Currently the on-topic page reads:

and it is not about...

  • general workplace issues, office politics, and job hunting (check out The Workplace instead)
  • implementation issues, such as code fixes (ask on Stack Overflow instead)
  • how to use specific tools
  • what language/technology you should learn next, including which technology is better
  • what project you should do next
  • where to find a software library, tool or other resource
  • product or service recommendations
  • career or education advice, salary, or compensation
  • personal lifestyle, including relationships and non-programming activities
  • legal advice or aid

the "where to find a software library, tool or other resource" line could be expanded to include tutorials and books explicitly.

I know they are covered under the "other resource" line but it's better to have it explicit. Maybe link to the meta discussion about it.

  • 2
    This site was created specifically to push useful but too general and/or experience-driven matters off stackoverflow, this was one of them. Now it's getting the boot here as well? Where is it going next then? Commented Jul 17, 2015 at 20:30
  • 3
    @ivan_pozdeev finding a tutorial or book has already been off-topic for a long time. I'm just proposing to make it explicit in the on-topic page. Commented Jul 17, 2015 at 20:34
  • Generally speaking, these questions will never stop coming. Getting a overview of a technological field (including the seminal reading on it) is a real and highly important task in the technology world. All we can do is invent a way to deal with them that is most consistent with SE's goal of sharing knowledge. Commented Jul 17, 2015 at 20:39
  • @ivan_pozdeev We have a canonical (locked) question for that for most languages. Commented Jul 17, 2015 at 20:50
  • If the "canonical answer" you mean is meta.programmers.stackexchange.com/questions/6483/… , it needs to have its phrasing updated, too. Commented Jul 17, 2015 at 20:54
  • Your q129641 link illustrates a completely opposite point to what you're preaching - that there can exist perfectly good recommendation Q&A's here. Sure, there are existing ones "for most languages", but there's always another field (including an updated version of the same field). Commented Jul 17, 2015 at 21:09
  • 3
    @ratchetfreak there's also some movement to try to get these book answers into the tag wiki. There's Is there a canonical book on Agile?, but agile tag wiki also contains that information, in a cleaner format (in that its not suggesting asking more such questions).
    – user40980
    Commented Jul 17, 2015 at 21:11
  • 1
    @ivan_pozdeev what you don't see there are the four deleted answers, including one that is in its entirety "After K&R, get Harbison & Steele. After that it's just man pages." -- we've had a very troubled history of trying to do book suggestions and we haven't seen any indication that people will ask them well, nor answer them well, nor moderate them well since then.
    – user40980
    Commented Jul 17, 2015 at 21:12
  • @ivan_pozdeev if you have points against put it in an answer Commented Jul 17, 2015 at 21:15
  • Well, how's this? Not bad, eh? *bursting with pride* Commented Jul 18, 2015 at 1:35
  • I would consider waiting for a while before deciding on this. We're getting some spike in books garbage questions recently and it's indeed worrying. However, it may be sort of side effect of STCI edits and retags which bump old questions to front page and may mislead some inexperienced visitors into thinking it's still okay. I'd wait for a while until STCI is completed and re-check if books questions still come at alarming rate or not
    – gnat
    Commented Jul 18, 2015 at 9:41
  • Stack Overflow already has this change. Commented Jul 28, 2015 at 16:54

4 Answers 4


The line now reads:

where to find a software library, tool, book, research paper, blog, forum, or other resources

The link goes to the same meta post as before.


I agree this change should be made.

  • Questions asking for book recommendations are already off-topic.

  • This is a very minor change in wording that does not change the intent or make the bullet point too verbose (it would be madness to enumerate all off-topic "resources").


Objective evidence provided in the initial discussion does show that "book recommendation" Q&A's as they are now are generally sub-par and unhelpful in the long or even the short run.

  • They tend to steer towards the "bad" end of Good subjective, Bad subjective
  • A question of just "recommending a book on a subject" is fundamentally unclear: any subject has any number of sub-subjects, and each book covers them in different proportion and relative quality
  • Books tend to become obsolete (the faster the more specific matters they cover), and the current SE model doesn't provide an endorsed-by-the-system way to keep the info updated
  • Answers to such questions are not self-sufficient: they don't actually address the asker's need (to get to know the subject), they only provide a reference that may or may not do this. Conversely, the basic SE's principle is to make an answer actually address the problem, only providing references for more information on some of its aspects.
    • Here, I can draw an analogy with meta-tags: they also are not self-sufficient and do not convey facts, but opinions
  • That's it for the problems. Shall I update it with possible remedies as well or it's better to post another answer to avoid making the discussion unwieldy? Commented Jul 17, 2015 at 22:05
  • 1
    The 'endorsed by the system' bit - while we don't have a clear way of marking something obsolete in an answer, every tag has its wiki (known as a tag wiki) which are often underutilized. However, such a place would likely be an excellent option for storing book lists and other related information. There's a bit of a chicken and egg problem there - no one is using them so we don't have many tools or much visibility into them so people are not using them. Look at the tag wiki for scala on Stack Overflow to see what can be done though.
    – user40980
    Commented Jul 17, 2015 at 22:06
  • @MichaelT I stand by the fact wikis aren't "a way endorsed by the system" as they are now. Users don't get much (if any) rep keeping them up-to-date, nor is the long (or even the short) description highly visible, even less so to potential askers. Commented Jul 17, 2015 at 22:09
  • Users less than 20k get 2 rep for an approved edit on the tag wiki. The alternative is a community-wiki and lock on a Q&A - which no one gets any rep for. The visibility aspect is very true, though that goes back to 'its been ignored both by SE and Users'.
    – user40980
    Commented Jul 17, 2015 at 22:15
  • @MichaelT What is 2 rep compared to a few dozens (if not hundreds) of rep and growing? The (in)famous Jon Skeet gets 200 rep daily just from upvotes (and just 200 because it's the cap). Commented Jul 17, 2015 at 22:20
  • By the way, a wiki is conceptually the same as a canonical answer - as the process of maintaining it: hard and ungrateful. Commented Jul 17, 2015 at 22:25
  • 1
    book questions (such as the agile one most recently), are best served by wiki locks rather than dozens of answers. If they have dozens of answers, its a historical locked question and no one can vote on anything on it. If it is a wikilock, the answer is a community wiki and no one gets any rep for editing it or votes. Two rep is more than zero that other book answers get.
    – user40980
    Commented Jul 17, 2015 at 22:28
  • dunno what's a "wiki lock", nor can find anything. could you add a hyperlink? Commented Jul 17, 2015 at 22:30
  • see New lock type for truly collaborative questions and answers
    – gnat
    Commented Jul 17, 2015 at 22:39

Post 7499 outlines the problems the questions cause. Despite this,

  • they reflect a real and highly important need of the technical crowd, so they will never stop coming
    • thus, unless we want to alienate users who have those needs and make them look for answers elsewhere, we must invent a way to address those needs, somehow, in a manner consistent with SE's mission to disseminate knowledge, here or elsewhere

This post outlines the possible remedies suggested.

  • Enforce the "Good subjective" principle: require an asker to specify what exact problem they need to solve with the information from a book (get an overview, compare, learn how to complete a specific task(s)), and an answerer to specify how exactly that book will address those needs.
    • Fixes 1)
    • partially fixes 2): a degree of uncleanness remains in any case
  • Maintain topic overview articles (a wiki page/canonical answer) containing book references, among other things
    • intended to fix 3) but a solution to make it "endorsed by the system" and visible to potential askers is most unclear.
    • is part or a larger problem of users getting little to no rep for updating existing info. This matter warrants a separate discussion
    • canonical answers are used successfully at Server Fault.
      • some do recommend books
      • in their case, it helps that enterprise sysadmins deal with a rather limited set of topics
  • Find another "sewer drain" to offload those questions into, like general questions were once drained here from StackOverflow

Note that nothing will fix 4) as long as the current "an answer solves the problem by itself" paradigm stays unshaken. Experience from the communities like Travel.SE that somehow managed to cope with matters that are by their nature incomplete and subjective, may help.

  • So, in the descending amount of realisticality IMO: boot to SRec, enforce quality, compile & maintain a DB of canonical answers. (wikis are even lower on the list due to their low visibility.) This is about the solutions that don't require additional accomplishments. Commented Jul 18, 2015 at 1:28
  • About the "boot" option: they will have to deal with the quality right away and intensively but, as the MSE thread says, this will fracture the community. Commented Jul 18, 2015 at 1:50
  • 2
    This post reads like it is trying to justify allowing book requests, when the question specifically states they are already off-topic and asks whether we should update the language in the help center to be more explicit.
    – user22815
    Commented Jul 18, 2015 at 15:13
  • you seem to believe that getting stream of questions is important. Consider that this is indeed not so. See Optimizing For Pearls, Not Sand: "We feel that the world is awash in questions..." etc
    – gnat
    Commented Jul 18, 2015 at 18:57
  • @gnat I believe that addressing the needs of the technical crowd is important, whether this means a stream of questions or anything else. I couldn't care less of how this is manifested and have analyzed every solution that was suggested so far. Commented Jul 18, 2015 at 19:10
  • it is but of a lesser priority than you probably think. That Stack Exchange blog article referred in prior comment, it explains why they had to compromise and cut off some stuff no matter how much technical crowd needs it: "the only logical thing to do is to maximize the happiness and enjoyment of answerers. If this means aggressively downvoting or closing unworthy and uninteresting questions, so be it. Without a community of people willing to answer questions, it really doesn't matter if there are questions at all, does it?"
    – gnat
    Commented Jul 18, 2015 at 19:14
  • @Snowman yes, my first post and your answer pretty much summarize the extent of the initial inquiry. I shamelessly used this discussion as an excuse to analyze the underlying problem in more depth and maybe work out a way to tackle it. Commented Jul 18, 2015 at 19:21
  • @gnat When did I say or imply the interests of the answerers (=the frequenters) need not to be considered in resolving this matter (or anywhere else)? On the contrary: e.g. the "overview articles" solution doesn't work in part because it doesn't address them. Commented Jul 18, 2015 at 19:26
  • per my reading, you said it right here, pointing "the needs of the technical crowd" while ignoring site regulars. Let's put it that way, if "crowd" wants book requests and answerers don't want these here, whose needs "win"?
    – gnat
    Commented Jul 18, 2015 at 19:38
  • @gnat No, the fact is, answerers want them here, too (the "and getting answers!" phrase and example link)! They just agree that the present procedures don't allow to handle this matter effectively, so a question needs to be removed to avoid cluttering the site - after the asker had a chance to read their message. Maybe I didn't emphasize this enough - in which case, you're free to improve the post. Commented Jul 18, 2015 at 19:47
  • wait a minute. Please answer the question I asked: if answerers don't want these, whose needs win? If you want, we can return to the matters of what answerers really want later, let's first prioritize whose needs are higher priority, crowd or answerers? (sorry if this sounds pedantic but I only want to see if thre is a common ground to discuss this matter further)
    – gnat
    Commented Jul 18, 2015 at 19:51
  • @gnat I'm not answering hypothetical questions. Askers' needs are important, too, and if they conflict, this must be handled on a case-by-case basis. The only thing that is absolute here is the network's goal - to share knowledge. Commented Jul 18, 2015 at 19:53
  • case-by-case basis, I see. Consider re-checking that SE blog article, you may find it difficult to discuss meta topics without this. Another (related, but bit more general and fundamental) article worth studying is A Group Is Its Own Worst Enemy
    – gnat
    Commented Jul 18, 2015 at 19:57
  • @gnat The post effectively says that the balance they need to maintain has become upset in the askers' favor, and they're fixing this. Shall I find another post that documents a swell in the opposite direction? Commented Jul 18, 2015 at 20:03
  • another authoritative post - yes please (random meta comments and posts don't count)
    – gnat
    Commented Jul 18, 2015 at 20:08

You must log in to answer this question.

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