I see tons of these floating around. Some examples:

And then of course there are some alternative but still related ones:

Can we have a single, canonical question that we all agree to flag duplicates against?

2 Answers 2


I'm going to try and answer in a general way such that this could itself be a canonical Q&A, if you had asked "How can we create generalized, canonical questions out of a series of localized ones?"

Can we have a single, canonical question...

Some people write their own Q&A when they notice something that hasn't been covered, which is great. I think about doing it all the time WRT things that have been covered, but (as with your example), not in a clearly canonical way. Stupidly, I still have yet to actually do this, and instead continue to type out answers to questions that should simply be referred to such a canonical version. [Maybe the game point system encourages this inadvertently?]

I have not looked at the actual Q&A's you reference, but if you think they could be summed up -- and of course you are free to refer to and quote them -- then I think you should go for it.

that we all agree to flag duplicates against?

Without a formal mechanism, that's a tough thing to accomplish.1 And although I do empathize with the desire for such a mechanism, I'm glad there isn't -- in essence you'd then have a system that doesn't need "us" to agree to flag anything. You'd just have one person consult a list that was previously agreed to and mark the duplicate, or whatever. The problem with that for me is I've seen questions closed inappropriately (IMO) on S.E. by five people. So either it takes 5 people to agree or just 1 or some other number...and/or X number of people agree to abide by a standard previously set by X number of people.

What we've got already is a sort of a democratic, decentralized "free market" system. No doubt this is laden with inefficiencies and ambiguities by nature -- e.g. I don't know what the heuristic is for the "Related" sidebar (presuming it includes votes, all those inefficiencies and ambiguities are also included), but it can't be made perfect, so it sometimes disappoints.

I can't recall seeing it done, but personally I have no problem with someone writing a canonical Q&A in a case like this and explicitly referring to the not-so canonical questions they are trying to encapsulate. In other words, literally list them like you've done here and then say, "I'm seeking a canonical answer to sum up all of these", then write what you think it should be. Anyone who sees that in future will get the point and, if you've done a good job, support your efforts by using it as the canonical reference. It will also help people looking for answers recognize what all those have in common, and hopefully why yours is the general answer.

1 Gilles mentions the reference question, which I was unaware of. I'm not sure if that can be applicable in all cases though.

  • OT: Maybe the game point system encourages this ← of course it does. The whole point of the game point system is to encourage content creation, priming quantity over quality. OTOH, meta, votes to close, moderation tools, moderators, etc... are about ensuring quality. It's about finding good points of equilibrium in this complex system.
    – ninjalj
    Commented Oct 16, 2013 at 11:18
  • 1
    @ninjalj the game point system encourages quality, not quantity. Adding content does not get you rep, adding good content does. Adding bad content costs you rep since people will downvote.
    – terdon Mod
    Commented Oct 16, 2013 at 12:35
  • 1
    @terdon I think it does both, although it is more effective at encouraging quality, which makes it worthwhile. But people do play it in ways that perhaps short circuit the "quality over quantity" intent. Eg., it seems to be SOP for many people on some busy S.O. tags to jump in with a quick, ambiguous sentence ASAP that amazingly bags them points while they edit in a bit at time -- in the end sometimes it is a good answer and sometimes it isn't, but nobody comes back to check their vote, and nobody bothers to add anything better because they can see the votes have moved on already.
    – goldilocks
    Commented Oct 16, 2013 at 14:24
  • 1
    @goldilocks yeah, SO is kinda different, true. There it seems that speed is valued over both quality and quantity. Go figure.
    – terdon Mod
    Commented Oct 16, 2013 at 14:26
  • People get excited when they know the answer to something and will upvote things that scan correctly if you already know the answer, but probably fail to help the OP -- and since stuff like that is easier to hammer out, people will favour speed of response over quality of response because they know the early bird gets the worms.
    – goldilocks
    Commented Oct 16, 2013 at 14:26
  • @terdon - I've noticed a few ppl here that attempt to do that, it drives me fnuts, when I see someone with a 10K rep answering a Q in 1-2 sentences max. Especially when there isn't even any code in the A. Then they get a UV here and another there. That's their game. These ppl should be harshly admonished, since they are, IMO, taking and not giving fairly.
    – slm Mod
    Commented Nov 9, 2013 at 8:50

Having a reference question with a solid answer is always a good thing.

The reference question should be clearly linked from the relevant tag wiki. That's how we can agree to use it as a duplicate target.

That doesn't mean that every vaguely related question is a duplicate! For example How to create a customized distribution based on Debian? is not about creating a distribution, and not even about creating a distribution derived from Debian (despite the title), it's about preseeding an installation.

  • Should we perhaps be editing titles in those cases to reflect what the post is really about?
    – kurtm
    Commented Oct 16, 2013 at 13:45
  • @kurtm Hmm, well, the title is correct. It admits other interpretations, because it's a matter of degree. You don't go the same way if you want to inject a few preinstalled packages and set up a few configuration values (like in that Debian-based distro question), or if you want to make a completely different distribution that includes Debian packages (like Ubuntu). Nonetheless I've given a stab at a better title. Commented Oct 16, 2013 at 17:37
  • That does state it better. I realize it's not easy, but when someone says "This one, but the title is really unrelated", I think "Why? We can edit that."
    – kurtm
    Commented Oct 16, 2013 at 17:55

You must log in to answer this question.

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