I have seen a lot of questions closed for being off-topic, that are only off-topic by the broadest of interpretations of the rules. Specifically here. This is an example of a question that might be construed to be within the rules.

Specifically pertaining to "managing the software of ...workstations" and "tools used for...automating these."

He was directed to Stack Overflow or Super User, but I can just as easily see the same question getting rejected from both of those sites because:

  1. He's trying to automate things for the benefit of other users presumably in a corporate environment. On Super User, corporate questions or questions that might be assumed to be corporate questions are directed here.
  2. He hasn't mentioned programming at all. There's no way a question that doesn't mention programming would last on Stack Overflow.

So, why are we so strict? Why do we reject questions so quickly? Why can't we give askers the benefit of the doubt and assume that it isn't off topic on edge cases?

Edit: Referenced question has since changed. Original question was:

I need to put a bootable image of a usb stick on a website so that our users can download it and burn it to USB themselves.

They aren't especially technical, so I don't want them to have to download any other software (e.g. Rufus / MagicISO etc).

My ideal is for something like the Windows 10 Media Creator (https://www.microsoft.com/en-gb/software-download/windows10).

I've seen some posts suggesting WinToFlash's Boot Replicator (https://wintoflash.com/home/en/) but the site flags up as malicious.

Is there a tool I can use that is safe and will create an exe which users can run to download and burn to a bootable USB?

  • 6
    We are not strict enough. We let too much crap through and are now victims of broken windows. – user9517 Sep 28 '16 at 7:45
  • 3
    Since you're new here, I just want to explain that downvotes here on Meta don't work the same as downvotes on the regular site. Votes on Meta don't affect your rep, and they are often used to indicate agreement/disagreement, instead of as an evaluation of the quality of the question. – Jenny D Sep 28 '16 at 7:54

In essence we're strict on topicality rules because any Q&A platform with only questions and no answers, is useless.

Unfortunately there are many more people asking questions, many new ones every day even, but far fewer people actively answering them.

Therefore the site rules, the badges and points, the community moderation tools, the actual moderators, they all exist for a single purpose: to retain the most valuable resource on any Q&A platform; the people actively answering questions.

They, I, we want to see interesting questions about our chosen topic, our profession even.

People want usefull answers and our mantra is good questions get good answers. Not so good questions get put on hold to allow the OP to edit and improve (or get migrated to a more suitable site) and the bad get voted down and deleted.
Admittedly you don't always need a great question to generate great answers but the number of questions coming in requires us to lean towards being too strict rather than too generous. But please flag and/or vote to reopen any closed question that you do want to write that great answer to. Those options exist specifically for that purpose and should be used. Please leave comment when you do so to clarify that.


We are strict on the rules because we are not the help desk for the internet. Our site has a very clear-cut scope: Professional system administration. As the question you linked was originally written, this sounded much more like a developer who wanted a way to distribute their latest product to the world and this is not our topic. The rewritten question made the system-administration aspect more clear, so it got reopened - I am still not convinced SF is the best place for this, though.


The question you are linking to is a request for a product. It clearly says "Is there a tool I can use [...]". Asking for product recommendations is clearly stated in the Help Center as being off-topic.

If you do consider that questions that are only slightly off-topic should be allowed to remain, perhaps you could point to some that are actually only slightly off-topic, instead of one that is clearly and blatantly so?

  • -1: he is looking for "a tool" in the sense of "I need to achieve a goal and it seems I might need a tool similar to those tools that I tried and that did not work", not in "what tool of those 10 is the best one?". You have to accept that some tasks just need tools to do them, it is not feasible to say "we cannot tell you which tool you need, just what you would need to do if you had to program it yourself!". It is good to keep it broad with general concepts, but sometimes you just have to do your task and need a tool that works. – user121391 Sep 28 '16 at 7:49
  • 3
    Yes, and in those cases you should go to a site that does software recommendations, not one that considers them explicitly off-topic. The fact that sometimes a recommendation is what you need doesn't mean that this has to be the place for it. – Jenny D Sep 28 '16 at 7:53
  • Administration is much more reliant on tools (as all programs are tools in this case, even OS and included software) than for example Computer Science (where tools are only helping, but strictly never needed) or even programming. What you are saying essentially equals to "If your problem does not fit our rules set in stone exactly, go away!" - Which is ironically exactly what Andrew Hendrix criticized. He asked why the rules are so strictly enforced, you say that they are (which is already known, hence this topic). – user121391 Sep 28 '16 at 8:10
  • 1
    Yes, that is what I am saying. That is what the entire help section, which I helpfully linked to in my answer, says. If you think the rule is bad, you should propose a rules change, not propose that the rule remain and people just pretend it doesn't. – Jenny D Sep 28 '16 at 8:13
  • I view this question as a proposed rule change, do you not? He asks "Why can't we give askers the benefit of the doubt and assume that it isn't off topic on edge cases?" which means "Why do we need such a strict rule?" Not all proposals have to start with "I propose the following:" - sometimes it is enough to say "guys, I think we could improve X, what do you think?", where X is a rule (that might or might not be changed depending on outcome of the discussion). – user121391 Sep 28 '16 at 8:29
  • 3
    @user121391 I disagree. If you want to propose a change, be specific. If you're not specific in what you want to change then Jenny's reply, which is just stating that she disagrees with the basic premise of that original question being only slightly off-topic, is a valid response. – Reaces Sep 28 '16 at 8:41
  • As I have said, this is not an edge case. You may want it to not be off-topic, but if so, you need to actually change the rules for what is and is not on-topic. Again, if you want to change the rules so that this question becomes on-topic, the thing to do is to suggest that the rules be changed, not ignored. – Jenny D Sep 28 '16 at 10:16
  • Actually, either ignoring or changing the rules results in the same outcome, see also my comment on my own answer regarding that. Andrew main argument is that any rules in the real world always produce a certain amount of edge cases and therefore changing of the rules achieves less than ignoring them (being more lenient). Also, you have stated multiple times that this was no edge case, but never why. One can only assume from your answer that all requests for tools are off-topic, but not all are closed. In fact, most remain open (and rightfully so, as they are useful). – user121391 Sep 28 '16 at 11:50
  • 6
    @user121391 I am sorry to have failed to express myself in a way that you are able to understand. Since direct statements with links to the list of what is considered off-topic is not clear enough for you, I am at a loss how to proceed, and so I give up. – Jenny D Sep 28 '16 at 12:25

I have been absent from SF for a couple of years and just recently started getting back in and I have two observations:

  • The number of posts that are off-topic has gone through the roof which isn't bad per se because it means the site has some more influx of new users and people are somehow getting here.

  • OTOH the (for me) new review mechanism fights those questions with fire. It seems to work to keep those questions at bay but it seems to make migration very quick which wasn't always that way. Looking back we also had questions that seemed off topic (and many were) but we commented on that the author had a chance to rephrase and show that it was in scope before it was migrated.

I have reviewed some questions myself and many times when a question got closed as off-topic there was no comment and no explanation given to the point where I had to read the question with my rules-lawyer hat on to see why exactly it was closed.

  • As a new user, I also experienced what you describe when I first started here about a year ago. Interestingly, it is very different on other sites, for example on Unix&Linux, the spirit is generally very helpful, the rules are used for really bad cases and it helps new users to feel more welcomed as well as increasing the quality of answers (and questions, later on). – user121391 Sep 28 '16 at 7:52

Product recommendations have pretty much always been off topic on all of the stack-exchange network - there's some notable exceptions recently with software recommendations (which I was a mod-pro-tem of) and hardware recommendations. Keeping quality and relevance of software and hardware recommendations is hard.

Reading through the original question as posted.

I need to put a bootable image of a usb stick on a website so that our users can download it and burn it to USB themselves.

Is a reasonable statement of the problem .

They aren't especially technical, so I don't want them to have to download any other software (e.g. Rufus / MagicISO etc).

This is... basically magic.

My ideal is for something like the Windows 10 Media Creator (https://www.microsoft.com/en-gb/software-download/windows10).

Which is an additional piece of software which the user needs to download and run. Often unsuccessfully

I've seen some posts suggesting WinToFlash's Boot Replicator (https://wintoflash.com/home/en/) but the site flags up as malicious.

See previous point.

Is there a tool I can use that is safe and will create an exe which users can run to download and burn to a bootable USB?

Is a software recommendation.

So, essentially we have an unanswerable question that's pretty much off topic on all the sites you mentioned. With cleanup, this may be on topic on software recs, but this closure seems to be the system working as designed.


If you look at the edit history, you notice something interesting: the edit which removed the off-topic designation did add very few useful stuff to the core question. It detailed why the asker needs to solve his problem, and it describes his surroundings. Those are not needed for the question itself in any way.

Of course it would be helpful to know what he has done already, so this edit is not completely useless, but that info might just be asked by comment ("Please add what steps or alternatives you already tried and what did not work") instead of a slap with the rule book.

I have seen several of those off-topic remarks in the past, but I always wondered what it should achieve. Professional system administration is extremely broad and also vague. So where do you draw the line what is included?

  • Limiting it to money/employment/freelancing services would mean to forbid perfectly good questions about managing a cluster of 64 virtual machines to deploy CI on a home lab that is up to best practices (which is perfectly possible with todays commodity hardware and open source software) and to allow someone who asks trivial questions because he is lucky to be paid for his incompetence.

  • Limiting it to difficult questions would mean someone would have to decide what is easy and what is difficult - an almost impossible task, as every situation is unique and everyone can learn and better themselves. You would end up with either no hard questions or with just them.

  • Limiting it to scope (size of company, amount of affected systems) is also problematic, as it excludes administrators of small companies, niché software/technologies, or problems that only surface on some machines.

  • Limiting it to things already covered on other sites of the network is futile, as you would have to exclude all database questions, everything about Unix systems, everything where own code is envolved, all website stuff, all network and security questions and even everything that is done on a single client PC. You could effectively rename the site to "Active directory and mail exchange", because that is all that's left.

So you end up with only one reliable measurement - quality of questions. This can be measured approximately in length, details, similar questions and upvotes. Best of all, it encourages the asker to edit/improve the question to get more upvotes and to stay visible. It also needs much less mod activity, so less work and more time to answer other questions.

  • 3
    We have changed our scope and reduced our quality requirements several times over. All it does it allow more crappier off topic questions. The end result being few people bother any more and the river of shit becomes Amazonian. – user9517 Sep 28 '16 at 8:48
  • 2
    You could conceivably ask any IT related question on Server Fault and expect to have it answered. That would make us (like many other places) an helldesk for the internet. That is not our mission. We limit the scope to a business environment so that we are not overwhelmed by crappy questions from home users and the hoards of cluelessness that surrounds Stack Overflow. Sadly we have failed. Hanging on in quiet desperation... – user9517 Sep 28 '16 at 8:52
  • 2
    What Iain said and also: Anything in a home environment is clearly off-topic, "best practices" or not. – Sven Sep 28 '16 at 8:55
  • @Iain This is a very negative outlook that would ultimately either destroy you or lead you away from this site. I think it would be better to reward good, interesting questions than to try to destroy or prevent bad ones. It is difficult to reliable say wether a question is good or bad, it is much easier to judge the effort that went in it. – user121391 Sep 28 '16 at 8:57
  • @Sven Yes, this was the problem I wanted to highlight (and what the asker also said) - the rules are too strict and exclude good questions by strict enforcement. To solve this, either the rules could become more differentiated (my proposal) or the enforcement could become less strict (Andrew's proposal). – user121391 Sep 28 '16 at 8:59
  • Broken windows theory. I search every day for good interesting questions that are well researched. I am always disappointed. – user9517 Sep 28 '16 at 9:00
  • 1
    Evidence of broken windows theory in action. We once allowed CPanel questions with the intent of gathering Q&A for sysadmins unfortunately all we did was bring in the clueless pointy click crowd. Now we can't get rid of them. – user9517 Sep 28 '16 at 9:03
  • @Hanginoninquietdesperation Broken windows theory is only that, a theory. I agree on your second point, searching for good questions is difficult here, but I think it could be improved by tools. For example, fuzzy tag search over multiple sites, include and ignore lists of users for search, generally stored search criteria, tag hierarchies/trees, multiple separate upvotes for effort/content/etc. and searching for them, ... that just from the back of my head. Similar to the web, I want to find what I want to see, not tell others they should do how I want them to. – user121391 Sep 28 '16 at 9:07
  • 3
    @user121391: I disagree with both you and Andrew. We (well, I) want questions that are good and within the scope of our site and I don't see why we should widen our scope or loosen the "enforcement" of our rules. I am fine with excluding home users - they have other places to go to. – Sven Sep 28 '16 at 9:09
  • 6
    @user121391: As a mod, I have zero interest in becoming a secretary for lazy idiots and do their work by tagging for them. – Sven Sep 28 '16 at 9:12
  • 3
    There are lots of things that could happen that don't. People could read the site help before posting. People could search the site before posting. People could use Google before posting. People could write interesting, well researched questions but none of that happens either. – user9517 Sep 28 '16 at 9:20
  • 5
    Closing crap questions is a social good. – user9517 Sep 28 '16 at 9:21
  • 2
    @user121391: We have discussed this again and again - just iterating this again is pointless. The majority of our active users want a limited scope with more or less strict enforcement of our arbitrary(!) rule set. – Sven Sep 28 '16 at 9:27
  • 2
    @user121391: We are a bit of an odd thing in the Stack Exchange universe with being one of the very few communities that are decidedly not open for everyone (the others being MathOverflow and Network Engineering, I believe) and I am not sure what the company thinks about it. Certainly they don't actively cater to our "special needs" (I am OK with that...). – Sven Sep 28 '16 at 9:36
  • 3
    @Reaces: Your Q&A ratio and rep proves that you are neither lazy nor an idiot (obligatory disclaimer: At least as far as this site is concerned) :) But seriously, my remark was purely about the semi-regular idea to let mods do some thing or another "to improve our site". Remember the difficulty class triage idea from a few weeks back? The occasional edit for a missed tag isn't a problem, but I would be weary of having to actually read every question to check if they are appropriately tagged. – Sven Sep 28 '16 at 15:46

Here's some feedback...
So, I asked a good question about Windows Server 2016 for a small business client. Normally, I'm a Linux person, but they wanted the latest Windows.
I got a bunch of downvotes and was told to go Google the answer. (Which I did a bit of research prior to coming here, but Server 2016 has been out only a short while.)
If that's the kind of answers people give, it is no wonder there are issues in the quality of both questions and answers.

  • If you have an axe to grind then you should open a new meta question and link to the relevant question n main. – user9517 Oct 29 '16 at 0:38
  • There's no sign of the question on the main site, so its difficult for anyone to comment on this. – Rob Moir Oct 29 '16 at 12:50
  • I got a bunch of downvotes, so I deleted it for fear of getting more dv. – MikeP Oct 29 '16 at 16:56
  • @RobMoir serverfault.com/questions/811761/… – MikeP Oct 29 '16 at 16:57
  • Despite your protestations your question isn't very good it's vague and lacks detail required to answer it.. You don't provide any information on the expected workload of the system so, as it stands you are asking for opinions and we don't do that. If you provided workload information it may be better but it would then stand a good chance of being a capacity planning question which again we don't do. – user9517 Oct 31 '16 at 5:50
  • @Hanginoninquietdesperation, workload is not relevant, in my opinion. No further detail should be necessary if someone knows what ZFS is, or how Windows does (if it does) SSD Caching. The question is a question of what works. Again, my point is there were several DV with no comments, no suggestions, no feedback, other than one useless comment. Thus, my point and experience matches the OP that the community is not helpful. – MikeP Oct 31 '16 at 20:27
  • Of course workload is relevant. It's particularly relevant to the first 2 questions you asked. for example if you want to run a database then it's likely you will want it on the SSD which might preclude their use for other stuff. – user9517 Nov 1 '16 at 5:53
  • @Hanginoninquietdesperation, workload and system function are two different things. The function will mostly be an all-in-one server, mainly file server. – MikeP Nov 1 '16 at 15:20
  • I don't have the rep here to view deleted posts, but generally speaking, show your research. If you have researched your question beforehand, make sure that research shows in the body of your question. Otherwise, for everyone else it's as good as a random idea. – user Nov 17 '16 at 16:36

You must log in to answer this question.

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