3 ms·

For what it's worth, I read the question without reading the flagging reason and could predict the flagging reason with high confidence. The question looks lik

by kqr 3y ago

For what it's worth, I read the question without reading the flagging reason and could predict the flagging reason with high confidence. The question looks like an "I'm out of my wits please debug my application" question which isn't atomic or precise enough to fit the SO format. I would have liked the person asking the question to provide a complete minimal failing example to even begin working on an answer.

nairboon 3y ago
Aren't all questions on SO of this type? With that reasoning you could close the whole SO. If you really debug hard enough and for a long time, you wouldn't ever need to ask on SO.
- oefrha 3y ago
  Not a moderator but have ~8k rep on SO and have spent some time on mod queues in the distant past. Most questions on SO these days (that are not "do my homework for me please") are garbage like this. Signal to noise ratio is horrendous. Gin, Next.js, whatever are pure noise, the actual problem is glossed over in a vague sentence: "my browser is getting the setcookie header correctly, it doesn't say it's blocked at all but my browser isn't setting the cookie". You could try to talk the submitter into providing the complete HTTP request, which could take a very long back and forth, but after a thousand questions like this, you probably decide it's not worth the effort to salvage the question; or like me, you decide it's not worth participating in the site at all.
- emodendroket 3y ago
  No. The purpose of the site is serving as a general knowledge base, not debugging random code that's only going to be useful to one person.
2-718-281-828 3y ago
i second that sentiment. i also find the way the "question" is put somewhat confusing and technically it's not even a question as it doesn't contain a question.
haspok 3y ago
This kind of thinking is exactly why I stopped asking questions on SO a long time ago. Because once you can construct the perfect question, you can most certainly answer it yourself. It should be OK for questions to be somewhat vague, to contain "noise" and non-related stuff (because how on earth should the person asking the question know if something is relevent if they don't know the solution already?). And I think in most cases a "complete minimal failing example" is way overkill. If someone in the future runs into the same problem, they have code already, and if there is already a solution accepted, then all that is needed is the description of the problem case and the solution (and in the best case, a root cause analysis, but that is an optional extra). Some mods seem to think that SO is the next Encyclopedia Britannica, and each question is an article in it. This was not true in the beginning, and I don't feel like this is the right direction. But who cares about what I think.
- kqr 3y ago
  > Because once you can construct the perfect question, you can most certainly answer it yourself. Not if what you're lacking is declarative knowledge, which is what the site tries to elicit, to a first approximation anyway. You should reduce the question to the point where you can say "between substeps 1.3.5 and 1.3.6 there seems to be a proposition I don't know. What is it?" If you can't reduce the problem to a missing proposition, there are better resources to turn to for help with debugging it step-by-step until you have isolated a problem. I've had success with mentoring and IRC.

by kqr 3y ago

nairboon 3y ago

Aren't all questions on SO of this type? With that reasoning you could close the whole SO. If you really debug hard enough and for a long time, you wouldn't ever need to ask on SO.

oefrha 3y ago
Not a moderator but have ~8k rep on SO and have spent some time on mod queues in the distant past. Most questions on SO these days (that are not "do my homework for me please") are garbage like this. Signal to noise ratio is horrendous. Gin, Next.js, whatever are pure noise, the actual problem is glossed over in a vague sentence: "my browser is getting the setcookie header correctly, it doesn't say it's blocked at all but my browser isn't setting the cookie". You could try to talk the submitter into providing the complete HTTP request, which could take a very long back and forth, but after a thousand questions like this, you probably decide it's not worth the effort to salvage the question; or like me, you decide it's not worth participating in the site at all.
emodendroket 3y ago
No. The purpose of the site is serving as a general knowledge base, not debugging random code that's only going to be useful to one person.

2-718-281-828 3y ago

i second that sentiment. i also find the way the "question" is put somewhat confusing and technically it's not even a question as it doesn't contain a question.

haspok 3y ago

This kind of thinking is exactly why I stopped asking questions on SO a long time ago. Because once you can construct the perfect question, you can most certainly answer it yourself. It should be OK for questions to be somewhat vague, to contain "noise" and non-related stuff (because how on earth should the person asking the question know if something is relevent if they don't know the solution already?). And I think in most cases a "complete minimal failing example" is way overkill. If someone in the future runs into the same problem, they have code already, and if there is already a solution accepted, then all that is needed is the description of the problem case and the solution (and in the best case, a root cause analysis, but that is an optional extra). Some mods seem to think that SO is the next Encyclopedia Britannica, and each question is an article in it. This was not true in the beginning, and I don't feel like this is the right direction. But who cares about what I think.

kqr 3y ago
> Because once you can construct the perfect question, you can most certainly answer it yourself. Not if what you're lacking is declarative knowledge, which is what the site tries to elicit, to a first approximation anyway. You should reduce the question to the point where you can say "between substeps 1.3.5 and 1.3.6 there seems to be a proposition I don't know. What is it?" If you can't reduce the problem to a missing proposition, there are better resources to turn to for help with debugging it step-by-step until you have isolated a problem. I've had success with mentoring and IRC.