4 ms·
No, it's possible to ask rather difficult questions and receive deep answers. To make an analogy, one can ask the programming equivalent of "why can alternative
by segfaultbuserr 3y ago
No, it's possible to ask rather difficult questions and receive deep answers. To make an analogy, one can ask the programming equivalent of "why can alternative current flow without a closed loop" and receive a summary of classical electromagnetism and Maxwell's equations. For some actual examples, I've read some questions and answers on how to interleave x86-64 instructions in a way to maximize instruction-level parallelism, some of the answers are comprehensive enough that they can easily be expanded into a conference talk or research report (or rather, these answers often summarize them).
So if a question can be answered with a braindump, either from facts or from (highly agreeable) experience, and doesn't look like homework, it's probably a good fit of the site's model. But if the question requires many rounds of conversations like many threads you see in a forum, it's definitely not going to end well.
- JohnFen 3y ago> if a question can be answered with a braindump, either from facts or from (highly agreeable) experience, and doesn't look like homework, it's probably a good fit of the site's model. Hmm. It's probably my fault and not yours, but this clarified little for me. At least, it doesn't make me feel confident that if I have a question, I'd be able to tell if it's one that would get a reasonable response on SO. Perhaps So would be better served if they separated out the "Q&A" from whatever entries they feel are worth being preserved, rather than combining the two. Doing that might make interacting with SO less unpleasant or more helpful, and then the particular ones that fit SO's model can be copied to a separate reference section. What I do know is that SO as it is puts an awful lot of people off, including experienced devs who might otherwise be willing to help answer questions.