2

I've been reviewing more close votes recently. There's been certain trend of a few questions being impossible to review correctly to my mind, and I wanted to ask if there's a community view on how to handle them.

I call them "impossible to review correctly" because the decision to close them very much depends on knowing the correct answer and the reason for closure is "doesn't belong on this site" meaning I very likely don't!

They all involve Linux and MS Windows interacting in some way. Depending on the answer, the question might be about something in MS Windows, or it might be about something in Linux.

Example 1

Pinging ipv4 addresses works while ipv6 doesn't - Kali Linux

This is about packet loss on a Linux guest Virtual Machine on a Windows 11 Host. Networking is "bridge mode" meaning the host hypervisor somehow emits packets from the VM with it's own mac address and IP and identifies incoming packets, routing them into the VM.

Tom Yan's Comment is very pertinent:

Note that WiFi client is not designed to support to be bridged. It could be that VirtualBox only implemented workaround for IPv4 to work

I have no knowledge if this is true or not, frankly I'm skeptical, but as a reviewer I couldn't be sure. What's more, this would be a detail about how Virtualbox Bridge networking behaves on MS Windows. That's totally outside my sphere of knowledge and way off topic for this site. So while there might hypothetically be someone here who could answer it, it's not reasonable for us to expect that.

FYI I voted to close because I think it needs a lot more detail to be sure of an answer if it is a Linux problem.

Example 2

Linux Groups - administration tables

This is about linux behaviour on the face of it. But it's Linux interacting with a remote Active Directory server.

Stephen Harris wrote:

You're using VAS which is an AD bridge solution and it can (and will) cache results. What you're seeing is how VAS works. Talk to the software vendor.

What's not totally clear is the extent to which this is about the behaviour of software running on Linux vs the behaviour of a Windows server.

On balance I voted to leave open and leave a vague enough answer based on what I understand might help the OP.

So what to do?

I find myself stuck between:

  • Either we leave questions open because there is "reasonable doubt" that they might still be about software running on Linux. This might be seen by some as "cluttering up" U&L with non-Linux questions. But at least it doesn't ask people to review questions based on details well outside their sphere of expertise. It keeps the review focussed on the question and not potential answers.
  • We slam closed questions because we suspect the might not be about linux, but we can't be sure. Personally I don't like this idea, I don't see it as actually positive. But it might sooth some people's OCD.
  • We skip any question we're not sure about for such grounds, in which case such questions may remain open because there's simply nobody who knows the answer and so not 5 people to close the question.

Honestly I'm not quite sure how to proceed with them.

7
  • 1
    I agree about leaving them open due to "reasonable doubt". In my opinion, closing a question is a significant measure taken onto it, therefore closing a question should not be taken lightly. The one closing a question should be "pretty sure" that the question is not mainly about U&L. Additionally, Windows-focused stackexchange sites may not have as many Linux users to answer questions, so "Linux-related" questions may be more appropriate here ayways.
    – Shidouuu
    Sep 3 at 16:35
  • 1
    Not directly related, the other day I was thinking about how good would be cross site questions: those that by its nature could be answered in multiple SE sites. I guess that it would be a nightmare to implement, but could be a nice option in some advanced/niche cases such as the ones you mention.
    – schrodingerscatcuriosity
    Sep 11 at 12:21
  • @schrodingerscatcuriosity for this reason I've always been sad at the splitting of SO. I realise it has benefits, but somehow it's made it harder to get answers sometimes.
    – Philip Couling
    Sep 11 at 13:37
  • FWIW, your example 1 would have failed even if you were running Linux through KVM on top of another Linux host. The fundamental problem is as @TomYan described: that WiFi client devices almost always cannot bridge. (I'd have left it open but answerable only as "you can't do that")
    – Chris Davies
    Sep 21 at 16:09
  • @roaima it's a funny one because the fundamental problem TomYan aludes to is one of MAC addresses with WPA, which are an OSI layer below IPv4 and IPv6. Therefore, the fact that IPv4 works suggests that you can do that (at least theoretically), and the issue is perhaps one of virtualbox implementation: IE a Windows program in this context.
    – Philip Couling
    Sep 21 at 19:27
  • @roaima indeed, thanks for the reminder, I've just checked back on my home WiFi... on MacOS with virtualbox 7.0.10 with Ubuntu 22.04 guest and IPV6 with bridge adapter over host WiFi works just fine. So I'm still in the dark as to whether the problem is Kali or Windows. My money is in the Kali user being the fault.
    – Philip Couling
    Sep 21 at 19:48
  • @PhilipCouling wow, impressed. I'll revise my opinion of WiFi client devices
    – Chris Davies
    Sep 21 at 21:43

2 Answers 2

Reset to default
5

The rule of thumb is: if in doubt, skip. So if you are not sure it is off topic, just skip it. That's absolutely fine. Only vote to close if you are sure that the question is off topic. If you are not sure, just skip.

As a general rule, we should err on the side of keeping questions, not closing them. Our objective is to answer questions, so we should only close when we know for sure that the question isn't a good fit for the site. Yes, this could leave "bad" questions on the site, but keeping a "bad" question is better than losing a "good" question.

1
  • The rule of thumb is: if in doubt, skip I think that's the piece of advice I'm most fearful of. The best analogy I can give is this: my friend describes himself as a "devout agnostic" believing both atheists and evangelists have too little data to justify their own belief. He feels certain that nobody can be sure. That's how I feel about these questions: I'm not "in doubt", I am certain I don't know and pretty sure the close voter didn't either. On balance I'd prefer to vote "leave open" unless there's a smell the close voter themselves knew something I didn't.
    – Philip Couling
    Sep 9 at 10:47
1

I leave questions open unless I know for sure that they should be closed.

Sometimes I'll put a reason for recommending closure so that others who are asked to vote have the possibility of seeing a reason for jumping one way or the other. (It might be a wrong reason, but if I leave one it's usually a considered opinion.)

I have also encountered questions that are flagged for closure - and indeed, sometimes closed - because the subject matter is a strange but relevant corner of knowledge. It's a fine line between voting to close because the question seems unclear and skipping because the question has (very) specialist subject matter about which a voter knows nothing.

You must log in to answer this question.

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