In the spirit of The Great Question Deletion Audit of 2010, I give you the great question deletion audit of 2012.

These are basically all of the highly-voted questions on the first "Most Votes" page that, if asked today, would quickly be closed as Not Constructive (this is not a complete list):

List of freely available programming books
What is the single most influential book every programmer should read?
Hidden features of C#
Hidden features of Python
What IDE to use for Python?
What are the lesser known but useful data structures?
What is your most productive shortcut with Vim? (Draft on the super user blog (link is editors only) of the top-voted answer, with accreditation.)
What makes PHP a good language?
What are Code Smells? What is the best way to correct them?
Hidden .NET Base Class Library Classes
New Programming Jargon you Coined
Mercurial for Beginners: The Definitive Practical Guide
The Definitive C++ Book Guide and List

Moved into tag wikis

Git for beginners: The definitive practical guide The Definitive C Book Guide and List
Most Useful Free .NET Libraries
Learning to Write a Compiler
Database development mistakes made by application developers

Which ones should stay, and which ones should go. Why?

Note that this is still a contentious issue, even among the mods. You could be a great help if you can articulate some guiding principles that could be applied fairly and equally to such questions.

For those users who are interested in preserving the information contained within these posts for personal use at a later stage, see StackPrinter made by systempuntooutless.

closed as too localized by Robert Harvey Mar 13 '12 at 18:55

This question is unlikely to help any future visitors; it is only relevant to a small geographic area, a specific moment in time, or an extraordinarily narrow situation that is not generally applicable to the worldwide audience of the internet. For help making this question more broadly applicable, visit the help center. If this question can be reworded to fit the rules in the help center, please edit the question.

  • 52
    And before anyone suggests migrating them to Programmers let me state here and now that I will close and delete any that do get migrated. – ChrisF Feb 12 '12 at 22:51
  • 33
    You are like a hunter, trying to kill the last dodo birds. – bobobobo Feb 13 '12 at 0:32
  • 34
    I don't see the need to delete. The quality of the site will be determined by the accessibility of the good information. Just the good stuff will float to the surface: it is the Way the Web Works. I'd suggest starting deleting at the other end (the ones with most downvotes...) – sehe Feb 13 '12 at 16:16
  • 10
    "This post needs 884 more deletion votes to delete" Like a drop in the ocean... (never did agree with that "popular questions should be harder to delete" feature...) – Adam Davis Feb 13 '12 at 16:25
  • 19
    dumb question, can only receive dumb answers. Why delete good content? – Tony The Lion Feb 13 '12 at 17:14
  • 10
    @Tony: It depends on what you mean by good. If by good you mean bikeshedding, big list questions that encourage other, far lower quality bikeshedding, big list questions, then I guess these are good. – Robert Harvey Feb 13 '12 at 17:15
  • 6
    Seeing as you proposed this action, I see it as up to you to provide some reasoning as to why these questions should be deleted, and if you answer is that is has been done before, I want my servants! – thecoshman Feb 13 '12 at 17:21
  • 4
    Dear Jeff, you have two weeks left in the office. Why not go out on a high note, and move all these old, silly, popular questions to a new section of SO—legacy.StackOverflow.com???—in read-only format, and put an end to this nonsense once and for all. – Adam Rackis Feb 13 '12 at 17:35
  • 28
    Can someone illustrate what the problem seems to be? Over and over, the best argument heard is "These questions make bad examples". Show me the boat loads of bad copy-cat questions, and I might start to see a reason to move/delete these 'examples'. – sehe Feb 13 '12 at 17:38
  • 11
    @AdamRackis "silly and popular", so you want SO to contain only "boring and never looked at" questions. I think I've either missed the point of SO or the purpose of this site has gone more corrupt than can ever be imagined. – Tony The Lion Feb 13 '12 at 17:42
  • 12
    @sehe: You can't see them, by design. Every time somebody asks one, it gets deleted by community action. Then the community has to defend itself on meta and in comments with the OP, because questions like "The Definitive C++ Book List" still exist, proof that such questions are seemingly on-topic. This process is not free; it costs the time and effort of everyone involved to explain repeatedly why your question stays, but their precious question cannot. – Robert Harvey Feb 13 '12 at 17:43
  • 7
    @Tony The point of the site is for programming problems to be solved, not a popularity contest. I can sort of understand people who like the questions and want them to stay, but I am completely not seeing how you could think the point of the site is a bunch of subjective lists. Or how it's "corrupt" to want to focus on solving real problems. – Matthew Read Feb 13 '12 at 17:46
  • 4
    I AM SAD. So many good FAQs gone, because of pedantry. – Raynos Feb 27 '12 at 1:01
  • 5
    The current state of this question fits the current state of that c++ books question, so that seems to be fine. I see no need to destroy that harmony. Can someone elighten me? – Johannes Schaub - litb Mar 3 '12 at 21:12
  • 5
    @sbi: I think it's great that there's a group of C++ tag users who maintain that list. I don't want to see it deleted. I think it's useful info. I am now aware that it's not in the tag wiki. Maybe we should print it up in poster format for every SO user. You and DeadMG are obviously really cool people who are way smarter than me and are absolutely going straight to Heaven++ with Bjarne when you die. Is that good enough? Can you stop beating me with your pointless conspiracy theories now? – Josh Caswell Mar 3 '12 at 21:17

20 Answers 20


I can't speak for any of the other questions, but deleting

The Definitive C++ Book Guide and List

is about as dumb a suggestion as is possible.

This is one of the two question I link to most often, and the uproar in the C++ chatroom seems to confirm my suspicion that this is the single most linked-to question in the C++ tag. What's more, this is the second-most linked-to question on the whole of Stackoverflow, only surpassed by a question twittered by Joel and Jeff!

This question is immensely useful, and relieves us of the task of writing this list again and again — in strongly varying quality. This question is one of the most important parts of the C++ FAQ, an effort started by regulars from SO's C++ tag to provide high-quality answers to frequently asked questions.

Also note that this question perfectly fits a plea from one of Joel's most famous SO blog postings:

Help us build a great library of canonical answers. If you keep seeing the same form of questions, whether it’s mod_rewrite rules on Server Fault, freezing computers on Super User, or how to use regular expressions to parse HTML, write a great, canonical answer, once and for all. Make it community wiki so that as many other people as possible can make it great. Work really hard on writing something that is clear, concise, and understandable by as wide an audience as possible.

This question does exactly that. It's also a perfect fit for the site's overall goal to make the Internet a better place. Yet you want to delete it despite it perfectly fitting the spirit and goal of the site, just because it violates some moderation laws which were, once, put into place to implement that goal, not to hinder it.

  • 22
    +1 "Help us build a great library of canonical answers" sums it up – sehe Feb 13 '12 at 17:29
  • 20
    The debate is mostly about whether book recommendation questions belong on the site at all. I don't think anyone disagrees with the quality this particular post has. But for every Academy Award quality post like "The Definitive C++ Book Guide and List," there are a thousand monkeys bashing away at a thousand keyboards asking book recommendation questions that will never see the light of day. Can the C++ book list go into the Tag Wiki? Do you have suggestions for ways these bikeshed questions could be better handled? – Robert Harvey Feb 13 '12 at 17:32
  • 15
    @Robert: Not really, because Tag Wikis suck. Who cares if, in someone's opinion, it doesn't belong here? What matters is that it's helpful. – DeadMG Feb 13 '12 at 17:35
  • 20
    You guys are missing the big picture. SE Corporate (especially Robert Cartaino) has been discussing for months ways to make Tag Wikis more relevant, places where communities like C++ and compiler-design can craft their own little personalized outposts. Why would you not be in favor of that? – Robert Harvey Feb 13 '12 at 17:37
  • 6
    @RobertHarvey: I am. If they make a post describing the proposed improvements, then I may well come out in favour of moving such content there. As they are now, however.. – DeadMG Feb 13 '12 at 17:42
  • 13
    @Robert - for every Academy Award quality post like "The Definitive C++ Book Guide and List," there are a thousand monkeys... That sounds like Hollywood today. I'm perfectly capable of watching Moneyball and ignoring Jack and Jill in real life, so let's just close the crappy book recommendation questions that pop up, and leave the good ones like this. – Adam Rackis Feb 13 '12 at 17:42
  • 21
    @sbi: To put it bluntly, all I keep hearing from the C++ folks are reasons why tag wikis suck, or comments about how moderators are lording over Stack Overflow, but not any actual, workable solutions. Frankly, I don't find either approach constructive. – Robert Harvey Feb 13 '12 at 18:24
  • 16
    @RobertHarvey: Oh c'mon! At least give us the courtesy to put in the effort and mock up a strawman argument, rather than attacking "the C++ folks" on a personal level. There is a very simple and obvious solution: Do not delete that question. If you cannot read this out of my answer, then you should put down that diamond immediately. – sbi Feb 13 '12 at 18:31
  • 10
    @jalf: The question states: "You could be a great help if you can articulate some guiding principles that could be applied fairly and equally to such questions." – Robert Harvey Feb 13 '12 at 18:34
  • 10
    I think @DeadMG has hit the nail on the head - the book list question would be perfect for the tag wiki if it could be easily migrated and sufficiently visible. Tag wikis are almost invisible to new users and can't be used as targets for close as dupe currently. – Flexo Feb 13 '12 at 18:43
  • 3
    The thing is, if we do move it to the tag wiki, what will happen to all the links that point to it? I link to the question more than any other question on this site. – In silico Feb 14 '12 at 2:51
  • 48
    @RobertHarvey: Well, such guidelines are rather simple to come up with: Do not delete valuable content, no matter what the moderation laws say. Stick to the spirit of the site, not to the letter of the law. Do not wreak havoc by deleting stuff that's linked to (literally) in a thousand places. Listen to your users. If they are offended, blame yourself. If in doubt (and why else would you be asking?), keep stuff, rather than deleting it... This is called common sense and has been requested here at meta quite often. What more do you need? – sbi Feb 14 '12 at 10:36
  • 4
    In short most of the "answers" to that "question" really are suggested edits to the actual answer part of the "question". That's not how a Q&A works. Don't use a Q&A engine to maintain a wiki article. – badp Feb 14 '12 at 11:38
  • 9
    @awoodland I do see the value in the content. There is no question that the content in there is valuable. It doesn't mean Stack Overflow has to host it in the Q&A area. What's the practical problem? Where's the expertise? That's not a question with answer, it's a book popularity contest. Also see this very awesome essay by (now community manager) Grace on a similar topic. – badp Feb 14 '12 at 11:56
  • 13
    @sbi I was dubious about your claim about the number of times the question is linked to (as I am of any unsupported claim). So I looked into it. The delete confirmation dialog reports 14 as the target for duplicates. This data.se query reports 299. This google link search returns 3,720. So you are indeed correct. – Some Helpful Commenter Feb 22 '12 at 19:19

Don't delete any of them. They represent the hard work of SO users who took time to answer the questions, some in great detail.

Stop trying to poach SO's history and leave the questions as they are.


Thought I'd add a screenshot of a comment from the linked long answer:

Good answer

  • 7
    You'll have to do better than that. The others that have posted here have provided a detailed analysis of their opinions. – Robert Harvey Feb 13 '12 at 0:35
  • 20
    I really have nothing else to say, the argument's been had. – bobobobo Feb 13 '12 at 0:39
  • 4
    +1, noting my agreement though I also have I have nothing "better than that" to add. As I said before, the software forcing us to chose between exposing broken windows and deleting popular content creates a stupid situation. – Guest Feb 13 '12 at 0:43
  • 2
    @bobobobo Actually, the answer to that argument is considered out-of-date, the current approach is not to lock these questions, but to close them, as a precursor for deletion. – casperOne Feb 13 '12 at 1:14
  • 10
    @casperOne: Then the current approach is stupid, wrong, and does damage to this place. – sbi Feb 13 '12 at 16:31
  • 8
    @sbi Then I would encourage you to actively work to help shape the place through the proper channels (meta posts are one way) instead of decrying it through comments, as comments are not really actionable items. – casperOne Feb 13 '12 at 16:35
  • 7
    @casperOne: OTOH, though, why would I need to care? This is but one of hundreds, if not thousands, of useful programmer forums on the web. Yes, it currently is the best, but if you delete that reputation, I could always switch to some copy of it with saner policies. I came here to ask and answer programming questions, not to be a politician, and I have deviated far from that already. As thousands of other programmers who come here, I do not want to police this place, I want to learn something about programming. – sbi Feb 13 '12 at 16:43
  • 6
    @sbi No one is asking you to police this place, but if you're unwilling to participate in the discussion about how this place is shaped, then I'm sorry as your concerns will never be heard, known, and possibly acted on. You've been given a tool and a voice, if you want to use the tool, but choose not to exercise your voice, then that's on you, not on the site. I appreciate that not everyone's voice on meta is heard as much as they think it should, but that's a problem of the scale of the site & prioritization with limited resources. Be part of the solution, not an unheard part of the problem. – casperOne Feb 13 '12 at 16:53
  • 6
    @casperOne: What about "I am currently writing an answer to this very thread" didn't you understand? – sbi Feb 13 '12 at 17:04
  • 1
    @casperOne: And what about "users come to SO to learn about programming, not to police, and when you won't let them do this, they'll simply leave" did you fail to grasp? SO is not a true democracy, that has been hammered down our throats very firmly. It's a website to earn money from while sticking to a nice and shiny goal ("making the Internet..."). You cannot force users to exhaust themselves in endless debates about how to implement this. And you cannot force them to stay if you fail. – sbi Feb 13 '12 at 17:04
  • 9
    @casperOne: Just look what happened to bobobobo, who did speak up. The regulars-turned-politicians turned him down violently. And that's not the first time I have seen this. OTOH, some of Jeff's answers where downvoted into oblivion, and still he wouldn't give an inch. Why should SO users bend over backwards and wear themselves out discussing here with the political cast, when what they actually want to do is to discuss programming? – sbi Feb 13 '12 at 17:09
  • 2
    @Jeremy - the software forcing us to chose between exposing broken windows and deleting popular content creates a stupid situation EXACTLY – Adam Rackis Feb 13 '12 at 21:05
  • 9
    @sbi if you feel so strongly, then leave. Otherwise work towards a solution. Your invective isn't helping matters. – Jeff Atwood Feb 20 '12 at 23:02
  • 8
    @Gilles: I think his point is that deletion is what needs to be justified. That the burden of proof should be on those wanting the deletion, not those wanting preservation. Deletion is the option of last resort, not the first thing you grab when you see something amiss. – Nicol Bolas Mar 3 '12 at 21:07
  • 8
    I thank sbi for speaking up for the silent majority against tyranny from certain high rep users – prusswan Mar 3 '12 at 21:12

Create a new site, perhaps at archive.stackexchange.com. Prohibit registration; this site would be for preserving content, not continuing discussions. Given the existing migration infrastructure, once a moderator migrates a post here it would be deleted from Stack Overflow, and anybody following an existing link will be redirected to the archived copy.

Migrate these questions en-masse, since they're not on-topic anymore. The content could continued to be mined for appropriate tag wiki material without needing to keep it on Stack Overflow.

No broken windows, no broken links.

  • 2
    Sorry, but how is this different from just locking the question? – André Caron Feb 13 '12 at 18:54
  • 10
    @AndréCaron It removes it from Stack Overflow. In previous discussions many people have objected to locking popular old posts because it ensures the top question lists will always be filled with bad examples, encouraging users to ask new bad questions. Deletion (or migration) is the only way to get rid of them. – Guest Feb 13 '12 at 19:09
  • 1
    @AndréCaron: because the users who would complain about not being to ask the question again wouldn't see it on the normal site. – user7116 Feb 13 '12 at 19:11
  • 1
    @sixlettervariables: it seems another discussion on meta already addresses that point. And you can see it applied in the definitive C++ book guide and list at the bottom of the post (though I would have added it at the top to make it more obvious). Besides, if it's not on StackOverflow, new users will assume it wasn't asked, which it already has. I rather closing as a duplicate than closing as "off topic". – André Caron Feb 13 '12 at 20:10
  • @AndréCaron: the problem with locking a question is an unfortunate subset of our users do not know how to read or write. That is why moving the questions to another site is different from locking the question. As for any book list question, I have no idea what an example would be of an on-topic/constructive question that is a duplicate of that question... – user7116 Feb 13 '12 at 20:12
  • @JeremyBanksʬʬʬ: I see your point, but removing questions won't prevent people (new users) from asking bad questions. And I really rather this discussion was already on stackoverflow so I can close as "duplicate" rather than as "off topic". The first is much more convenient, since it actually answers the poster's question. – André Caron Feb 13 '12 at 20:13
  • 1
    @sixlettervariables: there are no on-topic/constructive questions that are duplicates of that question. Everybody agrees that that type of question is off topic. Where we don't agree is whether keeping it on the site is harmful or not. – André Caron Feb 13 '12 at 20:14
  • 2
    @JeremyBanksʬʬʬ: it's certainly an opinion, but what I can gather from this (and other discussions) is that there is no general opinion. There is no concensus. And based on the amount of votes on the C++ book guide question versus this topic, I would tend to say the the "general" opinion is that keeping it on the site is a positive thing. – André Caron Feb 13 '12 at 20:15
  • @AndréCaron: folks like to point to it as a useful dupe target is all I pointed out. It's a bad example of a good question... – user7116 Feb 13 '12 at 20:35
  • @sixlettervariables: it's also a good solution to the poster's problem. I comment for improvements and vote to close bad questions all the time and feel that's a lesser evil and don't mind paying that price to keep good questions even if they're bad examples. – André Caron Feb 13 '12 at 20:37
  • @AndréCaron: as in, "to answer your question you should read one of the following books" ? Seems like a non-constructive question. – user7116 Feb 13 '12 at 20:39
  • @sixlettervariables: I don't recommend that list as a "solution" to beginner's problems, but I do point to it often to beginners as a comment to get them to know good books. – André Caron Feb 13 '12 at 20:41
  • 1
    @AndréCaron: ok we're on the same page now. I think a rather enterprising individual could migrate the content from that question to a website, slap a banner ad on it, and begin referencing it in posts. When the book list gets deleted they'd have a pretty nice stream of page views if the list is actually useful. – user7116 Feb 13 '12 at 20:43
  • 2
    @Jeremy, such a simple, intelligent solution - I don't understand why this hasn't been done yet – Adam Rackis Feb 13 '12 at 20:56
  • 5
    Related: Building an archive of deleted questions – Pëkka Feb 13 '12 at 21:12

There are several questions there whose content could be moved to a tag wiki.

- What IDE to use for Python?
- What are the lesser known but cool data structures?
- Most useful free .NET libraries?
- Learning to Write a Compiler
- What is your most productive shortcut with Vim?

Most of these are just "List of X" questions anyway, and they're mostly lists of links on top of that. The top answers are also mostly collaboratively edited CW posts, so no one will be losing out on any future reputation if we move the content to a tag wiki and delete the questions.

Note: I removed The Definitive C++ Book Guide and List from the list above because I don't think moving that content around is as easy as the others. There are a lot of links (both on SO and around the Web) to that page that it would be better not to break.

  • 11
    My bigger concern is, do we need to amend the tag wiki system somehow to better accommodate for this kind of stuff? – waffles Feb 12 '12 at 23:12
  • 1
    @waffles: I've been thinking about adding lists of books and lists of links to canonically-answered, frequently asked questions (so people can easily find the link when voting to close as a dupe) to the programming language tag wikis for a while now. That would only effect a small percentage of tags though, so I don't know if a change to the whole system would be needed. – Bill the Lizard Feb 12 '12 at 23:16
  • 1
    @waffles But then I just read your other comments... I like the sub-pages idea if we can just add it to specific wikis (as opposed every wiki having them by default. – Bill the Lizard Feb 12 '12 at 23:19
  • How would you move "Hidden Features of C#" to a tag wiki? It worked for the .NET libraries question, because the links were summarized in the question, and they are all off-site links. There's no telling how much text there is in the "Hidden Features" questions, and I doubt it would all fit in the wiki. – Robert Harvey Feb 12 '12 at 23:22
  • 1
    I think this is a bad idea. "Vim shortcuts" pages are all over the web; "python ide" will give you plenty of search hits; etc. Any list we have will be incomplete/biased and/or bad attempt to compete with a Wikipedia page. I don't want to see what some random tag wiki editor thinks is the most useful .NET library. – Matthew Read Feb 13 '12 at 16:48
  • 1
    @MatthewRead The alternative is to delete the content. I'm okay with that if it's bad content, I just wanted people to know that deleting isn't the only option. – Bill the Lizard Feb 13 '12 at 17:26
  • 3
    @MatthewRead: you know, stack overflow is overrated. All that information is all over the web. Where do you think we got it from in the first place. The web is overrated. Just hire experienced consultants. They'll have the actual experience, instead of just google's echoes of it. Hmmm. Did you mean 'SO is almost like a normal website'? – sehe Feb 13 '12 at 17:33
  • 1
    @sehe I'm not 100% sure what you're trying to say, but if you're saying that our lists of random crap are better than other sites' lists of random crap, I heartily disagree. SO excels at real questions, not lists of stuff someone thought was good, and not all our real questions and answers are all over the web. – Matthew Read Feb 13 '12 at 17:35
  • 2
    @MatthewRead: apparently some exceptions exist, or a bunch of people were drugged when they voted on these. – sehe Feb 13 '12 at 17:36
  • 1
    @sehe The lack of voting mechanisms on many other sites doesn't mean our lists of crap are either more popular or better. You might as well argue against deleting the questions; that discussion's been had quite thoroughly, so good luck with that. – Matthew Read Feb 13 '12 at 17:39
  • 4
    @MatthewRead: so "we want to hear the community voice" but... "voting doesn't count". I think it is the community did the voting. Again, someone show me the actual problem we're trying to solve. – sehe Feb 13 '12 at 17:44
  • 2
    @sehe If you want to post yet another question about why these sorts of questions should be accepted, be my guest. This isn't the place for it. Make sure you bring something important and previously undiscussed to the table or it will be closed as a dupe. – Matthew Read Feb 13 '12 at 17:54
  • 1
    Hmm, that's an interesting use of tag wikis. – BalusC Feb 13 '12 at 19:01
  • 3
    I can kind of see the rationale for moving to tag wiki, but what about the probably thousands of questions that currently link to the C++ book list, for example? I don't know if any of the others are as heavily referenced, but breaking that would be pretty bad – jalf Feb 13 '12 at 21:25
  • 1
    @Bill: As Conrad had found out there's thousands of links to this question all over the web. Deleting this would render all those links invalid, and would thus clearly be against this site's goal to make the Internet a better place. – sbi Mar 5 '12 at 9:32

I personally am of the opinion that the fundamental problem here is that tag wikis don't get enough love. Lots of this stuff could simply go in there. Perhaps if SO implemented some new goodies for tag wikis. I mean, right now, a "Tag Wiki" is a hundred lines, and you get to see about four of them when you click on the tag or mouse-over it. That's really not what we need here for stuff as complex as whole languages or toolchains or operating systems.

What would be great is if a tag wiki could be like, an actual wiki, so, for example, a subdomain which could have multiple pages linking to each other in true Wiki style, would make the idea of moving their content and deleting the question much more palatable.

  • 1
    +1 although I think separate "info wikis" and "tag wikis" would be more beneficial. – Pubby Feb 13 '12 at 16:24
  • Thank you DeadMG; that is a very valid observation, I never thought of that. Can we get per-tag moderators that do this? It is how the C++ tag self-organizes already, anyway – sehe Feb 13 '12 at 16:25
  • 5
    Suggest a feature request on how to make tag wikis better. – random Feb 13 '12 at 16:35
  • 3
    @random: Done. – DeadMG Feb 13 '12 at 16:43
  • Tag wikis are nice in that users don't have to look at them. If a user is so inclined to read them, they get the extra info. The rest of us can continue about our day. – user7116 Feb 13 '12 at 18:53
  • 1
    It's probably also worth noting that this is a woefully underdocumented feature. I didn't even know it existed until now. If some of my rep that disappeared (and there was a lot of it) was for questions that were moved into wikis, it's going into places I had no idea how to access. Someone might want to address the visibility of this feature. – Mike Hofer Mar 4 '12 at 5:00

These questions contain useful information. Now, SO is not in the business of preserving any old useful information. However, much of this information is useful to the business of this site: answering questions.

Being able to say, "Go read one of these" is a good thing. Now, where "one of these" is is something we can discuss. But one thing is vitally important:

We cannot delete these until we know where the information will go!

We don't want to lose any information. We want to transfer it. If a Q&A place isn't where it should go, if some of it should be absorbed into a Tag Wiki or something, then that needs to happen before deletion. Once the information is in its final resting place, the old question can go away.

That being said, there is one reason big reason why you can't use tag wikis: they're not discoverable.

A new user comes to the site. He decides to look for a C++ book. There are three things he could do:

  1. Search the site for "C++ book".
  2. Just ask a question recommending a good C++ book.
  3. Check the C++ tag's wiki.

Item #3 will never happen. Not for a new user. And why would it? A new user doesn't know what a "tag wiki" even is. If he searches for "C++ book", guess what he finds? Well, he doesn't find the definitive list, but he certainly doesn't find the tag wiki. Yes, there's a link to it right there, but he's not looking for it. He's looking for an actual post by actual people.

Before Tag Wiki's can take over this role, they must be discoverable. There has to be some way to actually find them with searches. If someone searches for a C++ book, then they should see a link to the C++ tag wiki that looks exactly like a question/answer. It needs to look just like any other post.

Therefore, what we need is the ability to have a tag wiki page that looks like, and is searchable like, a question/answer.

  • 3
    Post the part about making tag wikis indexed by the site search as a feature request – random Feb 21 '12 at 1:09
  • Just FYI, the tag wikis are indexed by Google. I'm seeing the Scala tag wiki as the 6th result for a search of "Scala Tutorial." If there's content there, people will find it. – Bill the Lizard Feb 21 '12 at 1:20
  • @random: Done. – Nicol Bolas Feb 21 '12 at 1:21
  • 2
    This. In fact, I would say, to leave most of the questions (especially the very valuable ones) locked but definitely undeleted, until it's demonstrated that that information is successfully findable elsewhere. I'm not sure how something like "recommended books" would work as a wiki -- even if it's not quite right for stack overflow, surely, surely, the reason it worked is both because (a) lots of people voted on suggestions, as well as (b) someone took a lot of effort to collate the useful answers?? – Jack V. Feb 23 '12 at 15:03

What if we moved: https://stackoverflow.com/questions/662956/most-useful-free-net-libraries to the tag wiki?

What kind of changes (if any) do we need to make to support that?

Git for beginners: The definitive practical guide is already mostly in https://stackoverflow.com/tags/git/info , we could move the rest

  • Ah, brilliant minds think alike. :) – Bill the Lizard Feb 12 '12 at 23:11
  • [status-completed] -- for the .NET Libraries question. – Robert Harvey Feb 12 '12 at 23:14
  • @RobertHarvey I can add "sub page" support if we need it, I want to hit the "tag wiki limits" first and then extend it to support what we need. – waffles Feb 12 '12 at 23:14
  • 8
    hmmm thinking ... we probably should have a closed/deleted as "belongs in tag wiki" that way we can redirect the traffic to the wiki. – waffles Feb 12 '12 at 23:16
  • So, are we seriously considering the policy that "big list" material is verboten on the main site, but OK in the tag wikis? I'm OK with that, I just never thought it would ever be considered proper. – Robert Harvey Feb 12 '12 at 23:40
  • 3
    @RobertHarvey I am not sure, I think big list fits way better in tag wiki format. In q&a it basically collapses. We probably will need to amend the tag wiki system to support big bodies of information, however I am totally open to that. – waffles Feb 12 '12 at 23:47
  • I would be ecstatic. That would solve a lot of problems. – Robert Harvey Feb 12 '12 at 23:49
  • 4
    If you put aside the pedantry of "well it's not technically a Q&A", big lists suck on SE because nobody wants to maintain or contribute a canonical list: you have to read all the piecemeal answers from people who couldn't bother reading the others or collaborating on one list. I don't see how "belongs in tag wiki" closures would effect a change in that mentality: who's going to take the time to go through all the answers and build the tag wiki? Do we leave the big list questions around until someone, some time in the future, copies all the information into a wiki? – user149432 Feb 13 '12 at 3:34
  • @MarkTrapp: I would be willing to do some of the work, but I can't do it all at once; it has to be done over time, and probably with the help of waffles' tag wiki enhancements. – Robert Harvey Feb 13 '12 at 6:24
  • 1
    @waffles: Can we write a way to convert posts (questions and answers) to sub-tag wikis (linked to from some specified parent tag wiki)? I have a fair idea that you use more or less the same mechanism for tag wikis as you do for regular posts. Why not have a mod function that converts these posts to hierarchically arranged sub-tag wikis? – Robert Harvey Feb 13 '12 at 6:27
  • 3
    Agree with Mark, and I don't want to see what some random tag wiki editor thinks is the most useful .NET library. Unless you're planning on taking Wikipedia head-on (bad idea and won't integrate well with the site, IMO) then you simply can't do a very good job at wikis. I'd like to see them continue to be quick summaries emphasizing the aspects relevant to their use on the site. – Matthew Read Feb 13 '12 at 16:50
  • Can you do something to increase the visibility of tag-wikis? I think a significant chunk of people don't know they exist. Even if it's just promoting the good, relevant ones on the RHS of questions near the ads. – Flexo Feb 13 '12 at 22:46

I've been using SO for over a year, and I have to say I have never googled a question, got sent here, and got help from one of the big subjective list questions. I don't think they belong here. Plenty of other sites handle that function just fine.

Keep SO clean in terms of its purpose, which is to host questions to which there are definitive answers. Clearly there is a fuzzy line somewhere, but many of the questions in the OP obviously cross it.

  • 9
    That's very anecdotal. Would it mean anything if I said I have gotten help from those big subjective list questions? Hidden features of C# helped, for one – Bob Feb 14 '12 at 19:45
  • 2
    Anecdotal or not, it is SO's stated purpose to be a repository of questions to which there are definitive answers. I'm just saying that my personal experience is in line with SO's stated purpose. – Almo Feb 14 '12 at 19:51
  • 1
    @Bob: well, yes it would help. If you were specific about it, told a story about how you found that question, the insights you gained from its answers, and how they made you a better developer... That would have made a pretty decent answer here. Of course, just saying, "IMHO, this is good / this is bad" with nothing to justify it doesn't add much. – Shog9 Feb 14 '12 at 19:58
  • 1
    Lists can be definitive too, though. They can be equally valuable, as well. Not sure I understand the point of a site with the most specific answers possible, as opposed to a site that tries to be the one place to find valuable information. – Bob Feb 14 '12 at 20:27
  • 1
    @Shog9 I found it by looking through SO's top questions (by vote). If I have a specific problem I Google it, and sometimes SO pops up, sometimes it doesn't. But looking through SO I expected to find some non-specific useful information, and did. Remove those and I just won't go to SO just to go and look around. The majority of specific questions are not helpful to me. – Bob Feb 14 '12 at 20:31
  • I come to SO for the answers to specific questions. General questions are easily answered elsewhere. – Almo Feb 14 '12 at 20:36
  • 2
    And others, like me, come for non-specific questions as well. Why is there not room for both? How are these list questions hampering your experience? – Bob Feb 14 '12 at 20:44
  • 2
    They add noise to the list of questions since they're all stuff that's officially "off-topic" for the site. – Almo Feb 14 '12 at 20:51
  • 4
    By definition, every question one is not interested in is noise, it doesn't mean they're bad questions. I'm not interested in the majority of very specific questions that get asked on Stack Overflow, even within tags I'm interested in. Does that make those questions bad? Does it hamper the experience of the whole site? No. I understand the powers that be (and the majority opinion on Meta) seems to be that they're off-topic, but I don't think they should be. – Bob Feb 22 '12 at 18:22
  • I downvoted this because, in response to the comment from @Bob, the answerer commented "Anecdotal or not ..."; you invalidated your own answer. – Kenny Evitt Jul 23 '13 at 17:37

Note that this is still a contentious issue, even among the mods. You could be a great help if you can articulate some guiding principles that could be applied fairly and equally to such questions.

I believe that those already exist:

Help us build a great library of canonical answers. If you keep seeing the same form of questions, whether it’s mod_rewrite rules on Server Fault, freezing computers on Super User, or how to use regular expressions to parse HTML, write a great, canonical answer, once and for all. Make it community wiki so that as many other people as possible can make it great. Work really hard on writing something that is clear, concise, and understandable by as wide an audience as possible.


I think that's all there is to it. Does the thread contribute to a "great library of canonical answers"?

That was the founding goal of SO, and I see no reason for it to change. And it can be applied to everything: questions, answers, long comment chains, anything.

We're programmers, we all suffer from some form of OCD. We want things to be nice and tidy and well-structured. We want to eliminate everything we can do without, or anything that can be inferred from the non-deleted content.

And that's the wrong attitude. It's a natural instinct for us to want to delete just because "it makes the site look untidy", and it irks us.

But really, read through the first year or so of posts on the SO blog, and we should have all the guidance we need: build a great library of canonical answers; duplicates are ok in moderation, the goal with SO is to be a useful resource for programmers, and make the internet a better place.

It is not intended to be a nicely "optimized" catalog containing exactly one of everything. It is not there to satisfy your OCD or mine.

That is the guiding principle. Or at least, it was. And I think it should still be.

  • 4
    I am a huge believer in canonical questions, but I don't believe that it's a fully-baked concept yet. How, for example, do you ask the question "The Definitive Python Book List" today, without the community closing it? More importantly, do we really need "The Definitive [subject] Book List" for every possible subject? How do you explain to the users that the C++ question is welcome, but their question is not? – Robert Harvey Feb 13 '12 at 18:47
  • Does it matter how you ask the question? Does it matter if it would be closed today? I think what matters is that it is the de facto canonical answer to a commonly asked question. It also has a huge number of inbound links, from other answers referring to it. And I think that if, say, the Python users feel the need for such a question, and organize themselves to (1) write a really really good one, and (2) use it and link to it, then that, too, should be allowed to stay. – jalf Feb 13 '12 at 18:52
  • 4
    The C++ one should not be allowed to stay just "because it's C++", or any special case like that. Simply because (1) it is relevant, and (2) it has proven itself. If I post a list of Ruby books, and no one ever refers to it, it's useless. If a group of Ruby programmers post a canonical list, maintain it, and refer to it, then it is useful – jalf Feb 13 '12 at 18:53
  • 1
    It matters if the question gets closed before you have time to add the canonical answer to it. How can a question prove itself if it doesn't get the chance to prove itself? – Robert Harvey Feb 13 '12 at 18:53
  • 4
    Also, I think the advocates for "Canonical Questions" mostly believe that canonical means, "A definitive answer to a commonly-asked question," not "An excuse to ask a big-list question with no definitive answer." – Robert Harvey Feb 13 '12 at 18:57
  • 1
    @RobertHarvey: tough luck. But even if other questions get deleted by the community, how does that justify deleting widely used, high quality information? If other similar questions don't get a chance to prove themselves, then that is the problem. The existence of the C++ question is not – jalf Feb 13 '12 at 20:23
  • 3
    None of the questions in this audit are such canonical questions. – Gilles Feb 14 '12 at 7:27
  • 1
    @Robert: "How can a question prove itself if it doesn't get the chance to prove itself?" That's a very good question. Have you considered this as an answer: "Allow the Python community to try to set up a book resource as good as the C++ folks did, and only consider deletion if they have failed after a few months." – sbi Mar 5 '12 at 10:26
  • @sbi: Try asking the question, and see what happens. – Robert Harvey Mar 5 '12 at 16:09
  • @Robert: I will most certainly not ask any Python question. And will you please answer my question? – sbi Mar 5 '12 at 19:06
  • @sbi: Canonical questions are not well-supported by the system, are not well-received by the community at present (how do you keep them from being closed?), and generally not well-maintained once they are asked, although there are exceptions. I personally think Tag Wikis are the best place for these kinds of resources, but that is the topic of an ongoing discussion. See also meta.stackexchange.com/a/122589/102937 – Robert Harvey Mar 5 '12 at 19:15
  • 2
    @Robert: Looking at the C++ book list question, I think that, while improvements would be welcome. such stuff is supported (and maintained!) well enough currently. And, "...are not well received by the community?" LOL. What's very obviously not well received by the community is that these questions aren't well received by the meta crowd. Just look at the very first sentence in the question you linked to: "I refer to moderators closing posts based on a strict interpretation of policy, regardless of their value to the community." The community values these posts, and laments their deletion. – sbi Mar 5 '12 at 19:34
  • 1
    @sbi: As I said, there are exceptions. The Meta crowd is the Stack Overflow community. It's not our fault you decided to disengage from it. One of the reasons I suggested you post the Python question is to see what happens to it; it would most likely get closed. Not by the Meta despots. Not by the moderator despots. By the community. – Robert Harvey Mar 5 '12 at 19:35
  • 1
    @Robert: Open your eyes man! Those "exceptions" are coming up by the dozen every each day on meta here. Your users are up and in arms about your policy. And you call that "exceptions"? Those questions are well-received by the users, and well-enough supported technically. What's not well-received is how you deal with them. Really, when everyone is driving the in the wrong lane, Robert, it might just be the moment to check whether you yourself are in the right lane to begin with, as awkward as that might seem. – sbi Mar 5 '12 at 19:39
  • @sbi: I have no idea what you are saying. It's not my policy. It's the policy established by the community. Moderators do not set policy, they only enforce it. – Robert Harvey Mar 5 '12 at 19:40

I think https://stackoverflow.com/questions/224867/what-programming-language-do-you-wish-would-quietly-retire should be added to the list for the audit (the question itself isn't highly up-voted, but some of its answers are over 100), and it should be removed. It's completely subjective and has no place here. It's inverse was removed for reasons of moderation, why not this one?

  • 4
    Given that its counterpart: What Programming Language do You Wish would Catch On? was deleted by 14 votes, I will retire this one also. – Robert Harvey Feb 14 '12 at 3:52
  • 2
    Along the same lines New Programming Jargon you Coined should be deleted...I fail to see its place on the site. – user7116 Feb 14 '12 at 16:00

These sites are not set up for list maintenance. Not in questions, and not in tag-wikis. As soon as a list exists it's obsolete. Using answer posts to contain list entries is cumbersome and prevents organization, subheads, and sorting. Tag wikis aren't set up for this job.

Let wikipedia own that problem, they've got the culture and the facilities. The 'canonical answer' to a list question can be a link to the wikipedia page.

In other words, 'kill kill' (pussycat) 'kill kill'.

  • "These sites are not set up for list maintenance." So you propose deleting everything to which answers with more than two items would make sense, because those would be list answers, right? – sbi Mar 5 '12 at 10:23
  • I don't propose anything. I argue in support of a long-standing policy against long-lived laundry-lists. – Rosinante Mar 5 '12 at 13:25
  • Well, it might well be that you do not want to propose that, but as it stands, I read your statement as condemning questions that can/must be answered with a list. Since we are talking SO here, a forum for programmers, who are used to having to express everything exactly and in detail, would you be so kind as to clarify that point? Thank you. – sbi Mar 5 '12 at 19:09
  • @sbi, none of this discussion is about a question which could get a single right answer with a list in it. It is about 'poll questions' which attempt to maintain a giant list in the form of all of their answers. That's what 'list maintenance' means. I feel no need to edit this answer given the many discussions here on meta about the subject. – Rosinante Mar 5 '12 at 20:49
  • Please bear with me, I haven't been a meta regular recently, and I'm still struggling to get those definitions right: What makes a question require "list maintenance" opposed to requiring an answer with more than two items on it? And is everyone of those "list maintenance" questions automatically also a "poll question"? – sbi Mar 5 '12 at 21:19
  • Classic example: What are the best books about c++? Each book gets an answer. As new books are written, if they don't go onto the list, the list becomes useless. It's all about the opinons of the people writing the answers. – Rosinante Mar 5 '12 at 21:20
  • That might be a bad example, because 90% of the C++ books out there are indeed utter crap, and the remaining 10% are well worth listing. But I wasn't after examples anyway, because it's easy to come up with an example supporting your (or mine, see above) POV. What I wanted is a definition that helps me understand how you sort questions into either of the categories you mentioned. We can then use the definition to sort questions into categories and see whether we'd both agree with the sorting. – sbi Mar 5 '12 at 22:09

I firmly believe that What is the single most influential book every programmer should read? should not be deleted. This list is not just a Google or Amazon search. It is a collective list of books and work that the greatest programming community on the face of the planet has deemed to be the most influential in their careers. Where else are you going to find books where some of the greatest work in that field is plainly listed, in one place, for anybody to peruse and look through? It might not fit but there's enough knowledge for me to overlook; this question should definitely stay.

I've already archived the top 100 answers to the question but I really don't think this answer should be removed. I know as a beginner it proved an invaluable resource for great books. I don't want to take that away from the next guy.


Spontanteously deleting questions that have been there for a while, means links get broken. This makes the Internet worse, can lead to frustration, and therefore broken keyboards as people smash them in anger, annoyed that they one piece of information they need crucially is no longer available.

The goal of Stack Exchange is to make the internet a better place, not to make the Stack Exchange network squeaky clean.

These questions aren't cluttering Stack Overflow, they sit at the bottom of the questions list and very very rarely surface.

Deleting useful information is wrong, archive/move/close to your hearts content, but don't delete. If the Internet was a religion, this would be a sin.


If you want a simple policy:

Delete them all

This applies to any other non-constructive question that's not in your list. And don't lock them!

If you're feeling like making fine distinctions, ok, I'll go and review each of your examples.

That's 13 deletes and 1 keep (plus one borderline question that isn't non-constructive). Yeah, when in doubt, just delete.

N.B. To clarify: when I say “delete”, I mean remove from the questions and answers part of Stack Overflow. The material doesn't have to be deleted from the face of the web. Some of it belongs in tag wikis (and I welcome improving support for large tag wikis). And if someone wants to copy the threads to another site, they have my blessing. Deletions of large threads should come with prior notice (spending weeks closed or locked counts as prior notice, since as far as I know no one's requested anything else).

  • +1 I like that you have no second thoughts about the vim post :) – Lix Feb 12 '12 at 23:35
  • 6
    well, delete, or move to tag wiki when possible. – Jeff Atwood Feb 13 '12 at 1:22
  • @JeffAtwood You're right, I should have made that clear. Edited. – Gilles Feb 13 '12 at 1:29
  • 17
    Two points. #1: the question is not about finding "a simple policy", it is about finding "the right policy". Thus, you don't answer the question. #2: "when in doubt, just delete" is probably the *single most harmful suggestion I have EVER heard regarding SO. Yes, it's very much in spirit with the meta police, and the idea that it doesn't matter what actual content you contribute to SO, *as long as you get to influence the direction of the site and the rules as much as possible". Problem is, there are still a few of us who think SO is more important than Meta. – jalf Feb 13 '12 at 17:36
  • 8
    Don't delete anything and move on with your life. – Tony The Lion Feb 13 '12 at 18:14
  • 8
    This answer is complete bull****. "What is your most productive shortcut with Vim? — the question is not constructive, but the top answer is pretty good." I'm sorry, but this clearly illustrates that it doesn't matter what's on topic or not. Everybody (including the bods) wants to keep the answers they refer to. I don't have enough rep on meta to downvote, but consider this a -1 from me. – André Caron Feb 13 '12 at 20:25
  • 3
    @André I did it for you – systempuntoout Feb 13 '12 at 22:58
  • 5
    @Jalf: Your first point is inane; Gilles is clearly saying that the simple policy he outlined is the right policy, so this does answer the question. Your second point loses any shred of legitimacy it may have had by its conspiracy-theoretical invocation of "the meta police". Those in favor of deletion care just as much about SO as you do; they simply have a different opinion of what's best for it. That's why this open discussion that Robert prompted is happening in the first place. – Josh Caswell Feb 14 '12 at 2:52
  • 4
    @JoshCaswell: If you're offended by the term "meta police", have you considered removing the cause for users to think of them that way, rather than blaming them for it? And I seriously fail to see "there are thousands like these, and this is one of the few maintained ones ==> delete" as an expression of care for SO. It's mindlessly suggesting to wipe out other users' work, in violence of the site's spirit, just because some algorithm spits out false, and despite the diagnosis that the algorithm fails ("one of the few"). I'd definietely call this "careless", not "carefull". – sbi Feb 14 '12 at 10:43
  • 4
    @sbi I suggest you read the full sentence. “Delete unless someone speaks up to defend that specific question”, which you did. I still don't think an SO question is the best place for a C++ book list, but if there are people actively maintaining it, it can stay in my book. I still push for deletion as the default policy because most such book lists are not maintained and end up an obsolete jumble of randomly selected books which don't help anyone. – Gilles Feb 14 '12 at 15:59
  • 6
    @Gilles: jalf wrote, '"when in doubt, just delete" is probably the *single most harmful suggestion I have EVER heard regarding SO', and I fully agree with that. This site's goal is to provide information — how could deleting information ever become the default here? I'm all for cleaning up, and do not hesitate to wipe useless crap off the site, but once a question has accumulated a certain amount of information, let alone after you yourself found it's actively maintained, how could deleting be the default? Sorry, but that's Just Plain Wrong™. – sbi Feb 14 '12 at 18:59
  • 1
    @sbi "When in doubt, just delete" applies to content that is manifestly not within SE's scope. What I oppose is keeping closed questions around forever (duplicates excepted), or refraining from closing questions just because they're popular. – Gilles Feb 14 '12 at 22:32
  • 1
    @Gilles That just doesn't make sense. If something is manifestly not within SE's scope, then by definition there is no doubt. – Bill the Lizard Feb 15 '12 at 13:24
  • 5
    @Gilles No kidding. Your comment still makes no sense. "When in doubt" can't apply to content that is "manifestly not within SE's scope" as you claim. "When in doubt, just delete" is a terrible policy proposal. Moderators should not be taking action on content until they're sure of what that action should be. – Bill the Lizard Feb 15 '12 at 15:34
  • 3
    @Gilles: "by default, content which has been established not to belong here should be deleted." Deleting content by default doesn't give anyone a chance to justify exceptions, or to move the content to a more appropriate place. – Bill the Lizard Feb 15 '12 at 19:28

I just try to explain why I think those very good answers are not meant to be in a tag wiki.

Imagine this scenario which everybody of us has once experienced:

You are in a hurry trying to find something on the Internet, the very huge Internet. What everybody does, is to use Google. Now assume, I want to buy a good C++ book and enter into Google, "best C++ book". What do I get back? 1 billion hits, which of course I can't read all. So, what everybody does, is just to click through some of them, mostly not spending more than a second on every site. Now, if you click on a link, called C++ tag wiki, where you have loads of text and a lot of other stuff you will be hitting return like immediately.

But then, there is this link called "The Definitive C++ Book Guide and List", you think "that must be it", you click it and you see a well-structured hitting headline which catches you immediately, a list, sorted for different entry levels, then you will get stuck and read. It is the same as with newspapers, you need a headline, which catches you and keeps you more than that critical second.

So hiding such very good topics in a tag wiki somewhere, where nobody is ever going to find to, would be a very bad idea. Deleting it, is even worse.

  • 1
    Do you have any ideas on how to make tag wikis more read, more visible. Or no? – random Feb 13 '12 at 17:58
  • 1
    @random no, but I think that the tag wikis are just not meant to be the place for great knowledge or FAQ like posts, just like on wikipedia, there are no pages explaining how to use git right or which c++ book to buy, but more likely to give you a general overview of what git or c++ is and maybe add some links to valueable posts/links. In my opinion,everything is fine as it is, because it works. – inf Feb 13 '12 at 18:06
  • @bamboon I have already posted a question regarding these questions which seems off topic to us but attract a lot of users to the community. What a tag wiki should hold should be given in the FAQ. I have already proposed that here – MozenRath Feb 13 '12 at 18:41
  • Except that it's isn't definitive. It's just some opinions of whomever felt like answering, voted here and their by their various fans. The Q&A format doesn't do this well. – Rosinante Mar 5 '12 at 3:19
  • @Rosinante You have to decide, stay conservative and drive the ship into the iceberg or change direction and keep on being a great source of knowledge. – inf Mar 5 '12 at 8:42

Due to all the down votes, I'll try to say what I want to say in different way.

Not every question has specific answers, but that does not automatically make the question a discussion, unanswerable, or whatever else is deemed negative around here. Yes, it means the question is broad and open, but is that not the point of a community-wiki?

It seems like the focus shifted so heavily towards specific questions with very specific answers that community-wiki fell out of favor. But I feel like the spirit of the site was lost in the process. At least, what I always took the spirit of the site to be. I thought these sites were about having a place where good answers to questions were to be found. Making the internet a better place. Cutting broad questions simply because they're broad seems antagonistic to that spirit.

Questions about books, hidden features, data structures, etc. contain useful information that's updated through the wiki process. Containing value information, I feel, is all the reason needed to keep a question around. Why debate about deletion? Moving the question? Why this discussion at all when it's a simple fix: change the way of thinking to accept that broader questions can be important as well.

  • There's nothing at all wrong with wiki-style Q&A; it's just not the raison d'être of Stack Overflow. – Robert Harvey Feb 14 '12 at 18:19
  • 3
    At one point I thought I understood Stack Overflow's raison d'être to be to make the internet better by having a centralized place for answers. What I don't understand is how a "list" and wiki answers are somehow a detriment to that goal. Very specific answers are valuable, but that doesn't mean broader questions and answers aren't or can't be. – Bob Feb 14 '12 at 18:25
  • 5
    So ironic how popular those questions are on Stackoverflow, yet how popular killing them all is on Meta – Bob Feb 14 '12 at 19:47
  • 1
    @RobertHarvey: Prove that. – sbi Mar 5 '12 at 10:28
  • 1
    @sbi: It's in the FAQ, and numerous blog entries and Meta posts. All ya gotta do is read. – Robert Harvey Mar 5 '12 at 16:08
  • 1
    @Robert: Please prove that wiki-style Q&A are not wanted on SO by pointing me to the appropriate blog entries and meta postings. – sbi Mar 5 '12 at 19:05
  • @sbi: Sorry, but I'm not going to do your research for you. That information is easily and readily available, if you have any interest in reading it. – Robert Harvey Mar 5 '12 at 19:48
  • 1
    @Robert: No, it's not. You're wrong. – sbi Mar 5 '12 at 20:06
  • @sbi: If I point you to all the material, you're just going to reinterpret it for your own purposes. Sorry, but I'm not willing to play that game. – Robert Harvey Mar 5 '12 at 20:08
  • 1
    @Robert: You mean that "material" can be interpreted? Well, that certainly would explain why you refuse to link to it. Thanks for making my point. – sbi Mar 5 '12 at 21:15
  • 2
    @sbi: See? You are a master at this. I bow down to you; I am not worthy. – Robert Harvey Mar 5 '12 at 21:16

Let me preface this by saying that anything that I indicate should stay, doesn't mean that I think it is appropriate for Stack Overflow (unless mentioned otherwise) and means that it should stay, for now.

Should stay for as long as people are putting effort into maintaining it. The first answer is well-maintained, the rest of the answers could use some cle.swer

(Note: The discussion around this is the one that sparked all of this in the first place)

It's purely subjective. While it has a list of a lot of great books, it's like asking which shade of blue is the best (which we all know is cerulean, BTW).

When the highest voted answer (which beats out the next highest-voted answer by 150+ votes) is not even directly related to C# (but rather, .NET), it's not a good sign.

Additionally, this was last edited in July of 2010, and the language has evolved since then (which begs the question, why hasn't the main post been updated with new answers?).

And why is this question is tagged when it's about C#?

It was last updated recently (December of 2011) but has 200+ answers. There should be one answer that is accepted which has the features added in. If not, delete.

Please try to include links to pages describing the data structures in more detail. Also, try to add a couple of words on why a data structure is cool

This question is a popularity contest to see who can dig up the most esoteric data structure:

WTF is "cool"?

I'm Cool Like Dat.

Undoubtedly, this caused a number of bad, highly-voted answers which are not really answers and don't indicate any of the practical applications of such structures.

As I go through this list of questions from top to bottom this is the best example of something useful so far. Only 86 (which is still a problem) answers and content is edited into the main question (although I dislike the use of the term "definitive", I'd change that).

As a .NET person, it pains me to say so, but there are examples of libraries that aren't as useful as they once were (case in point, the compression libraries that deal with zips are generally not as useful since the System.IO.Compressions namespace came about with .NET 2.0, although those libraries do still have some merit). If the list was better maintained, then possibly keep.

Please, if you must keep it, remove "most useful" from the title.

This is structured more like a FAQ, and is invaluable, IMO. Doesn't try to be "the best" or "definitive", just structured and helpful. Not fun either. I couldn't find something wrong with this if I tried.

This question would be deleted immediately if posted today. Additionally, who the hell is writing compilers on a day-to-day basis (not many, I'd wager)? I'm not saying people don't but the audience that appeals to in a truly beneficial way is small (and could be considered "Too Localized" on top of that.

Additionally, it's not about a single language compiler, it's about any compiler (although it mentions preferred languages). Basically, it's about listing every single book that was written about compiler resources ever. Delete as "go to Amazon".

The accepted answer is some TL;DR meta-manifesto which obfuscates a tremendous amount of useful information, and beats out the next highest-voted answer by 1200+ votes. (the animated gifs on the second-highest voted answer also annoy the crap out of me as well).

This is just a bunch of people trying to brag. "Yo, did you see how I task-cancelled that bug report today?".

And why is it tagged with ? Why does that tag even exist?!?

Barely though (there's a delete vote on it now)

There's definitely valuable information here, but like the VIM question above, how do you wade through the first answer, let alone all the answers?

PHP is not a good language (deal with it), just like no other language is a good language. It's about what you (and I mean a single individual, not everyone) can do with it.

Some useful information, but it's by no means applicable in a broad sense (and it's language-agnostic, which worsens the situation).

That all said, remember that deletions are not permanent and if a 10K user (or mod, or employee with a diamond) is willing to make the information available to someone so they can host it somewhere else, then do it with our blessing (just remember to attribute properly).

  • 1
    So if "lesser-known but cool data structures" were re-titled to remove the offensive "cool" word, it could stay? – Robert Harvey Feb 12 '12 at 23:43
  • 4
    The massive VIM answer should be converted to a blog post on the "community" blog.stackoverflow.com (if and when it exists) – waffles Feb 12 '12 at 23:50
  • 2
    @RobertHarvey No, I've updated my response to reflect why. Sorry for being so sparse on that. Not sorry because you made me spend so much time on the answer to begin with =) – casperOne Feb 13 '12 at 1:10
  • 2
    @waffles Yeah, how's that coming? =) – casperOne Feb 13 '12 at 1:10
  • 1
    @waffles, hey, delete and post to blog would be a pretty good way of 'getting it out of the system'... – Benjol Feb 13 '12 at 6:46
  • 1
    Don't drop the definitive from the C++ books post: if you do so, it won't convey the intent of the question any more; It is not supposed to be a discussion forum, it is intended to be the End-All place to go for C++ book recommendations. Don't water it down, or it will become what you expect. – sehe Feb 13 '12 at 16:24
  • @sehe That's a prediction which doesn't really have a basis in anything... But I respect the opinion. I think that other terminology could be used which satisfies both groups. – casperOne Feb 13 '12 at 16:28
  • We would welcome some questions as blogs in our newly born P.SE community blog, feel free to drop by our chat room if interested. /cc: @waffles – yannis Feb 13 '12 at 16:58
  • @casperOne: it's not a prediction, it's a statement of the de facto reality. If you want to break a popular, useful and highly popular question, go for it, you're the mod. But really.... ' – jalf Feb 13 '12 at 22:41
  • 3
    @jalf You can't swing around "de facto" to magically justify any argument you have no evidence to support. On top of that, the C++ question is the one question I said to keep, so someone is obviously not reading the post (and note that it's been that way since the first revision). – casperOne Feb 13 '12 at 22:46
  • I fixed that for you and removed ".net" from the tag list of that c# question. – Johannes Schaub - litb Mar 4 '12 at 13:27

Re Hidden .NET Base Class Library Classes

I don't think it should be deleted, but its duplicate answers with lesser votes should be.


Delete them all, surely. but compile all the answers to the tag wikis(this is a much bigger task actually). just adding the content to the tag wikis on an as is basis is not gonna help. So the material needs to be segregated and compiled properly into the tag wikis.

The above does not apply to: What is the single most influential book every programmer should read? just close your eyes and delete that one


I for one would not bother deleting any of them, sure they are not good questions going by the standards, but who cares? People like them and we get boatloads of crap every single day which may never be deleted that noone likes, so it's a waste of resources to fuss over these celebrities in my opinion.

  • 1
    lol I'm only commenting since I see you struggling daily to wade through tons of crap questions in the WPF tag, and fully understand your opinion of this. Oh and +1 because I agree we should use moderator resources to focus on cleaning bad content instead of good content. The question already says it's not a good on-topic question for the site, so I don't think you can argue that it promotes bad questions. – Rachel Feb 24 '12 at 1:43

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