Hi! I saw your note on TL. I thought I'd take our conversation about cross-posting to a separate room.

Seems relevant to us, as there's such a strong overlap between QC.SE and CS.SE (I'm guessing that many or most questions on QC.SE would potentially also be on-topic on CS.SE?).

I personally have a bias against cross-posting. I think it is good for the person asking (they're more likely to get answers) but bad for answerers (they're more likely to duplicate effort, discussion gets fragmented).

Between the two, I come down on the side of answerers.

As far as my personal experience, I've many times had the experience of answering a question only to discover later that it was cross-posted and already answered on the other site, or (more frequently) left a comment request clarification only to discover that it was cross-posted and requests for clarification on the other site had not been addressed. That feels bad to have my time wasted like that.

The consideration that most sways me is: what if everyone cross-posted whenever they could?

I bet that most questions on QC.SE would be on-topic on CS.SE. What if we said, cross-posting is ok, go for it, and everyone who posted a quantum computing question on QC.SE also posted it on CS.SE (assuming it's on-topic)?

That sounds like it'd be a crummy experience for answerers. I suspect it would also be bad for question quality (comments split between two pages, frequent cases of edits made to one copy but not the other, having to check both cross-posts to see what answers and feedback has already been provided). I'd hate to see us land there.

If that's an undesirable outcome, what should we do about it? My preference would be to disallow cross-posting entirely, so that there's only one copy of each question open at a time.

I don't like the perspective that says, publicly discourage it, but if you discover people doing it, oh well, no harm, just leave the questions be if they are on-topic. That seems to me to be unfair, because those who follow the rules are disadvantaged and those who ignore the rules are advantaged. If we're going to allow it, it'd be better to be honest and state that publicly and up-front.

I know there are some sites who have looked for compromises (e.g., require cross-linking; or require a delay). I'm not personally enthusiastic about those but I think a clear policy that's enforced is better than a situation where the public policy doesn't match what's enforced, or an unclear policy.

I don't think cross-linking helps much, because most of the disadvantages still exist: answerers still have to check both copies, discussion still gets fragmented, you often still end up with multiple variants where some were edited and others weren't.

Requiring a one-week delay, plus cross-linking, plus requiring the asker to update both versions based on feedback received on either copy, is better.

Right now, we have a situation where the default Meta.SE policy says 'no cross-posting'.

(There's a footnote there. The answer on Meta.SE used to say 'No', cross-posting is not allowed. A SE employee edited it to soften that, and leave wiggle room. I'm not enthusiastic about that change, but particularly because it seems to open the door to cross-posting without setting any clear requirements -- "it's best to..." is ambiguous about whether that's required or just a nice-to-have. But it is what it is.)

So folks who advocate leaving open cross-posts seem to me to be essentially in a position of having a vague policy, or a public policy that doesn't match what's enforced.

I see probably half a dozen or a dozen cross-posts on CS.SE per week, so I've gotten to see some common patterns that often crop up with them -- though there are always exceptions.

But whatever. That's my personal preferences, which probably don't count for much.

I know that each site tends to have its own culture around this, and I've learned to adapt to the cultures of different sites.

(For instance, math.SE and SO have a no-crossposting policy; if someone cross-posts and I flag it, mods reliably close the cross-post. Theoretical Computer Science has a nuanced policy that allows cross-posting only under strict conditions. There are a few sites that haven't acted on my flags identifying cross-posts, so I know not to bother flagging on those sites.)

My big question for QC.SE mods would be: what policy will they be enforcing in practice? If QC.SE wants to allow cross-posts, cool, more power to all of you -- I'd just like to know, so I don't waste my time flagging.

For reference, so far I've seen very few cross-posts between CS.SE and QC.SE, so it's a rare phenomenom. I looked through my flagging history on QC.SE and I found only two cross-posts between our two sites that I flagged on QC.SE. (There might be a few other that I closed on CS.SE instead of flagging on QC.SE, which wouldn't show up there.) Those two flags were marked helpful but the questions were left open.

Given the above, I was disappointed that those questions were left open, but I can adjust and refrain from flagging cross-posts in the future and just close the copies on CS.SE instead.

I also found the answer on QC's meta to be vague (e.g., 'allowed, within limits', without a clear statement of what the limits are; 'guidelines', without indication of whether they are required or nice-to-have), so I'm not sure what the policy is in practice or what criteria QC.SE mods will be using to decide when to close vs not.

Anyway. That's where I'm coming from. Phew. That was a really long dump. I'd be curious to hear your take on the subject and if you have any feedback or reactions for me.

and i kind of agree with a lot of your perspective here, and we definitely need to come up with a specific policy - we just haven't really sat down and said, okay, this is our policy, if that makes sense.

i would lean toward the rule of 'no crossposting unless it's been one week, there haven't been answers, and you keep things properly updated and link to everything'