Has anyone figured out why the staging ground is showing "page not found" for some users? I have more than 500 rep and have asked more than 2 questions in the past and still getting the same issue.

@anupsabraham Do you have an example? I haven't encountered that yet.

This is what I see when I go to staging ground link.

Ah, I see. I thought you were visiting a user's profile from staging ground. Sorry.

To answer your question: currently SG requires 500 rep and two well-received questions. And well-received is defined as score of 2+ or has an answer. Looking at your profile you have both - you have a question (answered) with score of 2+ and you have two additional zero-scored questions that still have an answer. That should have covered the requirements. But yet it seems you're still not qualified. Might be a bug.

The 2 question requirement is being dropped soon (this week, presumably), so hopefully that helps.

Ah ok. Cool.

Thanks. I'll wait. Very interested to see what this is all about. lol.

Relevant: What are the actual requirements to be able to review Staging Ground questions?. I'd suggest you wait a day or two until the question requirement is dropped. Then if you still don't have access - post a bug on meta.

It seems I only have one well-received question. Trying to learn what that means right now.

It's an entirely new metric that only showed up for SG. See the post above which is the first time it was explained. It means score of 2 and above or alternatively the question has an answer. You have three questions that fit this criteria.

@VLAZ can take their main profile URL and add "chat." subdomain

So I reviewed this question as major changes since it seemed that the same question was being posted by multiple users (Who might be either related or the same person). I am wondering whether that is appropriate or whether I should have reviewed in some other way.

@AbdulAzizBarkat ¯\_(ツ)_/¯ The problem is that unfortunately I don't see why it'd matter either way. If you publish or not, the user can just re-post it again. And has a chance to get to the live site. Quite a good chance. I'm not sure what portion of questions got to SG but if it's, 10% then they have 90% chance to avoid SG on repost.

Maybe "the intention" from SE is that you'd review it as any other question. Ask for specific data, give some feedback on how to improve it. Yet that's likely to just be ignored and you'd have wasted your time. While OP just re-posts it.

After I looked at it more - I modflagged it. The multiple user aspect is fishy. So, I asked for a review there.

@VLAZ I'd flagged it along with the review as well, I actually happened on those two while searching for duplicates and the deleted one and the comment there made me suspicious.

Quick question, would a question asking what to use in a Dockerfile for Windows 10 to be the base image be on-topic (Someone voted it off-topic)? This is can be answered stating that there's no Windows 10, etc. images (Wouldn't make any sense really) and they'll have to use a Windows Server image.

What was the off-topic reason? With the proper amount of mental gymnastics it fits down two pipes as far as I can imagine; "Not a programming problem" and "software recommendation". I'm not the biggest fan of shutting down Docker questions just because they are in the grey area of devops, but I wouldn't personally vote to reopen a question like this either.

One of the problems I see with SG is borderline on-topic questions being marked as off-topic and blocked like that. For example I think this is on-topic since it seems to fall under the "software tools commonly used by programmers" (AWS Cloud9 is an IDE) but it is closed as off-topic and a re-evaluation was declined.

@AbdulAzizBarkat I believe you can override closure in this case and push an approval through, or switch the status to major changes instead. An alert modal will pop up asking if you're sure, and saying "yep I'm sure" will reopen it. I haven't tested this on a question that is not in re-evaluate status, though, so...

If this bug meta.stackoverflow.com/q/430588 gets fixed, their votes should carry over after publication (or flags if they're <3k rep) (which you're aware of, just stating for the transcript)

@Spevacus I can do that, yes. Just waiting to see if anyone has different thoughts on why it is off-topic (In which case I'd leave it be) or whether we need any updates on it before publishing.

I chucked an edit at it. I think it falls under "tools used by programmers", but I'm unsure whether enough detail is given to solve the issue. The error hints at something being wrong with the setup and without knowing how that looks I dunno if someone can meaningfully answer. A google search on that error message doesn't turn up much, which suggests it's either not a real error message (i.e. it wasn't copy-pasted), or it's just a poorly-documented error.

But I'm no SME on that AWS stuff, so

Borderline on-topic questions being closed as off-topic is not specific to SG, that is a site-wide thing. The "software commonly used by programmers" as documented in the stackoverflow.com/help/on-topic is being highlighted here, but I must draw the attention to the fact that this is followed by "; and is a practical, answerable problem that is unique to software development". AND. Not OR.

I'd say it's unique to software development. Practical and answerable? Maybe with an edit?

You use the right words, "you say" :) In my experience curation is starting to be more stringent on what is a programming problem nowadays. Questions about configuration for example start to be closed more often.

Not exactly jumping for joy to be honest.

@Spevacus Me neither (At least about Cloud9, I've used various other AWS services)

@Spevacus The issue is that "DevOps" is considered, to an extent, as a separate discipline from programming. However, devops also has some overlap with programming. Docker being a very good example. So, Docker as a tool is on-topic for SO but not all possible questions. It's a bit fuzzy and I honestly can't tell where the line should be.

Topicality is a ✨spectrum✨

"I'm trying to administer a docker container to do <devops-y thing>" is off-topic, for example. Similar to saying "I want to administer an RDBMS". So, lots of Docker questions rely on details. And I am not really using docker, so I find it hard to judge which would be on-/off-topic.

@Spevacus a veritable rainbow

@VLAZ I get the point, but the question being discussed isn't really about DevOps? AWS Cloud9 is an IDE.

@AbdulAzizBarkat IMO Docker/writing Dockerfiles is on topic but then there's k8s questions which aren't according to the tag wiki but then what about GKE which is a collective for that kind of issues as well?

@dan1st GKE is basically K8s hosted on Google Cloud, you'd judge the on-topicness of GKE questions similar to how you'd judge on-topicness for K8s

like I also wrote a k8s question in the SG during the beta that I later deleted because I decided it was off-topic because k8s config and then reported something about the SG for askers resulting in people telling me it would be on-topic

@AbdulAzizBarkat idk how to judge these k8s questions

IMO just having K8s is not enough to call a question off-topic

Because the containers end up running code.

And often times issues happen due to the combination of how the containers (pods / deployments) are setup and how that interacts with the code.

I recently saw a question about passing a Spring secret from k8s showing that k8s gets the secret - I considered that on topic as a Spring problem (happens inside a pod)

From what I understand the community considers questions purely about K8s manifest files off-topic since that is considered configuration.

Today, yes that is also my observation. 10 years ago, such questions would have been no problem.

@dan1st Yes, I'd consider that on-topic. Anyways, I guess this discussion is going a bit off-topic for this room. Maybe we can hop into another room if you want to continue that discussion?

@VLAZ FWIW that was supposed to be changed the week following June 5th, though i suppose it's just running behind

there are 3 days left,

:shrug:

@AbdulAzizBarkat MY apologies, I probably missed some of the transcript. I thought it was docker. Yes, Cloud9 is definitely a tool used by programmers.

@Gimby This mildly annoys me because you're right - 10 years ago it wouldn't have been a problem. Also, 10 years ago "devops" wasn't much of a field. But in the last 10 years, not only did it establish itself (which is fine) - the programming field tried to push this discipline away. On SO that results in "devops" questions being considered off-topic. But also has real-world impact.

I've seen programmers IRL claim "That's not something I should do" and...if you develop an application but don't deploy it, what good is your code?

IMO, if a developer doesn't or cannot deploy their code, they haven't really finished their job. It's like if a cook says "OK, I've put all the ingredients in, and the thing is in the oven. My work here is done - somebody else should take over now".

@VLAZ tbh I can understand it being distinct in many cases

@VLAZ Same, I find the disconnect between the Ops role and the developers weird. For some reason both like to push away their work or don't want to understand even the high level stuff going on in the other end.

I think it is more that because there are so many questions and fewer curators, there is also less leniency. It must be closed, at all costs and with the least amount of effort involved. Hence people have started to lean more on things being "code" rather than "software engineering".

@dan1st I don't want to say devops is 1:1 with development. However, there is a significant overlap and, IMO, most developers should be able to handle at least simple devops tasks.

@VLAZ Yes but is k8s simple? I doubt it

Kubernetes scares me :)

@dan1st Depends on your usecase. If you go simple its really simple

@AbdulAzizBarkat My use case is writing 10k lines of YAML

@dan1st for real?

no

@dan1st You might want to consider something like Helm then. Although it generally bloats up the size of the YAML if you're writing those templates yourself.

But its great if you want to make manifest files in a flexible / reusable manner.

@AbdulAzizBarkat and the simple part is out of the window

@dan1st Not really, if you're using existing Helm templates, it's simple. If you're the one creating them, not so much.

I have my own custom stuff (for an application I wrote) that's also having multiple services in different namespaces on the same IP. Is it simple? No. Does it work? Yes. Do I need Helm? No. Would I consider most of it on-topic on SO? Not really (any more).

What's the opposite of scope creep? Scope... recession?

eh

if we're gonna allow some devops we ought allow all of it

Also it might degenerate in a tag like R. Who knows?

I had no experience with R on SO. Then I reviewed R questions in the SG. Now I mostly skip them.

R is for Rough to Review :/

@KevinB That is certainly an opinion. If you do that, the flood gates are opened for a whole host of questions for which the average visitor of Stack Overflow is ill-prepared. If you only allow some, well that's just the misery we have now where everyone makes up their own rules about what's on-topic and what isn't. Hard to really improve the situation either way.

that's fair, but would it also bring in a bunch of people who would be able to also assist with non-devops questions?

I think that station has passed. Stack Overflow is not hip anymore.

all the more reason to expand teh scope

if the userbase is shrinking, and the goal is to reverse that or slow it, shrinking the scope certainly wouldn't help,

I'll give you that - I wouldn't know. I have never been able to fathom the scope of Stack Overflow as it is; too flipping huge. The only thing I do know is that I really would like there to just be less sticks up butts :)

why the heck did they break Ctrl+K in the new editor

so annoying

because it's a text editor, not a programming question editor

it's built for teams

all it does is wrap the selection in single carets. least it could do is insert a new line above and below with three carets each

@KevinB Isn't that kinda aimed at dev teams as well?

not necessarily

they want entire businesses, large ones, to use it since they pay more

@dan1st Nah, is sold for essentially internal knowledge bases. That could be developers but it's not aimed at them specifically. Could be, say, to document stuff in an accounting firm or whatever.

@VLAZ Well, who knows the name Stack Overflow?

If it's aimed at developers it's woefully underequipped. You can't even indent/unindent code in the new editor with Tab/Shift+Tab. You can't do it in the old one, sure, but at least in the new one they should have done it.

idk

Here's their marketting page - the customers are varied stackoverflow.co/teams/customers

btw do you agree that things like this is off-topic?

> Java has compressed oops

Not like other expanded OOOOOPS

w-oops

@dan1st For that one - I'm not even sure what the question is. Is it "is this a good idea" or maybe "What were Java and NodeJS's designers thinking?" Both are off-topic-ish. I'm not really that well-versed with pointers and language design but I'm also not sure what OP describes as limiting factors are actual limiting factors.

@VLAZ It's "Why do Java and JS use compressed pointers limiting heap sizes to 32GB instead of the approach I'm describing (relative pointers)? What are the disadvantages of my approach?"

I think asking about the disadvantages is on-topic but it should be specific about that ig

OK, so more language-design-y then. I think it should be at home on PLDI. But needs rework for SO. Rework that won't really make it better, though. If it goes down to "what are the disadvantages of X" it might be a bit too "not practical" for some users.

Yeah questions like this are probably not received that well

But is it really off-topic? Can it be improved to a point where people are ok with it? idk

I personally think it's on-topic. Even with "language design is POB" I don't think that's really the case here. With that said, I also recognise I'm hardly an authority. Moreover, my personal opinion doesn't matter - I think the question would be seen by enough people who'd think it's off-topic.

Yeah I think that's a thing of trying the best to make it ask about the specific details and see what happens

And I feel like attempting to "improve" it won't really help as it probably needs to drop some of the more relevant parts. That or be really well-written but I mean above and beyond a good question.

Well I think background research from the OP could be useful but at that point, are they answering their own question with it?

I dropped a comment pointing them at PLDI.

Well maybe this comment here could be useful as well

@VLAZ ah that's what you meant with PLDI lol

Is it really a programming language design question or is it more of a platform question? Are platform questions on-topic on that site? idk

@dan1st It's for designing new languages as well as examining the existing designs. I'm not very actively following the questions there but, IMO, this one should fit. Probably a question like "Why Java chose to call it "boolean" and not "bool" - what were they thinking?!" won't fly.

Well it isn't really about the Java/JS languages but more about the JVM/NodeJS runtime(?). I don't know about whether these things are on-topic there.

well JIT compilation and things like that are on-topic so that's probably fine

Yeah, still on topic. I was more active during the beta and one of the big thing I remember that was discussed on meta was VMs (as in JVM) being on-topic and desired topics. At the start a lot of the questions were for syntax and some users were frustrated there weren't more advanced topics than that.

Yeah JVM is on-topic, there is no doubt about that

But why questions are - difficult

Like the off-topic reason would either be "opinion based" or "Other"