4 ms·

Sure, there are cases where seemingly-ridiculous things are actually needed. Where a lot of people commenting on this article are failing, though is thinking o

Sure, there are cases where seemingly-ridiculous things are actually needed. Where a lot of people commenting on this article are failing, though is thinking only about their needs and what they want. The people responding to your IRC message/ML post/etc are (usually) giving up their valuable time to help others for free, and making some accommodation for this is essential if you want to get anywhere useful. On top of this, when you join a room/list where you aren't well known, the responders have no context on who you are, your level of experience, etc[0] - but what they DO have is long, long experience with people asking XY questions where Y is both ridiculous and not going to solve X anyway. I know it's frustrating to have to go through your reasons when you are certain that damn_silly_thing_Y is actually the only way of solving your problem, but it is needed. After all, you're asking these people because they're likely more experienced than you are (or else why ask them?) and thus there's a chance you could be wrong. 0: And you can't often shortcut this - due to astronomically high rates of Dunning-Kruger cognitive bias in many of these venues, starting out with "I'm an expert, I know what I'm doing" without very convincing evidence is usually extremely counter-productive.

derefr 7y ago
I think you confused my point for the point of the OP, and then just typed a reply that is basically a reply to the OP. My point was the fact that, at the end of the justification chain, you've generally left the domain of expertise of the expert/forum of experts that you came to. Because of this, the people you're conversing with will fail to actually help you with your problem, because they've been dazed/confused/distracted by the higher-level problem statements that exist outside their domain of expertise. Even though the immediate problem is one they're eminently qualified to solve. It's unintentional nerd-sniping: they force you to tell them about a larger problem that they don't know how to solve—sometimes, in my case, because it's a research problem and nobody knows how to solve it—and they get distracted trying to understand the entire breadth of a topic they may not have ever considered before, in order to figure out how they would solve the highest-level problem themselves. They never "come down to earth" to solve your immediate problem. It's as if you went to a low-ranking soldier and asked something about how well some body-armor they use works in practice, and they asked you what the context is, eventually demanding you to explain the whole strategy of the war to them, at which point they get confused trying to understand the strategy and politics of the war they're fighting, which is usually inevitable because they're a soldier, not a politician.
- jfindley 7y ago
  I don't think I am. I can certainly understand that it's doubly frustrating when after all the explanations you still don't get an answer - but there's a couple of further issues with your point specifically. As has been noted elsewhere, questions need to be appropriate for the venue they're asked in. A general-purpose channel for <language>-users may not be the best place to ask extremely niche questions about algorithm design in edge-case scenarios where none of the usual answers apply. If what you're doing is hyper-specialised to your individual environment, there's a very real chance no-one will know the answer. If you put the effort in to explain, in detail, WHY you need this really strange problem, you may get lucky and someone smart and experienced in a closely related field may be around and willing to put in the time to help you. Or not. But you're still going to need to be willing to to in depth into the reasons behind what you're doing or they will likely ignore you. The lack of response probably isn't because people are dazed and confused, it's probably because your constraints are so restrictive that what you need isn't free IRC help, but a specialist contractor who you're actually paying for their time.
  - derefr 7y ago
    Do note that I'm not talking about asking questions on IRC; but rather, I'm talking about asking questions of people who I already know to be experts in the specific field I'm asking about. (The "specialized contractors" you're talking about, but more like e.g. university professors.) Like, if I have a concrete graph-theory problem, I'm asking a graph-theory expert about it. The problem doesn't come because the population I'm asking doesn't know graph theory; it's because someone who's an expert on graph theory can't be expected to also know e.g. DSP design, and the higher-level problem is a DSP design problem, and once they find that out, they say they don't know DSP design and refuse to help. Even though the concrete problem is a graph theory problem! This is, IMHO, one of the key psychological problems hindering interdisciplinary fields of study: everyone thinks that, because a problem is A&B, and they only know things about A, that they can't make a contribution. Even though everyone only knows either A or B, and the whole point is to get the people who know A talking to the people who know B, rather than to create some sort of A&B paragons.

derefr 7y ago

I think you confused my point for the point of the OP, and then just typed a reply that is basically a reply to the OP. My point was the fact that, at the end of the justification chain, you've generally left the domain of expertise of the expert/forum of experts that you came to. Because of this, the people you're conversing with will fail to actually help you with your problem, because they've been dazed/confused/distracted by the higher-level problem statements that exist outside their domain of expertise. Even though the immediate problem is one they're eminently qualified to solve. It's unintentional nerd-sniping: they force you to tell them about a larger problem that they don't know how to solve—sometimes, in my case, because it's a research problem and nobody knows how to solve it—and they get distracted trying to understand the entire breadth of a topic they may not have ever considered before, in order to figure out how they would solve the highest-level problem themselves. They never "come down to earth" to solve your immediate problem. It's as if you went to a low-ranking soldier and asked something about how well some body-armor they use works in practice, and they asked you what the context is, eventually demanding you to explain the whole strategy of the war to them, at which point they get confused trying to understand the strategy and politics of the war they're fighting, which is usually inevitable because they're a soldier, not a politician.

jfindley 7y ago
I don't think I am. I can certainly understand that it's doubly frustrating when after all the explanations you still don't get an answer - but there's a couple of further issues with your point specifically. As has been noted elsewhere, questions need to be appropriate for the venue they're asked in. A general-purpose channel for <language>-users may not be the best place to ask extremely niche questions about algorithm design in edge-case scenarios where none of the usual answers apply. If what you're doing is hyper-specialised to your individual environment, there's a very real chance no-one will know the answer. If you put the effort in to explain, in detail, WHY you need this really strange problem, you may get lucky and someone smart and experienced in a closely related field may be around and willing to put in the time to help you. Or not. But you're still going to need to be willing to to in depth into the reasons behind what you're doing or they will likely ignore you. The lack of response probably isn't because people are dazed and confused, it's probably because your constraints are so restrictive that what you need isn't free IRC help, but a specialist contractor who you're actually paying for their time.
- derefr 7y ago
  Do note that I'm not talking about asking questions on IRC; but rather, I'm talking about asking questions of people who I already know to be experts in the specific field I'm asking about. (The "specialized contractors" you're talking about, but more like e.g. university professors.) Like, if I have a concrete graph-theory problem, I'm asking a graph-theory expert about it. The problem doesn't come because the population I'm asking doesn't know graph theory; it's because someone who's an expert on graph theory can't be expected to also know e.g. DSP design, and the higher-level problem is a DSP design problem, and once they find that out, they say they don't know DSP design and refuse to help. Even though the concrete problem is a graph theory problem! This is, IMHO, one of the key psychological problems hindering interdisciplinary fields of study: everyone thinks that, because a problem is A&B, and they only know things about A, that they can't make a contribution. Even though everyone only knows either A or B, and the whole point is to get the people who know A talking to the people who know B, rather than to create some sort of A&B paragons.