11

This comes up a lot:

I was thinking of going through all the answers, making a list, posting it as a canonical question for this then CV'ing all those as duplicates.

Question 1: Any objections?

Question 2: There's a number of other "my headlights don't work" questions that don't fit "low beams out, high beams work". Assuming no major objections, should I try to put together a general troubleshooting guide that goes beyond just low beams being out, or leave it specific?

4
  • 1
    What does one do when one is CVing?
    – dlu
    Jul 24, 2016 at 23:53
  • 2
    @dlu Click "close" under the question then vote to close. Mods and gold tag holders on the questions tags can close questions as dupes immediately. The rest of us (with enough rep for the privilege) just vote and it goes to the close vote review queue.
    – Jason C
    Jul 25, 2016 at 0:43
  • I've got this on my todo list but it's a ways down.
    – Jason C
    Jul 27, 2016 at 18:57
  • 2
    re. question 2, there was something from an SE staffer on another site's meta saying that canonical questions are supposed to still be answers to a question, not massive encyclopedic tomes that try to house all knowledge on a topic area. If there's another general pattern/answer that works for another common headlights problem, give it a second canonical answer, but only if works as a Q&A. Jul 29, 2016 at 16:55

2 Answers 2

3

Let's wipe out these duplicates, but I think we should do what they do on Serverfault and make these canonical questions clear that they are canonical questions, and they are all that the community have to say at this point.

1
  • Agreed. I like what they do there.
    – Jason C
    Jul 27, 2016 at 18:56
3

The fundamental failure modes may be the same but the circumstances can be different enough to warrant keeping the question

I'm not sure how to feel about the suggestion to close vote these repeat questions and replace them with one canonical question.

Yes, having a canonical Q&A will help to reduce repeat questions. I'm totally for the idea of avoiding exact dupes.

But what constitutes a dupe on mechanics.SE?

I just click on three random questions in the list above and no two accepted answers pinpointed the same root cause of failure. If we were to mark a new question as a dupe, which one would it be?

Of course, we could do a Q&A to document the typical failure modes, but what I'm driving at is that each question has the potential to be unique, because each case that comes through is a slightly different application of the same fundamentals, kind of like when a person is asking for help with how to use awk, find or printf to do X. We don't usually close such questions as dupes and ask them to consult the reference manual/canonical question.

This quote from the canonical question blog strikes a chord with me:

That's why we actually don't mind having several versions of every question, where there are variations in wording or circumstances. The more chance that someone types a question into Google and finds their exact question already answered, the better a job we've done.

5
  • Sure, but keep in mind the other major purpose of duplicate links: To act as a sign post for folks searching for the same problem. None of these questions are particularly likely to come up in a search relative to the others. After the OP's specific issue is resolved, we want the questions to have lasting value, and not just take up space. It's to everybody's benefit to point future readers searching for "low beams don't work, high beams do" to something more complete than an essentially random post. Anyways, I intend to look for commonalities in the solutions before doing anything.
    – Jason C
    Jul 28, 2016 at 21:29
  • (Here's an experiment: Try finding one of those questions by typing an exact problem description into Google. And don't include a description of the solution in your search query, since you don't know the solution yet.) I want to clean up the existing ones to make them more useful, I don't want to squash new ones. I think we can improve the info quality on this site in general with a philosophy like that, for other topics too, not just headlights. Check out blog.stackoverflow.com/2010/11/….
    – Jason C
    Jul 28, 2016 at 21:31
  • 1
    +1 basically, have a canonical Q&A for the standard approach to the problem, but allow questions that have specific circumstances that don't fit (e.g. if it's intermittent, if it only happens when X other factor is true, etc), don't allow the existence of a canonical to prevent people with similar less common variants on the problem getting answers Jul 29, 2016 at 16:58
  • ^ Exactly. I'm sorry if I implied a different approach. That philosophy is obvious to me but I suppose I could've restated it.
    – Jason C
    Jul 29, 2016 at 19:11
  • Indeed @zaid. It's the nuance. Jul 29, 2016 at 22:54

You must log in to answer this question.

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