In the nearly 4 years that I’ve participated on Server Fault, it’s been a community that I’ve been proud to be a part of, and I’ve been able to build my own skill set a lot and help people out at the same time by answering questions. Everyone who’s helped to build this community should be proud of what it’s been able to do. We really have one of the best places on the internet to find answers to real problems that sysadmins face every day, instead of just inscrutable documentation or forum threads filled with crappy cargo cult solutions.

For most of us, the profession of system administration is fundamentally changing: instead of duct-taped together automation scripts, we have a huge ecosystem of configuration management tools; instead of applications running on one server in one data center, we have applications running in distributed data centers or hosted in the “cloud”; instead of just needing to be an expert in an operating system, we need to be able to solve problems with a myriad of OSes, applications, databases, and protocols.

All of these new problems need new solutions, and require sysadmins to reach outside of our comfort zones in lots of new ways. Instead of being able to kick the network issues to the network person, our systems need to be aware of the network, have the network configured in the right way, and even manage the network as software defined networking starts to catch on. Instead of getting an application out of the box that just needs a next-next-done install, we need to be able to understand how best to architect a solution for our specific environment using several different software solutions, and how to integrate those pieces of software.

We need to be able to learn and evolve extremely quickly to keep pace with technology and the state of the craft. Staying up to speed is incredibly difficult in this profession, and getting in the door is getting harder all the time. For me, Server Fault is a great resource for maintaining my own knowledge and improving it - accidentally referring back to a problem I solved years ago because it comes up in a Google search, or having coworkers ask me to help clarify an answer of mine they found while trying to figure out a problem (this happened at my last job), tells me that we’re really helping to advance the field of system administration as a whole.

But, I’m worried that we’ve gotten off track lately. There’s been a drop in participation through contributions - post rate, answers per question, and relatively active users (5+ posts per week) have all been trending downward for a while.

posters

Since June of this year we’ve seen a major jump in traffic - nearly 100% - due to changes to Google’s page ranking, with an accompanying jump in participation through community moderation (mainly close votes and down votes):

votes

...but, crucially, not an increased volume of questions over the same time frame; they’ve essentially stayed flat:

posts

There are a lot of valid arguments on why this is happening, but I think we can all agree that this is not a great direction to be going in.

Server Fault still gets a ton of bad questions, which need to be closed - and there are now fewer regular contributors among whom the load of “taking out the trash” is now spread. As a result, instead of participating through posting and voting on useful questions and answers, many of our most active users, the core of our community, are spending their time on the site looking for questions to downvote and close out of frustration.

We don’t want our best contributors feeling like the most important contribution they can make is to find stuff to get rid of - and more importantly, we want to avoid deterring people from joining the community and participating by being over-protective of what we want the site to look like. Narrow interpretation of the scope with rigid enforcement hasn’t slowed the volume of poor quality questions, but it has given Server Fault a rather hostile and insular reputation and a tendency to give a poor first impression.

I wouldn’t be comfortable telling a sysadmin I just met, especially a beginner or intermediate level one, that they should go try participating on Server Fault without warnings and caveats, and I suspect I’m not alone on that. We need those intermediate level people to be comfortable with dipping a toe into contributing, getting involved, to take the load of answering easy questions off of the veteran users who are tired of the simple stuff. This needs to be a place that’s welcoming of all system administrators, regardless of skill level, in order to accomplish that. Some of our veteran users will reduce or stop their participation, which is expected in any community; we need to have new contributors coming in to keep the community alive.

More eyeballs from Google should be a good thing for Server Fault; it means that this resource we’ve all created is reaching and helping more people, and it means more people who might take an interest in the site (and have expertise that would be useful to have around!) are getting exposed to it. Instead, the we’ve taken a harder line than ever to fight the perceived flood of low quality questions.

Server Fault should be a community that is welcoming of the easy questions. More importantly, it should welcome the kind of people who will be interested in answering those questions, who can can grow their own expertise while at the same time preventing our core contributors from having to deal with every easy question. It should be a community that changes and evolves alongside the practice of system administration, not one that gets stuck on a specific, rigid definition of what a sysadmin does or doesn’t do. It can’t survive and thrive and continue to be a great resource to so many people if we make it a community where only experts are able to ask a good question - after all, our experts are good at answering their own questions too, and rarely need to ask one of the community. Becoming more insular, more selective, more “expert”, is not an option that will lead to a sustainable future for Server Fault.

We should be encouraging questions on subjects like Docker and continuous integration, about trouble with configuration or troubleshooting on specific blank-as-a-service platforms, and about how to architect a server infrastructure for the NoSQL flavor of the month. Instead of trying to reject these questions and the people who ask them, we should push them in the right direction for how to be better at what they’re doing and how to solve their problem in the right way. Getting that knowledge out there, in front of the eyeballs from Google, is how we can really make a difference.

A few months after I started at Stack Exchange, the community team asked me to look at what I would do to try pushing Server Fault in the right direction. What I’ve arrived at is to remove the “professional” criteria from the scope of the site - it’s getting used too aggressively in closing questions, as well as interpreted to limit the set of topics that are accepted here.

Now, before anyone panics..

What I’m about to propose isn’t intended to lower Server Fault’s quality bar, or open the door to everything that belongs on Super User being on-topic here.

I’m asking for questions to be considered on their merits, taking away the close criteria of “infer whether this person has the right job title” because that can’t be what we’re focusing on when we decide if a question fits.

We still want to make it clear that if it’s for someone’s home network, it goes on Super User. And if it’s a weak question that can’t be understood or has no detail, I expect the close-hammer to fall just as fast as before. However, someone having problems with setting up hobby project on a VPS even if they aren’t a sysadmin by day, or a developer trying to work out how to deploy their application successfully to a robust server stack instead of their development machine, are the kind of questions that I want to be in scope for the site now.

This isn’t going to directly address the volume of low quality questions that we’re currently dealing with. What I’m hoping for is a change to the community’s approach to some of those questions: spending time editing the salvageable ones, giving users a push in the right direction instead of closing their question as quickly as possible in an attempt to avoid “broken windows”, or downvoting the question because they think the answer is too obvious.

Removing the requirement to be a professional means telling these users “your question isn’t appropriate as it stands, improve it by...” instead of telling them “you don’t belong here because you aren’t what I consider to be a professional sysadmin.” Over time, my hope is that we attract more intermediate-level contributors that we’re scaring away now who can handle the easy questions so the veterans won’t need to.

Question quality, and making sure contributors aren’t spending their time sifting through questions they aren’t interested in, is something that’s been getting a lot of attention on the Stack Overflow side of things, which can benefit Server Fault as well. Some of those efforts are applied here automatically, like rolling rate limits - we’re also going to test out how well some of the specific optimizations on SO work on SF, like the recommended tab, and see if we can tune them to work well here.

I want for Server Fault to be the great resource to so many people that we all know it can be, and the great community that I’ve known over the last 4 years, and continue to evolve as a resource for everyone in this profession - and I believe that it can do that as a place for everyone working on the kinds of problems that we are, whether they call themselves sysadmins, SREs, devops, or anything else.

My hope is that everyone in the community will join me in giving this a shot.

  • 8
    I was with you until I got the impression from you that SF should be SO and SU's helpdesk. – Colyn1337 Nov 21 '14 at 17:44
  • 10
    @Colyn1337 I'm not sure where you got that impression, can you clarify? I hope I made it clear that the expectation will still be that questions are on-topic, clear, and answerable. – Shane Madden Nov 21 '14 at 17:59
  • 4
  • 1
    @TheCleaner The plan is to make a change to the wording in the scope soon. As far as the duplicates, two of those are old discussions from 2011 and the third is this question (?), this is different in that it's a "here's something I want us to try" than an attempt to clarify the old scope; consider it an announcement if that makes more sense :) – Shane Madden Nov 21 '14 at 19:56
  • 1
    Gotcha. If the scope is going to change, fine by me, but please bear in mind that one of the close reasons that was "promoted" to be used over move to SU was the "not professional" close reason. So I would presume that close reason goes away entirely? – TheCleaner Nov 21 '14 at 20:15
  • 1
    @ShaneMadden I'd like to see that close reason and the "minimal understanding" reviewed. They feel awfully close to the same thing. Perhaps ask a new Q on mSF as to what the wording should be? – Chris S Nov 21 '14 at 20:18
  • 1
    @TheCleaner Yeah, there will need to be some housekeeping there to prevent confusion. – Shane Madden Nov 21 '14 at 20:24
  • 6
    I see that roughly 850 questions a week are being asked, but I think the most important metric is also the one that's not easily quantifiable. Sure we have 850 questions a week like we did last year, but the discussion centers around how those 850 questions are seemingly of a decreasing quality, and rightly earning their downvotes and closures. Gone are the days of casting disparaging glances at ExpertsExchange. Gone are the days of attracting experts. It seems like this has become a "think of the children!" discussion where we're supposed to onboard brand new admins and appeal to non-pros. – Wesley Nov 22 '14 at 0:32
  • 4
    ...and maybe that's the way the culture of the site is shifting. Maybe that's okay. Maybe we should re-write the FAQs, and mission statements, and start from the ground up. ¯\_(ツ)_/¯ – Wesley Nov 22 '14 at 0:33
  • 12
    I just want to make it known that I strongly disagree with this approach because it will not only not help attract better questions but instead it will encourage people to post even more crap. And no, I don't think we need to try it because we have edited around our various FAQs and help texts and whatever for years now without any positive effect. The only thing that would help is that SE would acknowledge that their "everyone can ask" approach isn't working everywhere and implements changes around this. The fact that you as an SE employee only came up with this tells me that won't happen. – Sven Nov 23 '14 at 15:49
  • 5
    @Sven we have edited around our various FAQs and help texts and whatever for years now without any positive effect All of those edits have been angled toward "how can we make it clear to the people we don't want that we don't want them?" This a different approach - I understand if you have doubts about how well it'll go, but it is a different direction than what's been tried in the past. – Shane Madden Nov 23 '14 at 20:49
  • 6
    @ShaneMadden: This edits where a consequence of the problem we face for years: A steady steam of terrible questions by people who don't care at all and just want to dump their crap into the first textfield they find. Your approach doesn't help at all with this - on the contrary, I fear, as it will remove a (weak) barrier that was erected for a reason. In the end, I would like this change a whole lot more if it came together with effective tools to block crap from being published in the first place, but that would require SE to acknowledge the problem and commit to try to solve it. – Sven Nov 24 '14 at 17:29
  • 7
    @Sven The idea is to drop the professional requirement, since it's being used as a hammer to close a lot of questions that are perceived to be from "non-professionals", whether they're devs or sysadmins who are asking something that we think a professional wouldn't ask. We should be handling Qs on whether they're topical and answerable, instead of trying to make all questions that come in be interesting/complex problems (and punishing the ones that aren't). Using the "professional" reason how we have causes us give a bad impression to potential contributors, and doesn't stop bad questions. – Shane Madden Nov 25 '14 at 9:43
  • 4
    'hostile and insular reputation and a tendency to give a poor first impression' - this. as a beginner coming from a programming background, I have asked some bad questions that deserved their fate - but the atmosphere here has dissuaded me from asking more that could have been better, or contributing otherwise. I think there is a wealth of expertise here and it's a shame to see it used mostly to delete stuff that isn't considered 'professional'. I'm glad this is changing and that from the other answers, alot of others think it should change too. – jammypeach Feb 12 '15 at 14:00
  • 1
    My $0.02 on this is that the poster is correct. Many sysadmins have a (deserved) reputation as insular, near-autistic pr*cks who won't even talk to you unless you can convert ASCII characters to UTF-8 in your head. Sad to say many of them seem to hang out here, waiting for someone to innocently ask a question so the sysadmin can slap them and scream "STUPID!" That will kill this site eventually. – user1071914 Sep 5 '16 at 17:59

I'll be the first to admit that I've stepped away from SF for a while. Those reasons are my own, and everyone should be able to step away when they want to.

That said, I've been looking at the site more recently, partly because it's been brought up in team meetings and partly because I miss answering questions and helping people.

This is not the community I have great memories of. This is not even the community that I left. This is now a community full of hate and isolationism. Personally, I was very disheartened to come back and see that state that this community has changed into.

We are not helping to make the current, future, and past generations of system administrators better here. People should not be shouted down because they are perceived to be not good enough. New people to our profession should be welcomed with open arms, taken under our wings and taught how to become a good admin in today world. We all had basic questions when we where young admins and there were people who guided us. Now that we are sitting at or near the top of the mountain, all we are telling the people who are not as skilled as us is you are not good enough, you are dumb, go figure it out for yourself.

For me, answering questions here, and on other forums in the past has been a way for me to give back to the community that made me the Admin I am today. To help mold the future so that our profession can come out of the dark basement, can be better than what we are now.

If we continue to work the way that we have then all we are doing is killing this site, and driving the sysadmins of the future somewhere else. And I believe the SE network is the best place to get amazing answers to your questions (not just because I work here, I believed that before I was hired).

  • 26
    I feel much the same way. It saddens me to see so many people basically being told "You're too dumb, go away from me before I catch your dumb disease". I, for one, am guilty of doing this in part. This was wrong of me to do this. I back Shane's proposal for the same reasons you do. I want to give back. But I want to give back to a community that I am proud to be a part of. – Mark Henderson Nov 21 '14 at 20:25
  • 2
    I wish I could have worded my disappointment as well as George did but I think I would have slipped into a rage based rant. So I'll just say "Hear, hear!" – Kyle Brandt Nov 21 '14 at 20:54
  • 11
    I'm gonna hazard a guess and say that you didn't get where you are by asking "gimmie teh codez" or other barely intelligible collections of words, and those requests are what we deal with most these days. We're not doing the sysadmins of the future any favors by answering those types of questions, or letting them clutter up the front page, either. Learning how to ask good questions and communicate effectively, however, is important, and we'd be wrong not to reflect that in our standards. (Which, frankly, seems to be the direction this whole thing is heading in.) – HopelessN00b Nov 21 '14 at 21:47
  • 3
    It's amazing what comes lout of the woodwork when there is a bit of a flap on. It's sad that y'all haven't seen fit to play the game or deal with the (apparently) obvious issues before now but I look forward to seeing you improve the place, Kyle, George and Mark. – Iain Nov 21 '14 at 22:41
  • 18
    @HopelessN00b I'm not saying cater to these people at all. But on any given day I can go on the front page and see people giving someone who is trying to learn, or doesn't have perfect English a hard time for trying to improve themselves. That is what makes me sad. – Zypher Nov 22 '14 at 0:06
  • 2
    @Zypher There are occasions where this happens and it is handled poorly. That said, there are also many occasions when someone is clearly in over their head and there really is nothing more to be done than inform them of this. Helping someone make the next step is all well and good, but if they're trying to take the entire staircase at once someone is gonna get hurt. – Andrew B Nov 25 '14 at 15:25
  • 1
    I tried to point this out previously only to be shouted down because unfortunately a lot of the elitists are moderators themselves. It is good that this issue is coming to people's realisation but what does anyone suggest should be done about it? – JamesRyan Dec 5 '14 at 16:13
  • I agree, and this is like a war on who did help the more. The way it work, there is no team work at all. No one got all answers, but you can direct the user to the good diag, but other step in to tell you are dumb, or such. Only 6 days I'am participing, and I even seen a OP that edit is question, but my answer then was no longer the best, so I was getting back feedback.. thats stupid. For SF comment should give rep, on SO it's maybe another story, but there you can't write a best answer without diag, and you just hope no other answer while you diag.. – yagmoth555 Dec 7 '14 at 2:35

We seem to do this every so often. What we have is a long tail problem. Here, have a made-up graph:

Surly target market http://sysadmin1138.net/mt/blog/2012/09/08/surly-target-market.png

Years of experience is not as good of a metric as it used to be, especially as more and more sysadmin types are code-types who are trying to glue/automate systems together through frameworks like puppet, and are getting their answers on StackOverflow (they have more puppet questions than we do, by the way). The career path of such code-focused sysadmins is a different one from those of us on the rack-n-stack side of things so they feel... developery, and their questions get VTCed a lot. Which is bad, since they're the feeder system to the surly target market.

The future of the profession is much more code-focused than it used to be, and chasing away coder-types is a great way to further the decline of the site. Or push us back into the arms of StackOverflow from whence we came. I don't ask questions on SO because I get answers back in deep dev-speak, which I have trouble decoding, and they're their own kind of antagonistic towards n00bs over there. Questions asked here select for an audience who thinks like I do, so the answers are more comprehensible. New automators are coming into the market with software-engineering training so actually get value from SO, unlike me.

The code-like nature of things to come makes it even harder for others to figure out what SF is for. We're more than the Systems Annexe of StackOverflow. Heck, we were given a moat when they took away the ability to migrate SO questions to SF. The fact remains that the surly target market we have right now came of age in a time when things were quite different in the admin space and up-and-comers do not sound like we did back when.

Thoughtful, narrow focused questions that show research and give relevant detail are hard to write. The author needs to have a grasp on the problem that's complete enough for discarded theories (provide proof, please), but not so complete as to know just what the problem is. That's a pretty narrow scope, and is our platonic ideal.

Questions like this one come from the beginner space: https://serverfault.com/questions/646007/how-to-perform-all-sorts-of-administration-tasks-both-locally-and-on-remote-com

Yes, that IS an incomplete understanding of the problem. They know it's possible, and are looking for the lead in how to solve it. This question didn't show research or trial theories.

If they had asked a question on how to configure WinRM to talk to other machines via PowerShell, it might not have been VTCed. That shows they're on the right path, but just need some help with an implementation detail.

Back in 2010 I would have answered a question like that without second thoughts. Today it took less than an hour to go from ask to closed. This is the kind of person that is starting out in systems administration. is another tag with more questions on SO than SF. There is a reason for that.

  • 13
    I don't really feel that a more code-focused system administration style is the problem here. It's a flat out failure to apply critical thinking and basic troubleshooting to a problem before someone asks a question that's the problem. We're not keeping the standard high, but caviling about how we can lower the bar more and pretend we're not. That question you posted would have been great if some research had been done and a more direct question formulated. It seems like people want to use ServerFault as a stateless IRC channel nowadays. PostSecret for button mashers. =/ – Wesley Nov 21 '14 at 22:10
  • TL;DR: We don't hate the game, we hate the playahs. =) – Wesley Nov 21 '14 at 22:11
  • 3
    (Please don't take the word "hate" too seriously, because I can feel the clutchy-pearls types already buffing their pitchforks.) – Wesley Nov 21 '14 at 22:12
  • @Wesley I know that an increase in low quality questions is the perception, but I think it's just perception. If people are burned out they should take a break instead of just being angry, see the last paragraph here. – Shane Madden Nov 22 '14 at 0:15
  • 1
    @ShaneMadden Do you think the 850 questions a week we get right now are different in merit than they were in November of 2013? November 2012? Is the percentage of well researched and documened questions vs poorly researched and documented questions higher or lower than it was, say, four years ago? – Wesley Nov 22 '14 at 0:36
  • I think the question volume and overall question quality has stayed largely the same, with a minor jump in poor questions. I think the ramp-up in question closure and downvotes since July is a reaction to a perceived increase in bad questions; since people heard we were getting more google traffic, they adjusted their perception of incoming questions to assume they were from the google hordes. Compare these graphs - if the google hordes brought the bad questions, as is the theory, why didn't the downvotes ramp up until long, long after the traffic spike? – Shane Madden Nov 22 '14 at 0:45
  • 3
    @ShaneMadden Interesting. I never heard about the jump in Google traffic. Personally I just started seeing a marked rise in really poor questions which was alarming, so I took to the downvote, VTC, and flag cannons. My completely unscientific guess about it all is that the poor questions increased, which slowly got people to become insulted and indignant which snowballed into the greater use of the only tools within our grasp to work with: downvote, VTC, and flag. That's how it was with me, anyway. – Wesley Nov 22 '14 at 1:01
  • @Wesley Fair enough - it's very hard to objectively measure if question quality has gone up or down, so I don't know. Maybe this is due to the cleanup load being spread among fewer people. In any case, this change doesn't purport to fix question quality directly - hopefully it'll be fixed indirectly by having the cleanup load shared better. – Shane Madden Nov 22 '14 at 1:15
  • 7
    Yes, the devops/code-focused thing is happening... but it's not happening everywhere. Not everything is about hyper-automated web-scale systems. And that's not really the way the bulk of the industry (from my vantage point) is going. When I see questions that focus solely on that type of architecture, I wonder, "where are your mentors? Who is really responsible for the systems? And how did you get placed in charge of them?" It's like sysadmins have been erased from that landscape and the path of mentorship and learning basics is just over. – ewwhite Nov 22 '14 at 1:46
  • 1
    @ewwhite Yeah, that's a fair point. I think easier access to systems is going to mean that'll continue to be an issue, and I think we can help to push those people in the right direction. It isn't happening everywhere, for sure - I just want us to make sure we're accepting of "both ways" instead of just the way we're used to. There's good we can do for the devops/cloud types. – Shane Madden Nov 22 '14 at 4:03
  • @Shane: "it's very hard to objectively measure if question quality has gone up or down". One measure would be to look at the fraction of questions opened in each day which were still open 14 days later, and/or the average vote score on each question after a similar period, for all days in the last three years (say). Is there any way of doing that? – MadHatter Nov 22 '14 at 15:51
  • 7
    My other problem with this "new class" of users is that they usually don't return, provide feedback to (my) followup question or even bother accepting answers. A quick query on data.stackexchange.com shows that I've answered 1,168 questions where the OP didn't even select an answer. Many of these were abandoned. They're treating this site like Yahoo Answers. – ewwhite Nov 22 '14 at 21:41
  • 1
    @sysadmin1138 By the way, I showed Tom the graph, he was amused :) – Shane Madden Nov 23 '14 at 11:42
  • 3
    @ShaneMadden: that sums it up very pithily indeed; thank you for that. I'm unconvinced by your solution because we've had the problem for some time; a solution is now badly needed, or more of the people in the room will just give up and leave. By all means, once the room contains more people, unbar the door; but for now, a barricade is needed. – MadHatter Nov 23 '14 at 13:20
  • 3
    I think it is a non zero chance that sources of questions from anything not directly Google related were pushed out at about roughly the same ratio as the Google traffic increased. But I could also see the plummeting quality being as a result of momentum that was set in place long before the Google bump. Maybe the bump from Google is a red herring. I don't really know, but it seems like something to be explored. Nevertheless, the solution to either cause of the problem is the same. Some simple barriers to asking a question. Nothing snobbish. Nothing pompous. – Wesley Nov 24 '14 at 3:09

The greater the number of participants, the lower the quality of the average participant. That's just the law of averages at work. I rarely participate here anymore, because I can't get help for my problems. There comes a point where fake Internet points isn't enough to keep someone around.

I think that many of the most upset people should have stopped participating in Server Fault long ago as part of the normal attrition rate of a community like this, but instead are hanging on to the way things were 5 years ago when there were a small fraction of the page views that there are now. That doesn't help anyone. If you are truly upset at what the community has become and you haven't been able to figure out how to fix it in years then maybe there isn't anything to fix. Maybe that's just how things are now and you should try to accept it.

As an "outsider" now, it seems silly to see people debating the merits of the words "professional" or "off topic." The people on this site that truly care about that type of thing are statistically irrelevant when compared to total page views. So, I see two options:

  1. Reduce the barriers to entry. This will likely drive away a few more top users (who maybe should already be gone), but will make it a more welcoming place to random googlers.

  2. Greatly increase the barriers to entry by having some type of new user approval process. I don't think that this is realistic, as private communities are not part of Stack Exchange's mission.

I think #1 aligns with Shane's suggestion.

  • I'm very guilty of hanging on in (not so) quiet desperation but I think I'm pretty much done now. – Iain Nov 23 '14 at 18:14
  • For point #1, as a newcomer I wasnt able to comment on question as I was under 50 points. That lead to generic answer when the question was unclear to try to gain point to reach the 50 points. That does not help the quality. – yagmoth555 Dec 7 '14 at 2:48
  • 1
    @yagmoth555 : status by design – user2284570 Dec 9 '14 at 23:13
  • @user2284570 - then we cant told anything for not so good answer, by design. On SO a comment asking for more code help, but anyone can ask that, on SF, if I ask you to validate in ntdutil for all role if a alive server got them, it does not have the same impact as i lead the diag – yagmoth555 Dec 10 '14 at 0:08
  • @yagmoth555 : this kind of problem raised and as been answered on Meta StackOverflow hundreds of times with that answer. An other answer (which also had been answered several times) is just gain the necessary reputation. – user2284570 Dec 10 '14 at 0:16
  • And i add that the fact we dont reward comment, make the community need a question with a possible direct answer, thus lead to discution like it has there. – yagmoth555 Dec 10 '14 at 0:17
  • @yagmoth555 : Seriously just read meta.SE first, this isn't even about SF. Or feel free to open a duplicate feature-request. – user2284570 Dec 10 '14 at 0:19
  • @user2284570 well, dont know. But the question tell Server Fault need... , so i thougth we talk server fault – yagmoth555 Dec 10 '14 at 1:29

Obligatory preamble as to where this is coming from: I would consider myself an outsider of Server Fault. I'm very young (not legally an adult), and so I'm learning system administration in the age of Puppet and Docker and all that. I would consider myself an intermediate system administrator. This is not so much an answer as a couple of potentially useful observations from my point of view.

A lot of the posts above have been talking about negative feedback, bad quality questions, the effect of influxes of users from Google, etc. But I haven't seen a lot about how to attract new regulars, and that's an important part of turning Server Fault into a healthy community again. The way I see it, Server Fault is now in a position where it has a bootstrapping problem. I'm interested in being a regular on Server Fault, as I genuinely really like system administration, but when I've tried to start answering questions, I can't find any that are within my skill level to answer. I'm not an organization, I'm a hobbyist. I run my home network like a corporate network for the fun and utility of it, but I can't and don't use anything that requires a lot of money. Because of this, any question that is about (e.g., just off the top of my head) HP-UX, Windows, iSCSI, large-scale SANs, etc. is automatically something that I can't answer, because I don't have the equipment and/or licenses.

This is indicative of a larger problem that Shane (I believe) mentioned: Server Fault has no intermediate questions and no tolerance for intermediate questions. The question is, how do we generate an influx of intermediate-level questions that is sustained long enough that a set of intermediate regulars can form? I don't know. Maybe part of the answer is "be nice for a while and see if people come." Maybe it's something else. I genuinely have no idea.

It's very late here - one minute to midnight - so I apologize if I've just rambled on and reiterated things that have already been said, wasting your time, but I don't think I have. I hope this post offers a useful alternate(-ish) perspective.

  • 19
    This may be part of the problem. Many people wouldn't view your type of questions as intermediate. They would classify it as beginner or hobbyist not intermediate. "You don't know what you don't know" applies here. You don't have exposure to the high-end of the profession, so you don't know what's out there. Because of that, you may incorrectly classify your skill level, which may lead to hurt feelings or friction when you post on the site. Note that I don't mean you specifically, but folks like you who would likely not be considered intermediates. – MDMarra Nov 23 '14 at 12:57
  • 7
    Having quickly looked over the questions you've asked on Server Fault, I'm confident saying that they're beginner or hobbyist level questions, not intermediate questions. I don't have a problem with simple or beginner-level questions and users, as those are necessary and even valuable, and I don't think that's what the majority of us crusty old users get pissy about. It's the shitty questions that we close and/or delete dozens of a day, and the users who don't even know enough about the basics to help. – HopelessN00b Nov 23 '14 at 17:13
  • 10
    "Server Fault has no intermediate questions and no tolerance for intermediate questions" I would love it if we had more intermediate questions on the site. Everyone would love it if we had more intermediate questions on the site. They'd be tolerated all day and night. The problem is that the lion's share of questions coming in are not intermediate. They're beginner at best (which is okay in itself I guess), and completely misunderstand all that is good and holy about computer science / system administration at worst. – Wesley Nov 24 '14 at 1:13
  • 2
    @MDMarra hmm, okay. I think that's an interesting point. the way I would define beginning would be "barely knows how to use a commandline, has at most a rough understanding of how to use e.g. dpkg, ping and other basic tools", I'd define intermediate as "knows to document things, knows to think about tools like etckeeper or maybe Puppet, can handle dpkg with ease, etc.", and I'd define advanced as "eats database corruption, RAID failures, mysteriously dropped packets, OpenStack-type clustering, etc. for breakfast" - I'm assuming this doesn't match your definition? – strugee Nov 24 '14 at 1:51
  • 3
    @Wesley, MDMarra - could you give an example of an intermediate question? this is not so much for the discussion as for my own perspective. I think MDMarra nailed it when he said that perspective may be part of the problem. what I call intermediate you may call beginner, and what I call beginner you may call "doesn't even register". it may be that even if there aren't less doesn't-even-register questions on the site, it would feel better if the signal-to-noise ratio was better. in that case, s/intermediate/beginner/ and the answer still applies. maybe? – strugee Nov 24 '14 at 1:57
  • 4
    @strugee This was the best intermediate question I could think of in the last few days. Might have been the only intermediate question that popped it's head out of the swamp of despair: serverfault.com/questions/646175/… – Wesley Nov 24 '14 at 3:13
  • 4
    @strugee WFIW, I've been doing systems administration for 17 years and I'd consider myself intermediate. You are most definitely falling into the Dunning-Kreuger Pit – Chris S Nov 24 '14 at 15:55
  • 4
    @Wesley there are hardly any intermediate questions here because all intermediate questions are answered elsewhere already (even in the respective product's documentation) and the "intermediate" admin knows where to look. The remaining questions are either beginners' (who don't know where to look due to lack of experience) or ones not answered easily (and thus possibly never even posted due to the unlikelihood of getting an answer). – the-wabbit Nov 24 '14 at 21:58

Shane Madden said:

We still want to make it clear that if it’s for someone’s home network, it goes on Super User. And if it’s a weak question that can’t be understood or has no detail, I expect the close-hammer to fall just as fast as before.

I'm glad Stack Exchange is taking this position :-)

What I’m about to propose isn’t intended to lower Server Fault’s quality bar, or open the door to everything that belongs on Super User being on-topic here.

I’m asking for questions to be considered on their merits, taking away the close criteria of “infer whether this person has the right job title” because that can’t be what we’re focusing on when we decide if a question fits.

Regarding "questions to be considered on their merits"... I agree, and a few weeks ago there was a strangely similar discussion about a guy who insisted that Network Engineering should help him build an ethernet cable for his home office. Aside from the coincidental timing of the drama, I wanted to share a few thoughts... perhaps we agree, but I'm going somewhere with the discussion.

SO has had a long-standing policy about judging questions on their merits, instead of the job description of the person asking. I agree with this policy, however, let's consider some of the differences between common SO topics and SF topics:

  • SO: HTML and JavaScript
  • SO: Compiled languages: C, C++, C#, Java, go
  • SO: Scripted languages: Perl, Ruby, Python, PHP, shell


  • SF: Unix / Linux system administration
  • SF: Windows administration
  • SF: Database administration
  • SF: Network administration

I see a few notable differences between SO and SF questions:

  • On SO, it's quite possible to replicate the OP's problem without hours of labor; particularly since there is an emphasis on creating a minimal, complete, valid example. People might not be terribly good at doing it, but it's at least feasible in many cases.
  • On SF, it's quite challenging to expect people to build the same minimal example, because many questions are specific to the OP's work environment. For instance,
    • I can clone someone's git branch and experiment with their code if I need to, but how exactly does one clone their linux server, SAN traffic, hadoop cluster, or network congestion?
    • System administration also has a higher barrier to entry than programming. Even if I was inclined to spend hours building a virtual or physical mock up of an average SF problem, there are non-trivial licensing issues involved with replicating many environments (i.e. Windows, Solaris, HPUX, Juniper, Cisco...). By way of contrast, a lot of software can be written with free / open-source tools available to the masses (gcc, Eclipse, git, scripting languages, etc...).

So, what is my point?

The higher barrier to entry, and complex systems make it all the more important to get good documentation from people asking on SF; however, getting them to figure out how to document something is a chicken and egg problem.

We all say the same thing... the questions largely stink. On another site, I attempted to help people ask good networking questions, but even that (largish) document has not shown itself adequate yet. I'm not sure a self-scored checklist is the right answer; they need targeted feedback from the community.

Rhetorical question: How do you help a clueless person even know the right things to document? Speaking from very limited time moderating another site, it is pulling teeth to get people to put IP addresses and interface names on diagrams. If they simply knew how to document their problem adequately, I argue that at least 30% of them could solve the problem themselves. Yet, after trying for the last year to get good question documentation on a small beta site, I'm concluding that many people simply are not willing to submit to pleading in comments for even the simplest documentation, such as posting their configuration and an adequate topology diagram. NOTE: I have a feature request below that might improve the question feedback process.

My SF Feature Request:

I have a theory, which I hope Stack Exchange could help prove:

My theory: If we give people very specific feedback on what is missing from their question, as well as a complete example of what we expect to see then we stand a chance of getting better questions.

Feature Request: Give every SF question a scorecard, which indicates what is lacking from the question, and a good example of what it might look like. Every person who views the question can vote for whether the person needs to add information to the question.

The score card should look something like this to the OP:

  • 54 people graded this question:
    • 30% of reviewers are tired of seeing trivial questions from you (yes, I understand this bullet is controversial)
    • 80% of reviewers are not happy with the current state of the question; 10% of reviewers think the question is adequate. 2% of reviewers think you did a good job asking the question.
  • Requests
    • 90% of reviewers think this question needs a good network diagram (include a hyperlink to what we think a good network diagram looks like, including subnets / interface names).
    • 75% of reviewers think you need to add the version of operating system you're asking about (include a hyperlink to how you document the operating system for common operating systems)

The score card should be associated with each revision of the question. After a question is edited, the OP should be asked whether they included what was asked for. If they did, then the score for that element should reset. If they did not, then the score should carry forward. If people consistently lie just to reset the score card, they potentially could be suspended, or loose reputation.

In order to build a decent score card, we need to designate score card tags, such as , , , , etc... The tags would determine which score card is used for the question.

Finally, the score card must render well on mobile devices; most of the time I'm on stack exchange, I'm on my phone.

I realize that this suggestion is really a whole separate discussion itself, and I've hardly done a good job explaining the requirements. The point was to not just to gripe about how much questions suck, but hopefully offer a practical suggestion for improving status quo.

  • So for the feature request, are you thinking this will supplement or replace the current "descriptive close reasons plus comments" system for getting more information out of people? The main concern I'd have would be that getting those scorecards populated with good information would take a lot of effort on the part of the reviewers, and might not be the most rewarding way for them to spend their time on the site. On the mobile front, have you tried out the iOS/Android apps? – Shane Madden Nov 22 '14 at 19:43
  • The requested feature should happen while a question is still open. As for the effort required to populate the scorecard, it's up to the community to help people ask good questions. Scorecard templates is an effort to make it easier to give constructive feedback, as well as prioritizing what changes the community wants. I think Stack Exchange should do more to reward community review efforts, and I include this feature as the kind of effort SE should reward with points to reviewers. – Mike Pennington Nov 23 '14 at 0:43
  • The reason people don't want to document things is that they're not systems or network people. They see no value in it or the systems or networks they are just another tool they want working. They are hack coders and all they really want to do it get on with thrashing at the keyboard to get the next block buster app out the door. – Iain Nov 24 '14 at 15:22
  • 4
    Regardless of what they want to do, we need to make it clear what's required to help them in each situation. Over time, they might discover the value of fully documenting their question, because they wind up getting the help they want, and might even learn something in the process. – Mike Pennington Nov 24 '14 at 15:42

Feel free to try, not that you need my permissions or anything, but I'm pretty sure your proposal addresses the wrong problem.

I don't think that the "not professional" reason gets used as an excuse to clamp down on high quality, but home/development questions. In fact, my experience has been the opposite - the regulars will save high quality questions from that requirement, rather than using the "professional requirement" to slam shut good, quality questions. I have personally edited out "at home" from questions to save a good questions from closing, and I've witnessed other high-rep users do the same thing with similar good, but out-of-scope questions they felt added value to the site, in spite of not being strictly professional. And this is above and beyond what happens more regularly with reopen votes and the close vote review queue.

Given that, I think it would be really helpful if you find some examples of questions that add value, but got slammed shut simply because they're not professional. I really doubt you'll be able to find many. What I see far more often is a bad question, or question about "doing it badly" getting slammed shut for not being professional, the OP saying they are, in fact, a professional, and then being pointed towards the meaning of "professional" meta thread, with butthurt usually ensuing. Of course, butthurt usually ensues anyway, because it's evidently not nice to tell someone why their pile of shit masquerading as a question is a pile of shit, even when done politely and constructively, but that's a bit of a tangent.

I really do think that questions are judged on their merits as it stands now, and the professional close-reason is used as a supplementary close reason most of the time. There isn't really a clear "your question sucks" close reason, but the professional capacity close reason is fairly close, especially if you read the meta thread on what "professional capacity" means in these parts.

The data you cite would be interpreted more accurately if looked at through the lens of drive-by window-lickers becoming more active as the high-quality regulars (and even just anyone with half a clue) becomes less active. It is my opinion that the window-licking horde of crap-flingers are driving off the people who would otherwise be asking quality questions, but I will grant that it's not something I can prove, or offer more than anecdotal evidence for.

In any event, I think that's where the effort needs to be focused. Slow the influx of crap, and hopefully retain more of the regulars we have, or allow new people to grow into that role. (Ways to do that, like stopping unregistered users from asking questions have been suggested and not implemented by SE.) But I don't see us retaining existing members worth a damn, or growing new ones while the front page is full of embarrassing fail. And not just embarrassing fail from a professional IT perspective, but from any perspective. Seriously, right now, all the Server Fault main page does is validate my low opinion of humankind. And having said that, that is why I can't recommend Server Fault to friends and colleagues or co-workers any more. Not because I'm worried that someone will hurt their feelings or close their question, but because I think they'll look at the main page of the site, see the river of shit and lower their opinion of me and my skills accordingly.

  • 9
    Your premise is that bad questions from the google drive-bys drove people away, but the decline in regular posters began long before the jump in traffic, and again, the jump in traffic didn't cause any jump in question volume, so that doesn't add up. Yes, I absolutely agree that some regulars have frustrated to the point of giving up on the community by poor quality questions, but I believe that our bigger problem is that we're not bringing in fresh blood. – Shane Madden Nov 21 '14 at 18:31
  • 5
    The "not professional" reason is the most commonly cited close reason among currently pending flags. The second most common is the "minimal understanding" reason. I can't think of any custom close reason on the network that focuses to that degree on the person asking rather than the content of the question. You've got a participation problem, so that seems counter-productive. – Jon Ericson Nov 21 '14 at 18:31
  • 4
    @JonEricson the professional close-reason is used as a supplementary close reason most of the time. There isn't really a clear "your question sucks" close reason, but the professional capacity close reason is fairly close, especially if you read the meta thread on what "professional capacity" means in these parts. – HopelessN00b Nov 21 '14 at 18:36
  • 11
    But why does the question suck? To an outsider it sounds like the reason it sucks is because the people on this site are elitist dicks. (Your profession is already saddled with an unfair negative stereotype. Why do you want to perpetuate it?) Editing out "at home" hardly makes a question better. All it does it make the question fit the arbitrary rules of the site. Making a question better involves things like retitling, retagging, adding detail and removing fluff. The "not professional" reason tells people not to bother with those things because they just aren't allowed to ask at all. – Jon Ericson Nov 21 '14 at 18:56
  • 8
    @JonEricson I'm all for having a close reason that says "your question is bad and you should feel bad," but for whatever reason, that idea never gained the approval of SE. So, we use the tools we have at our disposal. (And I can tell you from experience that using the custom close reason to say "this question should be closed because it is just awful" only causes drama.) In any event, expecting community members to take more time to explain why a question sucks than the OP put into asking it is doomed to failure. Not gonna happen. – HopelessN00b Nov 21 '14 at 19:01
  • 6
    You don't have to explain to the OP why their question sucks if you pick "too broad", "unclear what you're asking" or "primarily opinion-based". Fewer clicks even. – Jon Ericson Nov 21 '14 at 19:03
  • 14
    @JonEricson Well, to be blunt about it, I don't have to explain to any user, ever why their question sucks. And as I can tell you from experience, it just causes them to get hostile and start flaming in the comments, flags get thrown, the mods message me to be nicer, and I get pissed off. So that would be why I'm not going to bother telling some random drive-by question asker why his question sucks. It's just wasted effort that results in me being pissed off. Been down that road before, and I learn from my experiences. ... – HopelessN00b Nov 21 '14 at 19:06
  • 13
    Hopeless isn't the only one either... – Chris S Nov 21 '14 at 19:07
  • 3
    @JonEricson Having said that, if you would like to be the one to explain to all the windoe-lickers why their question is just horrible, feel free, and even encouraged. Start with this recent question, rinse and repeat a few dozen times a day, and let us know how long before you burn out/want to kill all the new users/commit suicide/give up. – HopelessN00b Nov 21 '14 at 19:08
  • @JonEricson "Does TOR share IPs with other users" is "too broad"??? It's pretty clear what's being asked, and not opinion based either... – Chris S Nov 21 '14 at 19:10
  • Re: the current discussion, this is addressed in the paragraph that starts with This isn’t going to directly address the volume of low quality questions that we’re currently dealing with. Trying to further make the point that there are currently bad questions is not necessary, because there isn't disagreement about that. If you're frustrated to the breaking point by the bad questions, why is the solution to wade further into the problem? – Shane Madden Nov 21 '14 at 19:15
  • 7
    @JonEricson The "not professional" close reason is intended to apply to the subject matter, not the person. I can see how it could be misinterpreted in that way, though. It was very difficult to write; we ran into the length limit while composing it. I'd be happy with a clearer version of it. – Michael Hampton Nov 21 '14 at 20:51
  • 1
    FWIW, @Jon: Michael & I discussed that question a bit after he raised it in chat - while there's a plausible interpretation of it, it's far from clear that's what the author is actually asking. If someone wanted to answer it, I would encourage them to re-write the question for clarity; failing that, closing as Unclear is most likely to lead to the necessary improvements. – Shog9 Nov 21 '14 at 23:49
  • 3
    As someone not new to SE (SO and SU user) I've avoided asking questions on SF, even when they might have been on-topic because I am not a paid, professional sysadmin. Recently for class, I needed to setup a DSC Pull server as part of an automated Moodle deployment, I ran into some odd permissions issue, I did consider asking here, but thought better of it. As an outsider to this place, it's really hard to know whether someone in my situation, someone not strictly in the profession, but with a professional mindset and using "professional tools" is welcome to ask questions here. – Crippledsmurf Nov 23 '14 at 11:51
  • 9
    @Crippledsmurf: I don't get to be authoritative on this matter, but if you want it, my take is: yes. Anyone, regardless of mindset and tools, is welcome to ask questions here provided that they have done their basic research, and tried to answer their own question first. If you trundle by with a question at almost any level, that shows you've thought about it, and tried to skill up, I'll help if I can. If your question makes it clear that we're your very first stop for learning about what you're trying to do, I reach for my close button really quickly. – MadHatter Nov 23 '14 at 21:51

I agree that SF should be a site for professional network administration. Even though it is a valuable resource for me, I don't feel it's as good as it should be.

In fact, it's scope is too narrow: Q&A for professional system and network administrators.

I am employed by a small enterprise, I'm responsible of a network with 7 servers at the headquarter, and about 30 shops. I have real and difficult system and network issues to deal with, but I'm not considering myself a professional network administrator, because it takes only a small part in my time schedule. Programming and helpdesk take the big share.

In my opinion, SF should be open for questions about system and network administration in a professional environment.

When I face a problem for which I cannot find the solution online, I often don't feel comfortable to ask here on SF, because the question seems too convoluted. Or I wonder what might be so special in our system setup, that nobody has that problem. Sometimes I also wonder why I don't get an answer to a problem that seems simple : don't others have the same issue ?

To conclude, I think that this proposal goes in the right direction. I hope this site will remain a valuable resource, and that in the future it will open up it's scope and even get more helpful.

  • 4
    +1 - People aren't following the traditional sysadmin path. It's not a full-time role in many organizations. It should be, as there's a world of best-practices that come with experience in the field. I think in your case, you should try to form good questions. I have no problem helping someone who may not know how it "should be done". I do have a problem when the effort placed in describing the issue is too low or if it's difficult to obtain the pertinent information from the question asker. Also, going through that effort, I expect a followup: "It works", "thank you" or accept the answer! – ewwhite Nov 30 '14 at 20:41
  • 3
    For me, when I use "professional" here, it's been more about the standard of the question being asked than the resume or job description of the asker. – Rob Moir Dec 2 '14 at 10:36

We're caught between a rock and a hard place here. Server administration on the professional level requires good communication skills, and the ability to effectively break down (and convey) a problem in a way where others can understand the environment, the intent, and the challenges.

Unfortunately, these same qualities pretty much fly in the face of the nature of the internet and SE in general. This is sad, because I think there is value to what this community wants to accomplish in making its userbase conform to this bar. The value isn't just making it easier on ourselves (the people who answer): it reinforces the problem solving skills that make or break a good system administrator, which are takeaways that can help a person in a broader range of scenarios than the one they're originally asking us about. The language barrier really gets in the way of this unfortunately.

I truly find value in being able to help people get exposure to this kind of problem solving mentality on the internet. The problem is doing so in an effective way. I know I'm not really providing any solutions here, but I think it's worth saying: while it may be inevitable, I will be very sad to see our "professional" qualifier go.

  • 2
    Too often we're closing as "off topic! n00b!" and not asking those diagnosis questions, which could potentially see the question either answered or migrated somewhere sensible. Migrating says "good question, wrong place" and is Good Customer Service; closing is like slapping them in the face and genuinely makes us look bad. – Andrew Nov 28 '14 at 4:13
  • @Andrew That said, there is a line: we're not supposed to migrate bad questions as it's just passing the buck. – Andrew B Nov 28 '14 at 8:40
  • And on the flip side - a good sysadmin is also a person who's tried a lot of the possible solutions already, and is really stuck with something specific - which may not be answerable, because everyone else on SF isn't using that make/model/version. – Sobrique Dec 10 '14 at 12:51

As someone that participates on SF sporadically, the main thing that turns me off about the current environment is seeing a high volume of down voting without explanation (Basing this on what I saw in Oct / Nov, maybe things have improved). Without explanation it's difficult for an intermediate level admin to understand the flaws in their question.

Here's a recent example of high downvotes without explanation: https://serverfault.com/questions/649936/change-documentroot-of-a-subdirectory-in-directadmin (it also isn't closed)

It's not a great question but the person was down-voted and as of now was not given any tips on how to improve the question.

On the other hand, people seem to do a good job closing poor quality questions quickly.

I like the strict quality criteria, because it makes me think harder about my question before I ask. The excessive down-voting I see on other people's questions however; easily creates a negative perception of the site.

  • 2
    So... you'd feel better if the front page had lots of upvoted, or 0-score crap? But also, you like the quality standards? I'm not sure what to make of that apparent cognitive dissonance there. – HopelessN00b Dec 8 '14 at 19:05
  • Thanks for asking. I'd rather people leave a note about why they downvoted. Without helping people understand why their question is poor, the person that asked is unlikely to create a better question next time. It's more likely they'll never ask a question here again. – JMC Dec 8 '14 at 19:09
  • 4
    Well, the downvote-without-comment thing isn't a "problem" that can really be solved (not without discouraging downvoting to the point that no one does it), and one that exists on all the SE sites. The tooltip when you hover on the down arrow tells you what a downvote is intended to mean, and in most cases, that's plenty. Trust me, telling people why their question is a steaming turd just makes them defensive and offended and flame-y anyway; it doesn't do what you seem to think it would. And frankly, I don't want people who ask crap questions returning to ask another one anyway, so... bonus. – HopelessN00b Dec 8 '14 at 19:13
  • @JMC That sentiment gets repeated on meta here often, and the answers tend to be the same. There are a number of regulars who consider themselves to be in the "nicer" camp who simply can't keep up with the rate SF gets bad questions, and the reward for any actual effort is usually being slapped with a trout. Someone who demonstrates effort and intelligence to solve their own problem will usually get useful feedback in the comments, even if they are misguided in some way. Not so coincidentally, those are not the questions that draw the most downvotes. – Andrew B Dec 8 '14 at 22:14
  • 2
    @JMC I would suggest that a downvote without a comment is the equivalent of downvoting then typing in the text from the tooltip that appears when you hover above the downvote arrow as your comment. You're not wrong about it leaving a negative impression on people, but I would suggest that not dealing with bad questions would leave a worse impression. – Rob Moir Dec 9 '14 at 10:41

You must log in to answer this question.

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