3 ms·
My issue with the site is the opposite. Asking a specific nuanced question results in a response of a) tell me what you are really trying to accomplish and i'
by mmd45 8y ago
My issue with the site is the opposite.
Asking a specific nuanced question results in a response of a) tell me what you are really trying to accomplish and i'll answer that question or b) don't do what you are doing.
Don't presume I don't know what I'm doing or know the right question to ask. This happens so often and is such a time waster I hesitate to use the site. Their culture even invites it by giving it a fancy name of the "xy problem".
Please provide a way to flag a question as "Don't respond with an XY question" or alternatively "I know what I am doing and still have a question".
- the8472 8y ago> Please provide a way to flag a question You can do those with prose. Explain that you know what you're doing. Give an underlying reason that shows that your X is an X. Answerers can only act on a) the information you provide b) heuristics they have based on seeing many other questions. If you do not provide a) then they'll have to fall back on b)
- mmd45 8y agothat invites a debate about the premise. i shouldn't need to justify my constraints. i should be able to pose a problem and STATE the constraints. if you don't believe that my constraints are real then you can feel free to not answer my question but the SO regulars always feel the need to interact even when they have nothing to add.
- the8472 8y ago> if you don't believe that my constraints are real then you can feel free to not answer my question But XY-problems exist and are not uncommon. Overconstrained questions also exist. How are people supposed to know that yours is not one of those without additional information?
- zbentley 8y ago> i shouldn't need to justify my constraints. If you're asking people for free help and to spend their time understanding how their relevant but not identical experiences may adapt to your different-but-related problem, then yeah, you might. If someone, even someone I respect, know personally, and work alongside, comes up to me and says "I need to install a nearly-20-year-old version of MySQL next to a modern software stack; help!" [0], I'm going to ask "why?" and maybe "are you sure there's no other way to approach your goal?" first. Not presumptuously or because I'm sure I know better, but because a) the answer will help me better understand the problem and goal, and b) because sometimes extremely competent, rational people really do overlook the obvious solutions, sometimes for days/weeks spent beating their head against the wrong problem. Sometimes the justification is as simple as "it's a business constraint imposed from above". Sometimes it's a more complex story. Either way, wanting to know that isn't asking too much. [0]: https://stackoverflow.com/questions/47350382 https://stackoverflow.com/questions/47350382
- alkonaut 8y ago> tell me what you are really trying to accomplish and i'll answer that question That IS a valid question. The site would be nearly useless without that. Of course there are reasons why you may not want to do the obvious thing Y but have to do X to solve a particular problem, but for a nuanced and specific question that's really the most important thing in the question. If it's not there, the answerer will (rightly) assume you are trying to bang in a nail with a saw - no matter how nuanced the question is otherwise. That is: to be a good question, you have to be very specific about what problem you are trying to solve, what solutions you can't use, and why it is that you have to go with a specific solution. The site isn't a question and answer site, it's a problem and solution site. People don't go to stack overflow to find an answer to a question, but rather to solve the problem they have at hand. So to make your question as good as possible for those people it really needs to have a problem context, including rejected solutions. If those rejected solutions aren't in the question - they will come in the answers. That's good.