-17

In a post earlier I showed a couple of sites that are held back by a community that doesn't exist, and thus can't prosper. The tldr there is even though some want to provide more value, they can't because value is given with reputation and many sites aren't producing reputation. There seems to be two technical solutions at hand,

  • Rules: change the network rules so a site is more hospitable to forming a community organically. I've suggested this over in another qestion.
  • Community: change the network rules so the site actually grows a community.

The problem is the latter isn't politically acceptable, but it seems very obvious to me. Some sites never form because they can't compete with Stack Overflow. Let's review,

  • This is a real material problem where I can not provide more value on a site.
  • Fixing the limits such that I can provide more value by lowering the rep limits to edit/tag as I put forward on the other post isn't popular.
  • But, everyone agrees with the numbers, that for those sites in their current state the limits are unrealistic.

How can we make those limits more realistic without changing them? We can grow one community at the expense of another with technology. This isn't a bad thing: why should communities compete with each other anyway? The network's content is a gigantic zero-sum game, for one site to get a contribution other sites must lack it.

Here is my suggestion, given a site like DevOps Stack Exchange which is really suffering, I think we should push [terraform] questions over to it. Terraform is a DevOps technology.

HashiCorp Terraform provides infrastructure automation with workflows to build composition, collaboration, and reuse of infrastructure as code while providing standardization for security, compliance, and management. The extensibility allows for integration with workflows to drive innovation with self-service workflows.

If a question is just tagged [terraform] there is no conceivable way for it to be more on topic on Stack Overflow than Devops.SE. Why not just help the users out, and the community which is struggling,

  1. Make the users aware of the DevOps community which is likely where it belongs based on the presence of one tag which can infer with great accuracy a better fit on the network.
  2. Better yet, offer the user the ability to migrate the question
  3. Better yet, just move it. Stop sites from competing when the material is clearly better homed on another site.

We can even operate with all of these based on certainty. This doesn't just go for DevOps either. There are a lot of cases for this,

If ALL of your tags are exclusive to the same site, force the migration,

  • [devops] [k3s] [azure] [aws] [terraform] [terragrunt] [chef-infra] [docker-swarm] [jenkins] [ansible] [gitlab-ci] [github-ci] [kubernetes] [k3s] [kubectl] [continuous-integration] [continuous-deployment] → DevOps
  • [sql] [postgresql] [mysql] [sql-server] → Database Administrators
  • [linux] [debian] [centos] [rhel] [fedora] [wayland] [xorg] [pulseaudio] → Unix and Linux Stack Exchange
  • [ubuntu] → Ubuntu Stack Exchange
  • [vim] [neovim] [spacevim] [syntastic] → Vim Stack Exchange
  • [emacs] → Emacs Stack Exchange
  • [latex] [tex] → TeX Stack Exchange
  • [android] → Android Stack Exchange
  • [qgis] [esri] [arcgis] → GIS Stack Exchange
  • [notepad] [windows] [powerpoint] [gimp] → Super User
  • [radare2] [ida] [ghidra] → Reverse Engineering Stack Exchange
  • [commodore] → Retro Computing Stack Exchange

And if you have MORE than just one the above tags tags, and another tag which isn't an exclusive tag, we can OFFER the migration without forcing it.

  • [vim] [perl] → offer, but don't force Vim Stack Exchange ... etc

I think this would,

  • Clean up Stack Overflow by making it more topical
  • Make the rest of Stack Exchange far more successful and popular
  • Give the space for communities to evolve cooperatively
  • Encourage people to do a better job tagging

Lastly, if you don't solve this problem then every community will be competing with Stack Overflow. Many will lose. We'll have more duplicate content. The experience on the network outside of a few very large inclusive sites will struggle. It's also a problem for the brand. If I was just a DevOps person and I knew Stack Exchange from that site, I would not be interested in ever participating with another community here.

6
  • 4
    Please provide evidence that sites such as DevOps, Database Administrators etc want this i.e. a highly upvoted Q&A on those site's Metas indicating they are in favour. Jul 6, 2022 at 23:17
  • @RobertLongson The problem is on those nothing is highly upvoted. That's what i'm trying to solve. Perhaps you disagree, but in this case I'm going to argue that regardless of what any site wants, it's in the best interest of the network to take action. I believe this is a much better policy then "let every site run independently, regardless of if the StackExchange network suffers as whole." However doom and gloom, the last Q on meta.devops was over three years ago where they basically complained about this very thing. Jul 6, 2022 at 23:24
  • I'll link a recent post like this one from SO: Let's gift wrap our (good) machine learning theory questions for Cross Validated... It expands on this topic in parallel by showing what some SMEs think about migrating and making stricter rules about cross-posting and site on-topic overlap.
    – bad_coder
    Jul 6, 2022 at 23:25
  • Did you see the site-crossover and cross-posting tags here and on MSO for previous posts on the subject?
    – bad_coder
    Jul 7, 2022 at 0:05
  • @bad_coder I didn't feel like any of them covered the suggestion. Feel free to tell me which one you feel matches. Jul 7, 2022 at 5:45
  • @EvanCarroll I don't know how the debate on site-crossover went (and it would likely have to be very specific per-tag/site, etc), but I'm guessing it would take lots of community and company support to gain traction on establishing a policy that clearly defines scopes and then goes on to start migrating everything to a target site from that point onward. So I find this post very interesting but I don't see the local audience interested in it. Robert Longson probably got it right that you'd have to start by gathering a number of interested SMEs and there'd always be pushback from another camp..
    – bad_coder
    Jul 7, 2022 at 9:52

1 Answer 1

14

To be blunt (sorry), forcing migrations is a really, really bad idea.

It would not make sites "more topical." Look at the [android] tag on Stack Overflow for example. You'll mostly find questions like How to get the names of all files in a directory programmatically. Questions like that are not on topic for Android SE, which specifically says in its Help Center: "Development/programming questions are not covered on this site." The questions that would be more on topic on Android SE (and off-topic on SO) don't always have a single tag (much less the one you would expect) nor are they usually of migration quality. Migrating these questions automatically would be detrimental to small sites which likely already struggle with moderation. It would also be detrimental to the people that chose the right site to begin with but maybe didn't choose the best set of tags.


Suggesting a user post elsewhere is already a feature that exists. Or rather several:

(This should happen before the question is posted, which is the best time.)


It's also relevant to note that there are already tools that mods can use to monitor tags on other sites:

  • Tag filters
  • RSS feeds (which can be inserted into chat rooms too)

From there, it's as simple as asking the other site's mods to migrate any questions that are off-topic. However, in all the months I've had tag filters set up for some tags relevant to my sites, there hasn't been anything that was appropriate to migrate. (But then again, I don't feel the communities I moderate are really "competing with Stack Overflow".)

5
  • Right off the bat, I don't find your example at all convincing your android example should very obviously be tagged with the language. In no way is any question about a programming task to do X divorced of the language sufficiently refined for the site. The answer presumes Java or Kotlin. My cell phone runs Perl, Python, and Shell too. This speaks exactly to the point of "Encourage people to do a better job tagging". Jul 7, 2022 at 3:07
  • 1
    @EvanCarroll but not everything on Stack Overflow needs specific language or code. There are valid language-independent questions about Android programming (e.g. API clarification, lint message → this one has [logging] tag, but it could be replaced with [android-log] tag). And almost everything on [android-*] is related to app development anyway. Those are still off-topic on Android.SE. Jul 7, 2022 at 4:17
  • @MetaAndrewT Again, another example of a bad tag. That's not supposed to be tagged Android: the "android" tag is for "Google's mobile operating system". That question should be tagged android-log which is explicitly for that. (fixed now). Also you're right on android-* I was specifically trying to target the versions with that. Like [android-11] and such. I'd say the same for [postgresql-12] and the like. Jul 7, 2022 at 4:31
  • 4
    @EvanCarroll Please do not remove relevant tags from questions. That question should absolutely be tagged [android], even if it's also tagged [android-log] (a tag whose necessity for existing I seriously question, given the existence of the [android] and [logging] tags).
    – Ryan M
    Jul 7, 2022 at 4:54
  • @RyanM If your argument is that questions on Android API should be covered by android you should edit the tag wiki and define it that way. It's not clear at all me that questions on the API are covered by that tag. It seems at first glance like it would be very confusing to use the same tag for the API, when the most popular question that just has tag is stackoverflow.com/q/2033914/124486 (and has 0-to do with anything programmatic). Jul 7, 2022 at 5:36

You must log in to answer this question.

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