5

After doing some digging (i.e. FAQ for Stack Exchange¹, What is migration and how does it work?², and Can we have an Off-Topic > Belongs on StackOverflow option?³), I'm unable to find any explicitly defined criteria for what it takes to get another site added to our site migration list under the closed vote queue:

enter image description here

Via implied commentary [³], it seems like there is a threshold of posts that need to be migrated to a different SE site before said site is evaluated and potentially added to the list found in the closed vote queue. This feels like a bit of the chicken and egg scenario as the only other way I am aware to have posts migrated is by having the admins do it [²] after raising a flag on the question in particular. I feel migrating posts in this fashion isn't very streamlined, and as a result, questions that should be migrated often just get closed out for other reasons, which in turn reduces the numbers we would reference to justify adding the site.

Is there any minimum post migration threshold we are beholden to before we can get another site added to this list, are we able to just make a logical argument for another site, or is there just an entirely different set of criteria I'm unaware of that we must meet before we can get a new site added?

I'd love to know because, frankly, I feel we're missing the ability to migrate posts to Server Fault as questions that tend not be database or developer related often qualify as a server issue instead.

In response to @Aaron Bertrand's comment:

dba.meta.stackexchange.com/a/1281/1186 - and as added context, there have been 0 migration attempts to Server Fault in the last 90 days. 18 to Stack Overflow, one to here, and one to Super User.

Thanks for the numbers and link to the additional meta question. It still doesn't provide explicit criteria, but it's yet another example of what I suspect is being evaluated.

As a follow-up to this though, I did a quick review of the vote closures I've been a part of that were migrated, and of the 7 I touched, 3 arguably could have gone to Server Fault (1, 2, and Hibernate does not close idle session [migrated] closed/removed so cannot say for certain). Two of these three were migrated to Stack Overflow, likely because it was the path of least resistance, and the other was migrated to Super User (which in my mind has a lot of overlap with Server Fault, and I know I flagged that question asking it be migrated to one of those sites).

This basically goes to further solidify my point though that we're in the chicken and the egg scenario. We're using loaded numbers to justify why we shouldn't add a site to the list. I suspect if we were to go through all the migrated questions, you would find some that should have been migrated to Server Fault instead of Stack Overflow, but because the UI is designed to streamline migrations to Stack Overflow and not Server Fault, the numbers are really just reflecting the easier migration path available to the general community.

July 18th Update: As it's been about a month since I asked this question and we've only had 9 total votes against both answers (not including down-votes), I figure that in and of itself is enough evidence this feature, if added, would go largely underutilized. I've marked the "status-quo" answer as accepted, but it's ultimately up to the admins on if they want to include these sites or not. Once again, thanks for the various insights into this process.

  • 1
    I've been thinking the same recently. Thanks for asking this question. – John aka hot2use Jun 19 at 7:06
  • dba.meta.stackexchange.com/a/1281/1186 - and as added context, there have been 0 migration attempts to Server Fault in the last 90 days. 18 to Stack Overflow, one to here, and one to Super User. – Aaron Bertrand Jun 19 at 12:34
  • @AaronBertrand Thanks for the info and I also quasi-responded to your comment with a question edit. – John Eisbrener Jun 19 at 13:56
  • There is some subjective here, too. We, as moderators, in order to change the list of potential migration targets, need to make an explicit request of the community team, and provide justification for them to invest in such a change. In order to do that we also have to be fairly confident ourselves that the work required is worth the effort. For three migrations, that could have easily been handled manually by any of us with a flag, I'm not convinced. Yet. – Aaron Bertrand Jun 19 at 14:36
  • @AaronBertrand Take note that this is only 3 of my 7, not 3 of the 19 you see that were migrated. I suspect there may be more, but I don't have access to the full list of migrated questions, only those I had a hand in nominating to be migrated. – John Eisbrener Jun 19 at 14:41
  • I still think even if all 19 belonged on Server Fault (highly unlikely), it would be more work making the case than to manually handle all 19. I'm only one moderator opinion, though. What might be more compelling to bring to the company is that [x] migrations to the targets we have available, which didn't involve moderator intervention, ended up in the wrong site (and maybe migrated again, when possible, or the user was asked to post a new Q elsewhere). That is not easy research, but the current state doesn't really demonstrate that a whole lot has gone wrong or extra work has been created. – Aaron Bertrand Jun 19 at 14:46
  • TL;DR: The "juice for the squeeze" bar for the company to implement changes to an established site is pretty high. – Aaron Bertrand Jun 19 at 14:59
  • @AaronBertrand I totally get it. I guess I'm looking at it from a percentages based perspective (3/7 ~ 43%, which is pretty high) instead of a raw numbers equation. I just don't know how much "work" it truly is to add another site in a radio button list. It feels like it would be a trivial change, but I don't have the slightest clue as to the underpinnings of the site. As always, thanks for the feedback! – John Eisbrener Jun 19 at 15:08
  • You should check out the main meta where you can see the opposition to changes that seem to be simple one-liners. It’s not just the work itself but prioritizing, testing, etc etc. All the things that make any software development take 10x as long, but then add the simple “it ain’t broke” and “resistance to change” attitudes. It’s the primary reason I stopped suggesting or championing change proposals at all. – Aaron Bertrand Jun 19 at 15:54
  • @PaulWhite, It's been about a month since I asked this question, so I'd be happy to mark either of these answers as "accepted". At the end of the day though, it's up to the site admins on if they want to move forward with this or not. – John Eisbrener Jul 16 at 17:53
4

explicitly defined criteria for what it takes to get another site added to our site migration list

I'm not aware of any such criteria. My understanding is that the original list is formed early in the site's life, and can be added to on demand after that. Demand would typically manifest as meta questions demonstrating the need for a new migration path, ideally with examples, and broad community support.

I feel migrating posts in this fashion isn't very streamlined

Flagging for a moderator to migrate requires one user to flag and one moderator to handle that flag. Migrating via the review mechanism requires at least five people to get involved.

questions that should be migrated often just get closed out for other reasons

This is the more interesting part for me. What is meant by "should be migrated" here?

How often do we get well-presented questions that are immediately and obviously on-topic for another site (that isn't Stack Overflow)? How often does the OP already have an account on the target site? How much time is saved in the best case? How much time is wasted in the worst? Which is more likely?

How often will we manage to migrate successfully via 5+ person review before the OP decides to ask the question themselves on the target site instead? How often will it result in a mess at the target end where their moderators have to get involved in merging duplicates?

How often do the dba.se reviewers have enough familiarity with the target site to make an informed decision? Stack Overflow is less contentious because there is some topic overlap, and many dba users also have an active account there. How often will that be true for other sites? We can't limit reviewers to just those with accounts on the target site with a certain level of reputation.

Separately, an easy option for "throwing the trash over the fence" isn't really something I want to promote. I can count the valid migration flags I have ever handled on one hand.

Closing as off-topic and leaving a comment that the OP can delete and re-ask on the target site after reading their FAQ and posting guidance seems most often the best outcome in my experience. If the OP neglects to self-delete, the closed question will normally be automatically deleted by the system after a short time.

Related here: Can we be allowed to choose a suitable forum if DBA.SE is not suitable?

Related on main meta: A proposed philosophy of question migration

5

This answer suggests we add migration paths for Server Fault and Super User.

We promise to use the ability to migrate responsibly.

You must log in to answer this question.

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