3

I'm looking for a good way to help OP's with conceptual issues. Unlike StackOverflow, the Raspberry Pi forum is bombarded with people in genuine need of help in thinking about their problem. These people are not lazy. They are simply needing a bit of conceptual help. Here are two examples:

1) I see people asking about X when Y is the problem (e.g., "why is my code broken?" => "use a multimeter").

2) We dread open-ended questions. But some questions taken to be open-ended are in fact questions about concepts. Surely there must be some way for an OP to ask for help about a non-controversial concept for which we can provide verifiable and concrete examples?

As you've all no doubt noticed, I've flailed a bit trying to address this myself, so I consider that flailing to be the research due a proper question. I don't like the current downvote solution as it detracts from our mission of "we're all here to learn, so be friendly and helpful!"

But nobody wants the Reddit wrangle of woe.

What other options do we have? Would a tag help? Or would we allow "What considerations are required to X?"

Notice that other forums seem more flexible in their responses to conceptual questions:

SE electronics #1 SE electronics #2

The above questions handle conceptual queries with grace and humor. I would love a way to bring that joy here.

8

The system we have on Stack Exchange works great for specific technical questions, e.g. "I want to achieve X, but Y doesn't work because Z. How can I fix Y?". Problems that can be definitively answered in three or four paragraphs and/or a little bit of code fit in well. Generally, there are only a few solutions at most, and sometimes there's just one answer.

The trouble with more open-ended or broad questions is that they can't be definitively answered a lot of the time. It's not necessarily that people haven't tried to get an answer1, but it's quite hard to elaborate on a problem if you don't know what you're looking for at all.

Teaching someone a concept is absolutely fine here, so long as you can fit it within a few paragraphs and provide a reasonable explanation. "How does an engine work?" asks too broad of a concept, but asking about one specific subsection would be fine... even though it's still conceptual.2

A tag wouldn't be something that would help here — Stack Overflow did that many years ago, but it simply encouraged people to ignore the site's standards. Even though you could filter these questions, they're still on the site, lingering, perhaps collecting inappropriate answers or other moderation problems. You can't solve a problem with a group of questions by simply hiding them from people; it just makes it harder to notice there is a problem.

The intended workflow is:

  • Open-ended questions that can't be answered are closed as too broad or primarily opinion-based as appropriate.

  • The author takes this feedback into account and edits accordingly (if possible)

  • The question gets sent automatically for users to review and decide whether it's worth reopening.

  • The question is reopened (or more feedback is given) and answered.

In theory, at least. In practice, something goes wrong between the question getting closed and getting reopened — as so few questions are reopened. Since 2017, 2341 questions have been closed (most of which have been automatically deleted now), but only 112 questions have been reopened. Unfortunately, very few people are able to turn their closed question into a suitable one for the site, and out of those who do, it's difficult to get your question voted open again3. As goldilocks points out below, some users instead just start afresh and post a new question, bypassing the opportunity for users to review the changes and reopen. In reality, that's necessary if you ever want an answer as reopening is just so slow.

There are some questions that simply can't fit and are inherently open-ended. Users are welcome in chat to discuss these sorts of questions, though, but the issue with that is the chat room isn't too busy and your chances of getting a helpful answer aren't too great.

I think if you're interested in helping people to think about a problem, a back-and-forth chat room discussion might be helpful. Using that, you can work with the OP to understand what they actually need, so that they can then come back and post a more specific question.


1 A statistic from the Stack Exchange team is that ~90% of Stack Overflow users did search before asking. It seems unfair to call someone lazy when the evidence is that they have tried. That said, the search facility here is not the greatest, and if you don't know the term to search for, you could waste days and find nothing. Sometimes I feel like the insistence on 'effort' misses the point of Stack Exchange. There are often problems with open-ended questions, but relying on 'effort' to make a good question is just going to leave you disappointed. A poor question with an attempt that is entirely hopeless is no better than a poor question without... but the one without might help other people in future more.

2 You often hear complaints that conceptual questions haven't put in enough 'research effort', and that users are being unfair by expecting us to answer their questions. That said, my most upvoted answer on Stack Overflow is in response to what could be considered a rather poor question. It could, if you knew where to look, be answered by checking the documentation. When I answered it, the question was actually downvoted to -1. Today, the question stands at 97 upvotes. It's reached over 60,000 people—a perfect example of a question serving many people in future rather than just one today. That would have been 60,000 people not helped, had the user been turned away due to a lack of 'effort'. Future questions can now simply link to the existing one, so there's no need for it to be asked again and again in duplicates. That's how the site is supposed to work. It just happens that a lot of helpful questions aren't considered 'good' questions by our standards.

3 When was the last question reopened by 5 users? A couple of months ago? Turns out it's actually been over 4 years since the last reopening by users, in July 2014. It's only ever happened 8 times in the history of the site! It would seem that there aren't many people voting to reopen (probably because few people even have the privilege).

  • The demise of the [suggestion] tag was most informative. – OyaMist Jun 7 '18 at 17:11
  • I've edited to propose question wording that could be enumerable and verifiable provided OP gave initial list as fiat for research given unsuitability of metatags. – OyaMist Jun 7 '18 at 17:15
  • 1
    wrt helpfulnes of back-and-forth discussions, certainly true. We see the same effect in comments - with the drawback of cluttering the comments. So, it is in fact advisable to use chat instead of comments. – Ghanima Jun 7 '18 at 19:02
  • yes. chat is an awesome tool. I've engaged OP's to do that on concrete questions. However, there is still a bit of a problem of phrasing a question such that it doesn't get booted to the hold of despair. Note that the SE electronics queries would have been zapped into hold quite quickly here but survived and flourished with great outcomes in that SE site. – OyaMist Jun 7 '18 at 21:56
  • I've edited the title to match the body. It was too open-ended. Apologies. – OyaMist Jun 7 '18 at 22:36
  • 1
    Great answer. WRT the number of re-opened questions, I think the (difficult to impossible) metric of re-asked questions might be higher. A lot of closed questions aren't worth editing if the new version is going to be very different (as I think it would with a lot of too broad things; the OP is being asked to rethink his/her approach to asking for help, including prior research). Also, especially here, it can take a excessive amount of time to get something re-opened. So I'm not sure that "the ideal workflow" is to pursue re-opening -- depending on how you want to spin idealism ;D – goldilocks Jun 7 '18 at 22:39
  • @OyaMist I think generally, good conceptual questions are well-received here. Questions which blatantly don't fit in our format (e.g. by trying to cram their whole question into their title) tend to be more poorly received, but a genuine, clear conceptual question is entirely welcome on this site, even if it is more broad than a technical problem, so long as it's reasonably answerable. – Aurora0001 Jun 8 '18 at 11:58
  • How many active users on this site have sufficient rep to vote on a re-open? Just a curiosity since you say it's been 4 years since it happened. There are a lot of marginal questions asked here that should be closed, but it strikes me as odd that it would go so long without even one being reopened. – Brick Sep 5 '18 at 15:25
  • @Brick There are 37 users eligible to vote. 10 of those have visited since August 31st (probably the more active users). Keep in mind that if 5 users reopen, that's half of the active eligible userbase voting. And in my experience that rarely is the case—maybe due to lack of interest, but more often I think it's just that people struggle to edit their posts into shape and so their posts are voted 'Leave Closed' instead of reopened. – Aurora0001 Sep 5 '18 at 19:50
  • @Aurora0001 Thanks. Interesting as a curiosity. It takes 5 to close too, so that's almost everybody involved unless some people change their mind. I do see that many of the questions that are subpar must stay subpar and not deserve to be reopened though. – Brick Sep 5 '18 at 19:53
2

This site requires evidence of effort. There is no evidence of effort in these open ended questions. They haven't even bothered to check to see if the question is appropriate for this site.

The most appropriate site for such questions is almost certainly the official Raspberry Pi foundation forums.

  • 1
    Here's an example of a conceptual question that other forums have handled gracefully and helpfully. electronics.stackexchange.com/questions/233851/… We have trouble with this type of question. Why? – OyaMist Jun 8 '18 at 12:51
  • We are not a forum. – joan Jun 8 '18 at 14:05
  • 1
    Forums like Reddit are for endless opinions. In contrast, conceptual questions are broad but easily answerable succinctly with little argument. The Raspberry Pi system has tremendous scope from hardware to user interfaces. Our OPs struggle with this complexity and often need succinct guidance for broad concepts, especially in cases where the Pi itself constrains the applicable solutions. Also please note that my use of forum was in the general sense. The link is an SE site exactly like ours. – OyaMist Jun 8 '18 at 14:25
  • 2
    This is now circular. The appropriate place for such questions is raspberrypi.org/forums – joan Jun 8 '18 at 14:30
  • 2
    "We have trouble with this type of question. Why?" -> Do you have a parallel example to demonstrate this? Keep in mind that EE has a much broader scope than we do. Almost all of our questions could be said to cross over with either them, U&L, or S.O. So very theoretical things (like that "What is electricity?" question, or stuff like, "How does RAM work?", or architectural things about linux) simply don't belong here. We would be performing a disservice to users if we set out to do otherwise... – goldilocks Jun 8 '18 at 16:54
  • 2
    ...We simply cannot provide the focussed expertise that those other sites can. This is not because there are no experts here, but because there are more of them elsewhere. Which you can take as a sort of an answer to why we have trouble with conceptual questions, although I actually don't concur with that premise -- I think your use of the term "conceptual" is too ambiguous to allow for an objective evaluation (you can simply hand pick examples from here and elsewhere to "prove" your point). – goldilocks Jun 8 '18 at 16:54
  • I have upvoted Aurora0001's answer because it states clearly that comments should progress to chat. Not forum. I agree with that. – OyaMist Jun 9 '18 at 12:51
  • Yes. Focused expertise is indeed embodied more appropriately in other SE sites. The Pi unlocks knowledge that spans many sites. In fact, our own due diligence requires us to direct the OP appropriately. I simply think of conceptual Q&A as a request for a map to inform an OPs search. – OyaMist Jun 9 '18 at 13:02
  • Here's an example of an SE conceptual mapping question with several answers. Standard SE interactions are flowing normally without moderator intervention. electronics.stackexchange.com/questions/378873/… – OyaMist Jun 9 '18 at 13:40
  • 1
    OyaMist, The examples that you give of "conceptual" questions from the EE site lend themselves to definitive answers. Especially the first one about where electricity goes after being used. This is evidenced by the presence of relatively short answers to the questions that address them and are on topic for that site. As was noted in the answer by @goldilocks, conceptual questions that are still limited enough to get an answer can also be on topic here. The difference is that this site is oriented toward a specific piece of hardware, where there aren't really so many open "concepts" to address. – Brick Sep 5 '18 at 15:30

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .