-6

I answered Apache SSL not working and it received downvotes and comments that made me wonder: is our primary purpose, with new answers to historical questions, to answer OP's post or to provide useful info to people who might turn the question up in a search?

I'd Googled a particular issue and found this Server Fault question came up first. After researching to find an answer that worked for me, I proceeded to answer with the solution that solved the problem in my case, since I thought it was a potentially useful addition for people who found the question via a similar search pattern to my own, even if their actual problem was slightly different.

Is the benefit to the community contextual, or is it best to answer the details of the question exactly as asked, even if it is a very old question?

14
  • Again the downvote is interesting. Is this a question I shouldn't be asking? Is there a lesson to be learned here? Commented Jun 18, 2017 at 16:17
  • As voting is community specific and you're using an example from Server Fault and your discussion seems to focus on that, the down voters might feel the question is better suited at the meta of SF. They feel in its current form it is not useful to be discussed as an issue for the broader SE network. Unless you clarify why your question is relevant for all of us and then address that in your question with an edit.
    – rene
    Commented Jun 18, 2017 at 16:28
  • @rene my discussion is general; were you unclear on the example being example only? I thought it fairly well explained that I was not asking for vindication or upvotes for my specific SF post. The question has very clear general relevance: it asks 'is it more important to answer the technical details of the question, or to bear in mind the sort of users who might find it and be helped or not by it?'. What did you not understand, rene? What made you focus on my specific example? Commented Jun 18, 2017 at 16:29
  • 4
    I have read the question twice now, I have no idea what you want to be discussed, not a single clue. Are you asking if we need to be mind readers and predict the future?
    – rene
    Commented Jun 18, 2017 at 16:32
  • @rene that's a shame as I have tried to explain twice in very clear English. Commented Jun 18, 2017 at 16:33
  • 3
    It really isn't very clear what you're asking. I think I get it now... kinda... but it's definitely not clear.
    – Cai
    Commented Jun 18, 2017 at 16:34
  • It's probably not your English then but my reading comprehension. I'm happy to leave it for the higher educated and native speakers, this is out of my league.
    – rene
    Commented Jun 18, 2017 at 16:35
  • If I'm using over complex forms for the situation I'm open to that as a relevant criticism. I am more than happy to try to revise the question. Give me a little while to have a think and work out if I can come up with a simpler formulation. Commented Jun 18, 2017 at 16:36
  • @rene I think what's being asked is if answers should always answer the specifics of a question or if it's ok to answer more generally a solution that may solve a similar problem (i.e. The example question is about a specific setup of a specific version and the answer only applies to a different setup of a newer version) ...or something like that
    – Cai
    Commented Jun 18, 2017 at 16:38
  • Is that right @Peter ?
    – Cai
    Commented Jun 18, 2017 at 16:39
  • 1
    Sorry, not very clear to me either. Are you asking if it's OK to answer a question other than the one OP had (albeit similar) because people from Google might have that other question?
    – Laurel
    Commented Jun 18, 2017 at 16:39
  • 1
    @Laurel I'm saying: is our primary purpose, with new answers to historical questions, to answer OP's post or to provide useful info to people who might turn the question up in a search (though that's overly simplistic, but best I can do in simple terms). Commented Jun 18, 2017 at 16:55
  • 1
    In theory, you're correct, and I myself often find useful new answers in old questions. However, that's not a rule. There are many cases where the question itself became obsolete, so any new answers are not really relevant. If anything, a new question should be asked, possibly with a self answer. That might work better in those cases. Commented Jun 18, 2017 at 21:39
  • @ShadowWizard I agree with you, but I wonder if there are aspects of search patterns that might be reexamined assuming a formal relationship with Google (if there is that's fine and before my ban period I was naive to think otherwise). This is above my rep grade atm, of course, and I am humbly bringing up the topic. Commented Jun 19, 2017 at 10:32

1 Answer 1

7

Answers must fundamentally answer the question. If I ask how to X the Y in version 1.2.3 of A, an answer that tells me how to B the C in that version isn't useful, and nor is an answer that tells me how to X the Y in different product D or in version 2.3.4 of A.

What should you do when a question that is close has great Google juice, and there doesn't seem to be an existing question asking how to do what you've just discovered in a later version?

  1. ask "how do I X the Y in version 2.3.4 of A"? yourself. (Consider putting the version number in the title, and choose your tags well.)
  2. answer it yourself with the simple technique you have discovered
  3. wait a day or two for the community to respond - there may be a duplicate, there may be a better way to do what you've just learned about, etc
  4. return to the similar question and add a comment that says "for the same issue in version 2.3.4, see " and paste a link to your question.

You may not have commenting privileges yet, but since you'll be gaining rep for both the question and the answer, if this is a genuinely useful contribution, you should get that pretty quickly. Once the questions are linked, anyone finding the old one will see a link to the new one both in your comment and on the side in the Linked section.

Now you have kept the site clean and relevant. In this case chances are nobody is using the old version of the product, but on questions where things move more slowly, I would be really angry to see "on the version you can't upgrade to for whatever reason, the thing you're trying to do is easy, just do this" kind of answers. They don't answer the question. But if someone Googling for the issue on the current version does land on the old question, they will see two signposts to the newer one.

5
  • My question was objectively bad by this measure and it's right that it was, however well intentioned. Nevertheless, I wonder if Google is given/giving answers that relate to unrelated questions if this should be fixed at source and should not have to be monkey patched with objectively bad answers. I understand this is likely difficult and I assume Google devs are much cleverer than me, as I assume most people also assume. Commented Jun 19, 2017 at 10:36
  • 2
    Google doesn't understand the search terms you use. Calling it "monkey patching" when you add links to related questions is misunderstanding both how the StackExchange system works and how Google search works. Besides, you can't change Google and you can change how you use this site. Commented Jun 19, 2017 at 10:44
  • I don't understand. My concern was that people searching would get answers to the questions they had. I changed how I answered the question because of a flaw in the Google search I encountered which failed to understand my intent and the deepness of reading I was able to engage in at the point of the search. Because this is a subjective viewpoint I sought outside opinion. Commented Jun 19, 2017 at 10:48
  • 3
    I can't understand your point. You searched for X. You found a question about Y. You wanted to put an answer about X on it. People told you that was wrong. I toid you to ask and answer an X question, then comment on the Y question linking to the X question. You called that monkey patching and suggested that SE should somehow change Google. Forget that line of thinking. Answers should answer the question. Period. Comments are handy for edge cases like this. Commented Jun 19, 2017 at 10:58
  • My point, I suppose, is that people are illogical and people are the intended users of the site, not bots. If mine was an outlier that's fine, but edge cases are often much more than outliers. A negative number can be an edge case, for example, but very much not an outlier and very certainly there for a reason. Commented Jun 19, 2017 at 11:02

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