13
\$\begingroup\$

There are some questions about the setup of the site that get asked time and time again on Meta. These include allowing code-golfing languages, language-based handicaps, duplicate-closing on the main site, and others.

Whenever these questions are asked, they are immediately closed as duplicates, because the question has been asked before. For instance, this question, asked yesterday, was closed as a duplicate of this question, asked over four years ago.

For another example, this question about golfing libraries was closed as a duplicate of this four year old question, despite the former getting an answer with 16 upvotes that recommends the opposite of the top answers to the latter. This is a case where the consensus has pretty clearly changed over the intervening period.

While in theory new answers can be posted to these very old questions, in practice they essentially never are. Moreover, opinions change over time, and meta answers are very often opinion based, as opposed to the objective questions on the main site.

Closing new questions in favor of ancient ones is detrimental to the sharing of ideas, and promotes the status quo regardless of whether the status quo is desired by the community. It also makes it harder for newer users to interact, because they might have a hard time finding the appropriate ancient question to post an answer to.

There are at least three options we have:

  • The current policy: Close all newer questions as duplicates.

  • Create canonical, high-quality questions which are kept up to date and easily reachable, perhaps with a special tag, and close everything else, including the older question, as duplicates of it. This might encourage continued debate in an organized fashion.

  • Don't close such questions as duplicates of questions that are at least year old. Post an answer with similar content as the top answer of the older one if appropriate, but allow renewed discussion.

I am strongly in favor of changing this practice, but I'd be interested to hear what others think.

\$\endgroup\$
  • 8
    \$\begingroup\$ On meta as on main, this community seems to be very hesitant to close old questions as duplicates of new ones. This is reasonably common practice on other SEs, and I wish we made more use of it, too. I suppose option no. 4 would be a middle ground between your options 2 and 3. Give the question some time to accumulate answers, and if they provide nothing new close it as a duplicate. However, if it turns out that today's opinion differs from what was discussed before, clearly the old one should be closed, since it's not only a duplicate but also has outdated and misleading answers. \$\endgroup\$ – Martin Ender Jan 3 '16 at 22:40
  • 1
    \$\begingroup\$ People seem to be applying a very loose meaning of duplicate on meta, as anything with the same general topic and opinion. \$\endgroup\$ – xnor Jan 4 '16 at 3:31
  • 2
    \$\begingroup\$ I am voting to close this as a duplicate of... nevermind. \$\endgroup\$ – Fatalize Jan 4 '16 at 14:25
  • \$\begingroup\$ I feel like options 1 and 2 are basically the same. We want the older questions to be high-quality, up-to-date, canonical questions. If we went with option 2, we'd basically end up with the same situation \$\endgroup\$ – Nathan Merrill Jan 7 '17 at 17:26
  • \$\begingroup\$ @MartinEnder if we have another meta post that ends up with the same answers as the older one, what benefit would we gain by closing the newer one? The newer one would be more visible, and have answers that are more similar to the current opinion. \$\endgroup\$ – Nathan Merrill Jan 7 '17 at 17:29
  • \$\begingroup\$ Last comment, sorry: Finally, while I would love it if we could revote on many of the meta posts here, I feel like the majority of the community would vote to keep status-quo than change the rules of this site. If somebody were to post "Can I use HQ9+ on this code-golf challenge", I would likely upvote "No", even though I don't agree with it. \$\endgroup\$ – Nathan Merrill Jan 7 '17 at 17:34
8
\$\begingroup\$

Close only if they would have the same answers

For the question https://codegolf.meta.stackexchange.com/q/7887/25180, the answers are supposed to be:

  • Yes, they are drowning us. (But we may want to be drowned anyway, or we may want to find a way live with it without changing the rules.)
  • No, they are not. (But we may want to ban golfing languages for other reasons.)
  • It would be only temporary, or whatever.
  • Some suggestion about making new languages better accepted, or easier to learn. (It has nothing to do with whether they are allowed or not.)

They doesn't seemed to fit in the old question. Not banning golfing languages as a rule doesn't give much information about how to improve this situation (or whether some improvement is necessary), especially when there are many users designing new languages which wasn't the case in the past.

Banning golfing languages is a really unlikely solution to the new problem (if we think it is a problem). And we decided not to ban golfing languages for good reasons. It's not the case that we should ignore anything that may lead to banning golfing languages no matter how bad it is.

Note that there are already several questions about golfing languages in the past, and they are not closed as duplicates of each other. Some examples are:

Those questions have some minor differences, but some of them might be closed if they were posted today.

Now the community has decided not to ban golfing languages. We should (at least technically) allow more of those questions if they were not exactly asking for banning golfing languages, not less. As we don't argue too much about whether we should solve the problem by banning some languages anymore, the discussions should be much more clear and relevant than before.

At the very least, we should close the questions as duplicates exactly for the reason they were answered in the old question, not for the reasons such as we don't want to talk about it anymore, or it contained some bad ideas we don't like. We can comment, downvote or close for other reasons in other cases.

\$\endgroup\$
0
\$\begingroup\$

Post an answer explaining that this has already been addressed

When an opinion or discussion-starter of yours is closed as a dupe, it feels strange to be told that your viewpoint is a redundant with someone else's. To quote Martin, "Yes, there isn't much of a point in writing the same kind of answer to every of these questions, but closing them as a duplicate probably doesn't make the author feel like their concerns are being taken seriously." It's especially frustrating if you wrote your post in full awareness of alleged dupe, to bring up a related situation or explore another option in depth.

I see dupe closures on meta that over-reach, not saying "This is the same as another question," but "My answer to this question would be largely the same as to this other question." These are effectively answers and should be posted as such.

The answer doesn't have to be long or repeat existing material. It should basically link to the previous discussion, explain how it applies to the current question, and say why you think it's satisfactory. If community consensus is with you, then closed it as a dupe.

I see many advantages of doing this:

  • People can vote to register agreement or disagreement. If the dupe claim is unsatisfactory or community opinion has changed, this will be apparent.
  • The answer explains why the dupe applies. This is especially helpful for new users who may be unaware of the context of past discussions, or be using different terminology.
  • Competing answers can be posted. If someone claims the situation is different than before, they can explain why and others can vote on it.
  • It feels less dismissive. Even a short answer saying that this has already been addressed feels like a person responding, not some faceless bureaucratic process telling you that what you said was invalid.

I should be clear that this is for posts expressing a view where claiming a dupe is itself expressing a viewpoint. If being a dupe is just a matter of fact, simply close it.

\$\endgroup\$

You must log in to answer this question.

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