@jpmc26 In my experience, comments are a pretty rotten way to communicate anything of significance. Chat is marginally better. This is why I suggested posting your own question. With a full post, you can communicate a lot more clearly.

@jpmc26 I think careful thought and effort are rewarded, but only sometimes and inconsistently. It will be endlessly frustrating if you expect it to be a direct correlation. Most of the time, especially in today's low-vote environment, getting reputation requires a bit of luck.

I used a sockpuppet to ask a question a couple of years ago that took many hours to research. Then I spent many more hours figuring out my own answer. The result: 10 reputation.

I got +35 for knowing the command to install a gem. Fair? Not really. But that's how things go sometimes.

You might be interested in my talk on the subject.

@JonEricson I was referring to more than reputation. I was referring to the fact that someone can totally misrepresent your point and garner agreement based on it. Careful thought is only rewarded when it happens to coincide with people's preconceived notions.

(FYI, one point in your comment above I didn't respond to before: the user took almost a day and a half to respond to my question about testing multiple runtimes. I suspect they only did so after being questioned, and even after that, they didn't clarify the question to reflect those findings.)

I picked that example specifically because it was so highly upvoted and received such wide approval and was written by such an avid user, but I had explicit evidence that the author hadn't done their research to contextualize the question.

These facts lend credence to the notion that going through even basic steps to look into your question and make sure you understand the context beforehand is simply not valued by the majority of users. But not doing so leads to a very large number of poor questions. Askers who want a regular expression aren't learning about them to be able to construct their own, leading to effectively infinite variations on "build my regex for me."

The same can be said for, "filter, transform, or aggregate my list for me" type questions.

But these are only the most blatant examples. It's equally true of questions like the one I linked; it's just more subtle.

@jpmc26 Why pick the subtle example? I'll be honest: if the author of this question literally did no research at all before posting, I still think it's a good question. In a way, the very fact that someone noticed the oddity shows they've spent years of their life researching and using Python. The question is as far removed from "build my regex" as I can possibly imagine.

Do you want expert questions? Don't hassle experts when they ask questions.

Now experts can be wrong, of course. They can ask dumb questions just like the rest of us. I just don't think this example demonstrates anything like an high reputation user asking a question they should not. Maybe use an example like that.

Related to this discussion:

I'd be curious if you agree with Tiny Giant.

@JonEricson I don't think who's writing it should factor into our evaluation of the question's quality. Remove wim's user name and reputation from the equation. Imagine a 1 rep user stumbled across the behavior on their second week of programming, which is a realistic possibility. Is it still a good question? What qualities in the question itself make it so?

@JonEricson I think Tiny has a valid criticism, but I think what makes debugging questions so typically bad is more what I'm discussing here: askers rarely take the time to delve into their problem and get a good grasp of the subject matter. They don't feel the need to do so, and there's almost no forces at SO pressuring them to, social or otherwise.

After the "minimal understanding" close reason was removed, downvoting became the only tool available for the purpose. Unfortunately, it doesn't do a good job. It doesn't clearly communicate the problem, and I believe downvotes are much less common than upvotes. I don't know is that's due to the rep barrier, the cost, the social pressure to avoid making feel bad, or maybe other factors, though.

@jpmc26 I don't think it's a realistic possibility that a new programmer would find this. But if they did, it's hard to imagine a more clear and direct question. Someone new to Python would be hard-pressed to explain a hash of infinity without being confused. That the question is not even a little confused by this esoteric feature suggests it is not an uninformed question.

(As a point of clarity, when I mentioned wim's experience before, it wasn't because of that experience that I judged the question to be poor. I was merely eliminating the possibility that he might not be aware of the documentation or have any difficulty reading it.)

@jpmc26 We should probably stop hiding the extensive blocks and hurdles:

meta.stackoverflow.com/questions/318174/…

@jpmc26 A good deal of it is that downvoted questions disappear and get deleted. Upvoted questions gather more upvotes.

@JonEricson I can only really go by my experience with the questions I actually visit. Questions that disappear because they're downvoted wouldn't contribute to that experience, right? Same for questions that are never asked because of hurdles.

The hurdles and deleted questions exist if you don't see them. ;-)

Yes, but how would they create the common perception of the site being flooded with bad questions and there being very few worth actually answering? I'm clearly not alone in this.

There's really no point in trying to convince people who ask clear and concise questions to show their research. It certainly won't convince the bulk of people who try to ask questions to do the same.

@jpmc26 When there is a question worth answering (and several people did answer the one we are talking about) it's part of the solution to this perception. Why treat it as part of the problem?

By definition an answered question is one that somebody found worth answering. Why take that oppotunity away?

@JonEricson Because it affects the culture of the site.

There are so many bad questions, but we seem to always be talking about questions that are miles above average. So counter-productive.

@jpmc26 How? So far I've just seen a slow ratcheting of standards to the point people seem suspicious of a self-answered question.

To be clear, I think the question I linked could rather trivially be improved. I even left specific comments indicating what needed to be clarified. But rather than just deal with the problems, the author just asserted he didn't care, and you seem to be asserting the same.

Just raising the bar is not helping stem the flood. It's mostly getting rid of the people who care enough about quality they might improve. Why bother trying to get better if the standard seems pointlessly impossible.

@jpmc26 Why not edit?

Comments suggesting edits are just obnoxious when the edit link is right there.

If you have a working code fix to an open source project, wouldn't you submit a pull request?

@JonEricson I did consider it. I've held off for two reasons. 1. I was concerned that my edits might conflict with the author's intent (because I don't know how they want to handle the multiple runtime issue), so I wanted them to do it rather than risk creating a conflict. 2. I didn't know how stuffing the old version into previous edits might hinder this discussion. The second reason alone probably wouldn't be enough to prevent me from doing it, but the first reason is also there.

Because you wanted to leave evidence the author didn't do research?

(People are good at reading the change log, by the way.)

Rollbacks are costless too. I don't know. Seems wasteful to argue about this. (Both for your time and the author's.)

I suppose. Not the most altruistic of reasons, I realize. But the first reason means I would have to engage in a long conversation to even figure out what the final version needs to look like. (Should it ask about CPython only? If not, is that Too Broad?) A lot of effort to go through with someone who doesn't even acknowledge there's a problem.

Have you considered that the way the comments were phrased might be overly confrontational?

"Please edit the question because the answers don't explain why other runtimes do the same," is overly confrontational?

That comment could have been more clear, in retrospect. "Please clarify which runtimes you wish to ask about," would have been more direct. But I don't think my comment was confrontational.

I have considered that perhaps my later ones are, but they clearly aren't the source of the author deciding there isn't a problem. Their mind was already made up before that.

In fact, it wasn't until after wim's reply to that comment about asking about the language in general that I even realized they hadn't researched the issue. I had assumed they had looked into it before that. Then I spent about 10 minutes checking 3 places in the docs that I could think of for information about the value of the hash and discovered it's runtime specific, except for the guarantees about it being stored in sys.hash_info.inf.

Instead of asking if it might be implementation specific, you demanded to know if they had tested it on several runtimes. Since it is the same result everywhere, that's pretty much a red herring.

Um... It's not. wim's deleted comment specified a runtime that has a different value. Something similar to micropython, I think it was called? I wasn't familiar with it.

...Or am I remembering wrong? Hm.

Maybe it was that the answer was specific to the CPython runtime, so I wasn't sure if wim wanted information on other runtimes or not. (Hence my earlier point about the question being unclear.)

But the answer makes all this clear. It was an arbitrary value selected in the process of fixing a bug. It doesn't matter if a different implementation picked a different value or just kept the arbitrary value already in the code. And it's hard to imagine why someone would bother to change.

@JonEricson Anyway, I think "demanded" is assuming bad faith that is not in evidence. I posed the question and then went on to indicate that if he was asking about CPython in particular, the question should be edited. I was just trying to clarify. My first comment was intended to be neutral, and reading more into it is imposing assumptions.

Undocumented, but standard behavior.

@jpmc26 Having asked many non-debugging questions, I can assure you that comments like that sound like demands to justify the question. It's annoying to users who know what they are doing.