • @Telastyn wow, didn't think I'd see someone like you on Programmers Meta, getting upvotes, no less. I'm still determined never to use this stack if I can help it, but if people of your disposition come to dominate here next time I look, I may yet change my mind. – Stumbler May 13 '16 at 13:55
15

Here are my thoughts on the solutions currently listed.

Change Name is BAD because it was already rejected by SE top brass and for seemingly valid reasons; namely, there isn't an alternative name that narrows the definition whilst staying true to this site's purpose.

Add Sub-title is GOOD providing that the right wording could be found. Sub-titles are less restrictive than the main name so they can be a little longer and they can be very effective at refining the definition of the site to a perspective user; they are more likely to read a sub-title than a specific page that defines what is/is-not on topic.

Make Scope More Accessible is GOOD because I agree that finding an authoritative description of the scope can be tricky and putting this front-and-center will at least reduce some of the confusion.

Add Blurb is NEUTRAL because I don't know how viable it is or if it would be read. If anything I would lean towards GOOD because, if it is worded correctly and is sufficiently brief, the fact that it is specifically targeted at new users could be a great improvement.

Change Scope is BAD because the site's scope appears to have been modified to its current form for valid reasons; namely to ensure that the site remains useful to programmers.

Change Enforcement is NEUTRAL because I think it is separate to the current topic. To my eyes, fixing the problem of lots of off-topic questions being posted here would naturally result in people being less trigger-happy with the border-line on-topic questions.

  • "there isn't an alternative name that narrows the definition whilst staying true to this site's purpose" - "SoftwareDesign" as proposed by @MageXy above seems a good, descriptive name to me; "SW-Developers" comes to mind, too, but may still be too ambiguous. – Murphy Mar 8 '16 at 21:56
  • 1
    @Murphy SoftwareDevelopers is too broad (it is no more narrow than programmers). SoftwareDesign is possible but would need some discussion as to whether this includes the whole scope of the site. FWIW I think that a subtitle/tagline would be more effective than a name change. – Marvin Mar 9 '16 at 1:28
  • 3
    This isn't bad, but I'd consider that things you marked BAD happened in the far away past. For example, the name change was rejected in 2012! Sites, people and rules evolve. Maybe it's time to re-evaluate those decisions? I happen to think the name programmers.SE is part of the problem (maybe I'm totllaly wrong; maybe I would have been wrong in 2012 but not in 2016?). – Andres F. Mar 9 '16 at 4:26
  • @AndresF. You may be right. I think that name change was also rejected at the end of 2015, but I may be wrong. – Marvin Mar 9 '16 at 10:38
6

It's the unfortunate fate of software which interfaces with different facets of the collective human behavior. These reactions from the users are inevitable; but can be managed.

A tagline (or subtitle) would go a long way. This tagline should serve as a reminder on all events: questions, comments, votes, and so on. It would serve as a means to return the user's mindset to the original principle which birthed the site.

A blurb would also work on first post. I suggest there should be a series of blurbs which gets less in the face of the user when the site "learns" the user complies more with its principles.

  • 3
    I like this idea, I wonder if SE will let this be changed if site name is out of scope for a change. – enderland Mar 8 '16 at 14:32
4

I think part of the problem is it is too easy to ask questions. Not that it should be difficult, but perhaps require taking the tour first even for established users (+100 association bonus). While this is a low bar, I doubt it would work.

It is difficult to force users to read and understand the help center, for example. I am reminded of every time I have had to take sexual harassment training on the computer for work. Play on my phone until the required time passed for a "lesson" then check the checkboxes and hope for the best. Asking on a Stack Exchange site is no different.

When I ask or answer on a Stack Exchange site where I do not already regularly contribute, I am careful to read the help center and some of the meta posts first. I look at recent questions that received both positive and negative attention. I ask myself "why did this question get upvoted, and another downvoted or closed?" In general, this has served me well (also, never post when drunk).

Certain sites go above and beyond the general scope and quality requirements. Code Review (6 "yeses"), Skeptics (citations), Programming Puzzles & Code Golf (challenge type, scoring), and a couple others come to mind.

Programmers is not one of those. We just want a reasonably-scoped, clear, objectively-answerable, on-topic question and answers that actually answer the question as asked. We have a help center and many meta posts describing our standards, which do not require an excessive amount of effort to meet: they are simply unclear to some people.

I do not believe there is anything more we can do to help the situation except for token gestures that in the end, users will ignore.


On another note, I know several of us do try to post meta-links in response to questions that fall into one of several common traps:

There is no point in rehashing the same points over and over: post a link, then if the asker is interested in working with us, we are generally willing to help. If the asker is genuinely concerned in learning and has enough reputation, we have even invited them to The Whiteboard to hash things out.


Finally, you make a point about possibly changing scope to match what users want. We tried this, it did not work. I recommend reading up on this site's history if you have not done so already:

What is the history behind the site scope change from NPR to “conceptual questions about software development”?

3

Actions Taken

Suggestion 1 is being advanced by a renewed request for name change, this is being advanced by @Rachel in Four years later - Can we change our site name?.

Suggestion 2 is being advanced by active community discussion about change in moderation style. This is being advanced by @DocBrown in How to improve our style for reacting on “low quality questions”?.

Also see Arron Hall's discussion about Can we resuscitate more content?.


Recommendations

It is now clear to me that there are two problems with two camps; these are:

  1. Improper questions being asked on P.SE; this pertains to users not understanding the site's scope and purpose.
  2. Improper moderation/closing; this pertains to questions being closed too abruptly and with too little communication.

I expect that any solution that looks to resolve one of these and not the other would simply fragment the community. So, I suggest that we move to address both problems simultaneously, with one or more people from each faction leading-up a solution.

Suggestion 1. Improve communication of the site's purpose. From the results, this can involve creating a site tagline, creating a blurb that is displayed to new users before they can post, or by displaying the site scope next to the ask-question interface. These all require SE involvement so must begin with feature requests for our preferred solutions.

More immediately, Snowman's suggestion for a series of Meta Q&As about common question phrasing issues could be implemented. This would be advantageous because the community can do it without SE involvement. It also has the benefit of actively assisting Solution 2.

Suggestion 2. Change the self-moderating style I would suggest that this is best addressed by those advocating change being given the opportunity to moderate how they wish for a certain period of time once a week. There can then be discussion about improvements that are seen and lessons that can be drawn by the rest of the community.

Survey Results

Broad support was expressed in one form or another for all of the 'improve communication' type solutions; the general conclusion is that all of the options presented would be helpful and the question becomes how much support we will receive from SE regarding feature requests. In fact, the only negative comments on some solutions (such as name change) were motivated by the expectation that SE would not be supportive of the measure rather than by a dislike for the measure itself.

There was also extensive support for changing enforcement / moderation style. Related to this was the suggestion to update the 'how to write on-topic questions' related Meta posts to make them more helpful for newcomers.

Any suggestion of scope change was universally condemned.


Conclusion

We now have two clearly defined areas for development and active community participation in both areas. Moreover, the two solutions being developed accurately reflect the entire community, which greatly improves the likelihood of success.

As such, this question has served its purpose and the poll is now closed.

  • Thanks for the summary, but I have to disagree with your statement that any suggestion of scope change is universally condemned. This answer is at +72, and I think accurately summarizes the type of scope change the community would like to see if allowed. The problem is it is condemned by the people to run the site, such as Stack Exchange and the most active meta users. It is actually something that is very much supported by the majority of P.SE users, and is in fact the scope the site was first founded to support. – Rachel Mar 16 '16 at 3:44
  • @Rachel I was summarising the content of this discussion; during which scope change was rejected explicitly several times without a single proponent. As for he link it is three years old; it is irrelevant to the current discussion. So I don't see why you are unhappy with my summary. – Marvin Mar 16 '16 at 3:48
  • I actually like your summary, but the line stating the scope change is "universally condemned" really bothered me. We were chatting about that answer last week in chat, which is why I was thinking of it. Thank you for putting together this question and summary though :) – Rachel Mar 16 '16 at 3:49
1

Alright I'll bite. I have a few general points, but first off, it was my question referenced by OP, and I'd like to point out that while I used questionable / bad wording ("Is it possible?"), there was more to my question than that, taking it beyond the yes/no answer format, and of the scope, it certainly seems to meet at least one point, if not a few other vaguely:

  • software requirements
  • software architecture and design
  • algorithm and data structure concepts
  • quality assurance and testing
  • development methodologies and processes
  • software configuration management
  • software engineering management
  • software licensing

But it certainly had the potential, with a few edits, to become a good question. Instead it immediately, within a few hours, got 3 negative votes, 3 votes to close, and zero suggestions for improvement.

This is the major trend for questions in need of improvement on this site, and my whole point is that a large portion of the self-moderating activity on this site is making the problem of poor questions worse by choosing to down-vote and close rather than suggest ways of improving the question.

Stack Exchange communities improve their user base's question output quality by actively and continuously suggesting improvement to salvageable questions. By immediately pulling the trigger on 3 down-votes, 3 close votes, you skip this step completely.

I pointed out a specific user who I've noticed over a wide range of cases, and a long span of time, has been the "first-responder", and potentially the catalyst for this trend, because psychology and common sense show that after the first vote is made, it will encourage further votes in the same direction. This goes for close votes, positive votes, and negative votes. People follow suit, suddenly one extremely active user is causing alot of damage when his behavior is followed by others systematically over a long period of time.

Just think of how many people you could have trained to ask better questions in the past year with a few moments spent suggesting improvements to questions rather than just cutting them down, closing, and offering at most a link to a related meta post with no explanation.

Thousands.

I made a suggestion on Meta.SE based on my thoughts here.


Side-note regarding your improvement suggestion:

If you want to make more clear what your awkwardly strict and confusing scope is, how about making it more accessible?

Right now, the scope accessibility is horrible in terms of UX. It requires 2 page changes just to find the full scope definition, and that's after having to read through all sorts of meta information to find the right link.

How about placing the scope definition as an expandable module right there on the question asking interface, and perhaps even use some form of subtle animation to bring it to the attention of new users?

Or even just make it one click away? If you're trying to educate users on your scope the last thing you want is to hide the information behind two links.

  • I've added the suggestion regarding scope to the list of solutions. I've also made the list of solutions in the question impartial and have moved my opinions into a separate answer. – Marvin Mar 8 '16 at 13:04
  • Also, for the purposes of this discussion, could you add a summary to the top of your answer? I expect that you would be expressing support for the Make Scope More Accessible and the Change Enforcement solutions? – Marvin Mar 8 '16 at 13:06
  • 7
    "Just think of how many people you could have trained to ask better questions in the past year with a few moments spent suggesting improvements to questions rather than just cutting them down" – The problem with this is that it only works if the average time to improve a question is significantly shorter than the expected time between arrivals of two bad questions, which in my observation is not true. In other words: crap is flooding in so fast that you cannot afford to take time to improve the crap because during the time you spend improving the crap you accumulate more crap than you improve. – Jörg W Mittag Mar 8 '16 at 14:15
  • 3
    Adding to what @JörgWMittag said, many of the problem questions are from brand new "drive by" users who do not ask repeat questions. Most of those that do ask repeat questions ask identical or very similar repeat questions and ignore the advice given by the community, argue with edits, or otherwise are contrary to this goal. For every 1 person who will actively edit their question and improve it, there are dozens or hundreds that don't. – enderland Mar 8 '16 at 14:29
  • 1
    @JörgWMittag Until you include the factor of improved crap turning around and helping you deal with the problem in the future, and this effect being exponential. When you create a friendly environment for people in your community, help them improve, more of them will become active members than if you just vote to close them. Each active member increases the result moderation productivity, and you can bet those people you welcomed and spent time helping will follow suit, and one day welcome people themselves and there is your exponent. – Viziionary Mar 8 '16 at 14:29
  • 2
    And for what it's worth, there is a chatbot posting any "post on programmers" comment on Stack Overflow - the education problem is there -- not here. I encourage you to join that chat and participate and see how many low quality questions get recommended by users of Stack Overflow to be posted here. It is easy for you to post a "todo" list for regular users here, but us regular users here see a deluge of completely off topic (as in, even with edits never going to be on topic) questions here every day. Understandably it's easier to blast the site on meta than participate yourself. – enderland Mar 8 '16 at 14:30
  • @enderland Just because moderation is difficult doesn't mean problems with it dont need to be solved. If, instead of spamming close votes you took a moment sometimes on potentially improvable questions, lets say in a year you write 1000 of such comments. If as Jorg said 1 / 25 attempts leads to an improved question, a user responding to a friendly community, and of those 40 successes, 4 of them become strongly active members on the site, you've personally multiplied your effectiveness by 4 and those 4 people will each be treating people they way they were treated, helping improve questions. – Viziionary Mar 8 '16 at 14:36
  • @enderland the following year that number becomes 20 people that instead of leaving after wording a question badly and getting cut down, were helped to improve, and went on to become effective members of the community. In two years on the site, your investment of time helping a few people now and then to improve becomes a massively exponential growth of positive self-moderation activity on the site, and that's with you acting alone. So if you consider my numbers too high, fine, but with 2-3 people doing this the outcome is still very good. – Viziionary Mar 8 '16 at 14:37
  • This is a much better post than your previous meta question, and I can't say I disagree with any of it. – Lightness Races with Monica Mar 8 '16 at 14:42
  • 5
    @Viziionary I encourage you to spend some time trying to edit, comment, and improve questions here before so completely assuming it is possible to "fix" as many as you suggest. It is easy to tell others what to do when you don't understand what it is you are asking them to do. – enderland Mar 8 '16 at 14:49
  • 1
    @Viziionary Why does that matter? Not being an "answerer" doesn't mean you can't help people make their questions more on topic, ask clarifying questions, edit to make more clear, etc. In other words, be the change you want to see. – enderland Mar 8 '16 at 14:55
  • 4
    @Viziionary, ah, so you'd rather other people change than actually change yourself? That is not a good way to affect change, demanding other people adjust their site (since as you say, you don't really want to be a part of anything other than having your questions answered) to meet your desires, while simultaneously saying "not my site to moderate!" It makes me not care what your perspectives are since you are saying, "I don't want to be part of the site, but I want the site to conform to my desires so I can get answers anyways." – enderland Mar 8 '16 at 15:09
  • 4
    @Viziionary Be nice. – yannis Mar 8 '16 at 15:21
  • 2
    @Viziionary that's fine. But consider that I am not only writing to you, but also to others who might read this - perhaps you are unwilling to do anything regarding community moderation. Others who read this may feel empowered to actually participate in community moderation (who were previously unaware this was an option - it's surprising how many people seem unaware that "regular" users can be participants in moderation). – enderland Mar 8 '16 at 15:21
  • 2
    @Viziionary If you feel someone is being rude, flag their comments. "returning the favor" won't get you anywhere (good, that is). – yannis Mar 8 '16 at 15:25
0

Of the list of suggested solutions, the only one that could plausibly fix the problem is a change of scope. The core of the problem is that given the scope, the set of strictly permitted questions is, if not quite empty, something you would have to search through a thousand off-topic questions to find a valid example of. Certainly it is rare for the front page to contain a single one.

Virtually all questions of software are either:

  • matters of implementation (Stack Overflow)
  • matters of computational mathematics (Computer Science/Theoretical Computer Science)
  • sufficiently well-defined to be implemented by a tool or framework (Software Recommendations)
  • opinion-based, or have multiple valid answers (Reddit or whatever)

If you accept that is the case, clearly no level of moderation is going to fix that. And if those who attempt an impossible task sometimes become less than perfectly polite, then that's hardly surprising...

  • search at the site suggests that there are over 28,000 "valid examples". And about 10,000 remaining questions aren't necessarily "invalid", some may miss just a minor edit to reopen – gnat Mar 17 '16 at 14:56
  • I read through the first three pages of your link, couldn't find any questions that qualified. Maybe you can point to a single example of a question you consider valid? – soru Mar 18 '16 at 10:34
  • all 575 pages at this link are qualified, including first three. These are questions that aren't closed, ie valid by definition – gnat Mar 18 '16 at 10:38

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .