As a moderator on ServerFault I
run into a lot of the iffy questions. Our users flag them for moderator
attention and we deal with them. Some are obviously wrong enough to get
dealt with through the normal close-voting process without the
mod-hammer being involved. Along the way, we mods do run into some
differing opinions on certain nuances of the FAQ. I'll be covering some of the more frequently misunderstood areas of that august document.
#4: In a Professional Capacity
The FAQ states: "Server Fault is for system administrators and desktop support professionals, people who manage or maintain computers in a professional capacity."
The most common misconception by new people, not old-timers, is this adage:
These are the people we throw at the FAQ. As the FAQ clearly states, it's not just any servers here, it's server-stuff in a professional capacity.
All kinds of people run servers, but not everyone does so professionally. One of the key differentiators between ServerFault and all of the other public sysadmin hangouts that have emerged over the years is that the scope is set to explicitly not include casual server-stuff. Wondering how to scale your Minecraft server is not topical (unless you're doing so for profit, at which point say so). Wondering how to scale your memcached infrastructure for better overall application performance is.
But what is "professional?"
The userbase has had a lot of talk about this in the past, and there are still some divisions. We've lost some top users because they thought the ServerFault defacto definition was not restrictive enough. One thing I've noticed over my career is that system administrators as a class have a low tolerance forstupidity cluelessness, they get enough of that at work. Cluelessness is not professional, so when the clue-free start asking questions and not getting shot down in short order it grates. So they leave.
"Someone who manages servers and networks for pay" is one broad definition, but there are a lot of clueless people doing the job that are completely out of their depth. Some of our users roll in certain minimum skills, such as sufficient research capabilities to have found and tried some things before coming to ServerFault for the fix. Still others seem to have a nebulous smell-test, which is subjective and hard to pin down in text.
The top system administrators I've know have all been good at one thing. And that is the ability to research problems, find possible solutions, test them, analyze how that changed the problem presentation, and refine the research parameters to begin the cycle anew. You learn a lot by doing this, and sysadmins need to learn a lot and do so continually.
When we see questions on SF that show this pattern, and they're here on SF because they've run into something intractable to their research skills, we smile and answer when we can. Sometimes all we need to do is show a research path they'd missed and off they go. Others, we point out faults in their analysis, which causes them to refine their search terms, and solutions ahoy!
These are our people. They belong here, even if it turns out they're hobbyists (though beware of the dreaded 'home' word). It's the outlook that reflects so well upon them.
People who deal with problems by gathering a bunch of data and throwing it at a bunch of nominal experts in the strenuous hope that an answer will fall out the other end are not our people. All too often, these questions lack the right kinds of data for us to come up with answers and we fall into iterative, "have you tried?" loops. Those sorts of questions are a better fit for full up forums, where extended discussions, back-and-forth, trial and error, and overall one on one extensive interaction are possible. Questions requiring this level of hand-holding are a very bad fit for ServerFault, and they tend to get ignored or flagged 'low-quality' as a proxy for a 'delete this crap' flag.
And yet...
We do get occasional questions from sysadmins who are working out of their field of expertise for some reason and need help. The Windows admin suddenly having to crack a BGP routing problem. The MySQL expert suddenly thrown into MS-SQL. The Cisco expert who just had a bunch of Juniper kit dropped in their lap. The Apache guru suddenly dealing with IIS (these tend to include swearing). When you're out of your depth, it's hard to ask good questions. You don't know enough yet to frame them right for the experts.
Is this a "professional" problem? Not so much, it's more of a "bad question" problem. Especially if they're a frequent user on SF, we will do what we can to help, going so far as to inviting them into a chat-room for intensive one-on-one. They've proven they are professional by their other work on SF. This is not a courtesy we provide to brand new users. Double-standard? Yep, but that's what reputation systems are for.
The second most common misconception is this one:
#4: In a Professional Capacity
The FAQ states: "Server Fault is for system administrators and desktop support professionals, people who manage or maintain computers in a professional capacity."
The most common misconception by new people, not old-timers, is this adage:
This is ServerFault, and I have a Server. Therefore this question belongs. QED.No.
These are the people we throw at the FAQ. As the FAQ clearly states, it's not just any servers here, it's server-stuff in a professional capacity.
All kinds of people run servers, but not everyone does so professionally. One of the key differentiators between ServerFault and all of the other public sysadmin hangouts that have emerged over the years is that the scope is set to explicitly not include casual server-stuff. Wondering how to scale your Minecraft server is not topical (unless you're doing so for profit, at which point say so). Wondering how to scale your memcached infrastructure for better overall application performance is.
But what is "professional?"
The userbase has had a lot of talk about this in the past, and there are still some divisions. We've lost some top users because they thought the ServerFault defacto definition was not restrictive enough. One thing I've noticed over my career is that system administrators as a class have a low tolerance for
"Someone who manages servers and networks for pay" is one broad definition, but there are a lot of clueless people doing the job that are completely out of their depth. Some of our users roll in certain minimum skills, such as sufficient research capabilities to have found and tried some things before coming to ServerFault for the fix. Still others seem to have a nebulous smell-test, which is subjective and hard to pin down in text.
The top system administrators I've know have all been good at one thing. And that is the ability to research problems, find possible solutions, test them, analyze how that changed the problem presentation, and refine the research parameters to begin the cycle anew. You learn a lot by doing this, and sysadmins need to learn a lot and do so continually.
When we see questions on SF that show this pattern, and they're here on SF because they've run into something intractable to their research skills, we smile and answer when we can. Sometimes all we need to do is show a research path they'd missed and off they go. Others, we point out faults in their analysis, which causes them to refine their search terms, and solutions ahoy!
These are our people. They belong here, even if it turns out they're hobbyists (though beware of the dreaded 'home' word). It's the outlook that reflects so well upon them.
People who deal with problems by gathering a bunch of data and throwing it at a bunch of nominal experts in the strenuous hope that an answer will fall out the other end are not our people. All too often, these questions lack the right kinds of data for us to come up with answers and we fall into iterative, "have you tried?" loops. Those sorts of questions are a better fit for full up forums, where extended discussions, back-and-forth, trial and error, and overall one on one extensive interaction are possible. Questions requiring this level of hand-holding are a very bad fit for ServerFault, and they tend to get ignored or flagged 'low-quality' as a proxy for a 'delete this crap' flag.
And yet...
We do get occasional questions from sysadmins who are working out of their field of expertise for some reason and need help. The Windows admin suddenly having to crack a BGP routing problem. The MySQL expert suddenly thrown into MS-SQL. The Cisco expert who just had a bunch of Juniper kit dropped in their lap. The Apache guru suddenly dealing with IIS (these tend to include swearing). When you're out of your depth, it's hard to ask good questions. You don't know enough yet to frame them right for the experts.
Is this a "professional" problem? Not so much, it's more of a "bad question" problem. Especially if they're a frequent user on SF, we will do what we can to help, going so far as to inviting them into a chat-room for intensive one-on-one. They've proven they are professional by their other work on SF. This is not a courtesy we provide to brand new users. Double-standard? Yep, but that's what reputation systems are for.
The second most common misconception is this one:
This is a pre-production environment I'm building for eventual deployment to production. That's professional!Not according to us, at least. We scope "professional capacity" rather closely around stuff that's running in production. This can be a hard distinction to make since we're supposed to solve problems in pre-production before pushing to production. We've found that scoping things this way makes it very clear that the following two cases are equivalent:
- I bought a VPS running CentOS and want to build web-sites on it. How do I install nginx to it?
- How do I get VHOSTs to work with MAPP on my mac?
Unfortunately, this also excludes questions like these:
- I'm swapping out all of my old Catalyst-based distribution layer for an Extreme-based programmable layer. How do I get VRRP working during the transition?
- During our test migrations to Exchange 2012 from Exchange 2003 we're seeing the dates on emails change. Where is that coming from?
We've deemed this acceptable in large part because the former questions (webdevs getting their environment set up) vastly, vastly outnumber the later questions.