-7

I know there's plenty of hardcore users here who despise open-ended questions more than anything else on the planet, but aren't we going a little too far sometimes? There are times when asking the community for advice could be a very powerful thing. Indeed, by not allowing at least some open-ended questions we're shutting off a potentially very potent resource... our group knowledge.

We entrust "open-ended" questions to decide who moderates the Stack sites, so why not allow certain open ended questions?

Consider the following questions:

  • Which IDE is best for coding in PHP?
  • Which books are recommended for learning C++ basics?
  • Which graphics application is recommended for web sites?

Yes, there may be no definitive answers to these questions, but the collective community knowledge, and the reputation of those who answer, can help shine a light and point people in the right direction... Provided there is only one question (no duplicates) and that the question is periodically re-asked, so that new developments get integrated fairly.

I understand the problems with "open-ended" questions, but couldn't we, as a community, be able to find a way to overcome them?

  • All of your examples are recommendation questions. Are you using "open ended" as a euphemism for recommendations, or is it just bad choice of examples? – Daniel Beck Jan 23 '12 at 18:35
  • I'm not entirely sure what you're referring to with "euphemism"? Are you suggesting "recommendations" are something I'm ashamed of, or that might offend people? I really have no idea what you're talking about here. It may be a bad choice of examples, I guess, but I have absolutely no idea what you're getting at. – Django Reinhardt Jan 24 '12 at 13:30
  • Sorry, bad choice of words. But recommendation questions, which is what your examples are, have been discussed many times before, and were dismissed even by the site owners AFAICT. So I'm asking whether you're asking about recommendation questions, but simply don't call them that (deliberately – i.e. trying to avoid the stigma – or accidentally), or whether your "open-ended questions" refers to something else and you just picked poor examples. It would be helpful if you could provide other examples of the kind of questions you want to discuss here, that aren't asking for a recommendation. – Daniel Beck Jan 24 '12 at 13:53
  • Ah, I see. No, I wasn't aware of that. I really don't see this going anywhere, which is a shame because, as I say, it could. – Django Reinhardt Jan 24 '12 at 14:38
  • For example, this could be considered open-ended. That's why I was asking. But if it's really only recommendations, I really don't think this site will work for this. Some of it is the format, it's difficult to get comparable posts, and some of it is that the product recommendation section will dwarf all of the other content on the site. – Daniel Beck Jan 24 '12 at 14:59
  • That's a good example, actually. I wonder if there could be a "discussion.stackoverflow.com" or something similar that was specifically designed for open-ended questions (both recommendations and any other kind). As I say, I just feel it's a shame for this community's collective wisdom not to be used to its fullest potential. And at the moment there isn't an "open question" equivalent to SE sites that crowd sources answers to such a large degree. – Django Reinhardt Jan 24 '12 at 15:19
  • Something more like Programmers.SE perhaps? – Daniel Beck Jan 24 '12 at 16:24
  • Yeah, I guess so. Although ironically you're not allowed to ask about programming tools there, so although the discussion is "open ended" (it seems) you still can't ask for a recommendation on something. – Django Reinhardt Jan 24 '12 at 16:34
12

All of your examples are:

  • off topic (clear shopping recommendation without any interesting aspects, e.g. exotic requirements),
  • not a real question, since they're just too broad, and
  • not constructive as the answers are highly subjective.

Let's ignore these categories for a moment, and see why the questions aren't useful.


Consider the book recommendation (ignoring for a moment that it's off topic on SU anyway). There was a similar topic on Stack Overflow: What is the single most influential book every programmer should read?

Let's ignore for a moment that it was finally closed. What do you see?

A popular, "open-ended" topic that managed to attract a lot of attention. So much attention, in fact, that a single, one-line answer could give users editing permissions on the site. A topic where users stopped answering the actual question, but instead added yet another book I liked to the list and so completely broke the entire topic.

When I first started, there was "Mastering Turbo Pascal" by Tom Swan. There is nothing terribly profound about this book. It was clear and concise with usable examples.

The single most influential book every programmer should read...

The same would happen to your topics. The book recommendation topic will list every book on C++ that was ever written.


Regarding editors...

What constitutes an acceptable editor depends entirely on what you are used to, and what features you regularly use.

It's difficult to recommend something to another person, because you just can't know their preferences. While some features are indispensable to one user, another hasn't even heard of them. And yet both use the same application. What good is one user's recommendation for a new text editor to another?

Without specific objective criteria, what do you consider to be a good answer to a question? The one you like best, the ones you know?

I prefer vi via SSH to Notepad++ on Windows. I prefer TextMate on OS X to both vi in Terminal, or MacVim. I haven't figured that one out yet. What editor would you recommend to me?

A topic about editors will be dominated by the usual suspects everybody knows, while there's also three pages of editors almost nobody ever heard about. As soon as another editor comes out, someone will append that one to the list. It will look like a Google search results page or Wikipedia list of text editors (except that the feature comparison tables will be missing, and the ordering is rather arbitrary). Late entries will have no chance to get to the top, even they are are popular and almost perfect for almost everybody.

Even if the state of the art changes, the same topic with the same old, now perhaps obsolete recommendations is still around, with new entries having to chance to rise to the top. What do you mean, the top 10 popular recommendation don't cover C++11? Time to open a new recommendation topic with implied requirements that change over time. PHP6? HTML5/Canvas? Nope, you're still building your web pages using tables and you better like it.


It's really easy to research programming books, text editors, or graphics editors on the web. A simple Google search produces dozens of viable results very quickly. SU is not an encyclopedia of top lists related to computers, and those topics wouldn't add anything to the site.

What adds value are specific questions. Questions that cannot be easily solved otherwise, questions that require in-depth knowledge. Questions that aren't answered by posting "Notepad++" or "Stroustrup".

So if you need a PHP5 (!) editor that has project-wide function and class name completion, some way to quickly find and open files that are part of the project by a substring of the file name, and needs to work on Linux and Windows, asking for that might work.


But your examples are too broad in scope, and too subjective to be of any use to others. They might be interesting to browse for someone with too much time, or to find new toys to play with, but otherwise offer little. And they don't provide nearly enough information to allow us to close the more specific recommendation topics.

  • Of course you mentioned periodically re-asking the same question. But they are unfailingly popular, and will dominate the site. Also, who decides when to re-ask? After a new C++ standard is out? After it has been adopted by compiler writers? After a new major release of one of the text editors is out? – Daniel Beck Jan 23 '12 at 16:57
  • The community would decide when it would be re-asked. I would suggest every six months, automated, to begin with, but that's just a wild guess and a starting to point to see if it would work. – Django Reinhardt Jan 23 '12 at 17:22
  • I'll just add: You bring up plenty of reasonable problems, but I still believe we should be able to find solutions. There's a huge amount of knowledge and expertise here that isn't being tapped. Yes, the answers given might not be perfect, but we are, as people, used to that: We accept magazine/site/blog review round-ups of books and hardware. We also accept Amazon.com reviews when we come to making a buying decision. This sort of knowledge CAN be channelled into a useful form. It's just a matter of overcoming the problems and not giving up before we begin! Isn't that how this site started? :) – Django Reinhardt Jan 23 '12 at 17:25
  • @JohnnyW Ignoring books for a moment, something like Wikipedia's lists of text editors, comparison of text editors, and one prose article per program is far superior for something like this. Just select the one or two programs with the desired feature set, and you're done. While free-form responses allow for good flexibility here, they're not useful to properly compare equally good or valid recommendations. – Daniel Beck Jan 23 '12 at 18:40
  • 3
    I will add that the point of the SE sites is not tap all the useful knowledge of it's users. The purpose is to be the Q&A site where there is as near as possible a one-to-one relationship between the question and the answer. This is not always possible. There are many times where there are multiple perfectly good solutions to the same problem. The fact that this unpalatable situation occasionally arises does not mean that we should stop striving for that desired one-to-one relationship. – EBGreen Jan 23 '12 at 20:50
  • I believe, also, that this site strives to be a means to provide answers to questions that you would be hard pressed to find answers to elsewhere, it they even are out there at all. The point of having a community of experts coming together to share knowledge isn't to catalogue all the common trivia that one in any particular field will come to know and to share our own opinions and preferences, no matter how well reasoned they are, but to solve problems and answer difficult questions that just aren't covered by your typical reference material or textbook, or which require that special combina – Ben Richards Jan 27 '12 at 5:08
  • 1
    I believe, also, that this site strives to be a means to provide answers to questions that you would be hard pressed to find answers to elsewhere, it they even are out there at all. The point of having a community of experts coming together to share knowledge isn't to catalogue all the common trivia that one in any particular field will come to know and to share our own opinions and preferences, no matter how well reasoned, but to solve difficult problems that just aren't covered by your typical textbook, or which require that special combination of knowledge and experience. – Ben Richards Jan 27 '12 at 5:09

You must log in to answer this question.

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