7

I flagged this question as off-topic: Feature detection for embedded platform OpenCV

My rationale was that the question was a general programming question about OpenCV that had no apparent connection to Raspberry Pi other than that the OP said that they happened to be using that hardware. The same question could have been asked on any hardware though.

My flag was declined, which is fine. That happens sometimes, but it seems to me this is part of a larger trend, so I'm asking about it here. The site help, under "What topics can I ask about here?" includes

Software specific to the Raspberry Pi. [Emphasis original]

It further specifies in the enumerated list of things that are explicitly off-topic:

General programming questions, e.g., "How do I implement a keyboard interrupt in python?".

The question that I've picked as an example here is the first one that I flagged, but it seems to me that the site is flush with questions that are either straight programming questions or straight Linux/Unix questions (which are also explicitly off-topic according to the site help).

So my meta-questions:

  • Is there some standard that this community is using to allow / disallow these questions?
  • Is the practice of the community consistent with the help?

The former would be good to document, even if only in meta. If the answer to the latter is "no," then it seems that an update to the site's help would be appropriate.

5

Thanks for bringing this question up again. I think it is reasonable to discuss it again to determine how the community thinks of this now and how we agree to handle it. Maybe it's time to reconsider how we apply our own standard and make the helpcenter reflect the practise.

This ist the most recent discussion on the topic (early 2016): What defines the borders between RPi and other related Stack Exchange sites?

Personally I used to go with:

When in doubt, ask there first. Questions where the Raspberry Pi is not a significant factor will likely be closed. [emphasis mine]

However as a moderator I try to not hammer things down without a few votes by the community (sometimes I still do if I feel it is the right thing to do or if the question is blatantly off-topic). So usually votes need to be carried by community - that is, five voting members to agree an a certain position. So as Goldilocks' has put it in his answer to the meta question linked above: "this depends of the willingness and availability of participants" and might take some time.

While we need questions to be at least a little specific to the RPi Milliways has brought up an important point (again in the discussion linked above):

I agree, although pragmatism sometimes prevails. If a new user asks a 'NIX question I will often answer, because the questioner is often new to LInux and the Pi is the first and only experience. Besides Unix & Linux (Operating systems) can be intimidating. Users there don't seem to understand absolute beginners and confuse them with technicalities.

This is usually fine if a question receives a helpful answer at our site. If it does not however, one might still consider asking at the larger and more general sites such as SO or U&L. Note that migration of such questions is not always the best approach. Usually hinting to ask alsewhere while closing a question here works better if the question need some polishing to work well on another branch of StackExchange.

I have experienced that most users of our community tend to give quite some leeway to questions of users that are new to the Raspberry Pi and/or StackExchange... and again if the question does not blatantly violate the scope of our site and the respective user gets a helpful answer here we could call it mission accomplished. Given time and a few guiding comments one could still point out the nuances of the workings of StackExchange and how and where to post additional questions.


The openCV question that you have linked however is an example of how it's not done well. It is technically completely unrelated to the Pi and a solution that would perform well on any other Linux board is likely to do the trick on the Pi too. The question received just one short answer and one could very well argue that the significantly larger user base on SO could have been beneficial to better answers (or impromptu close as a duplicate).

Closing it now however will change little as it will still persist on our site (question with answer). It is also too old to migrate away. The only good that it would do is a sort of setting a precedence here... but I think this is done better discussing the issue here on meta and handling new questions.

  • If there's a difference of opinion about whether it's on or off topic, that's one thing. If the position is that I was correct but the mods are going to decline it anyway, that's bad moderation. The "decline" as opposed to "disputed" or "aged away" is intended to be a slap on the wrist of the person who flags and comes with some penalties to the flagger. (meta.stackexchange.com/questions/141292/…) If you're declining flags that are nominally valid you are setting a precedent. – Brick Aug 8 '17 at 0:12
  • Also as a point of fact, your comment about the question persisting because it has answers is, AFAIK, wrong. Questions with answers can be deleted by moderators or by the community: stackoverflow.com/help/deleted-questions (I don't think that's specific to SO.) – Brick Aug 8 '17 at 0:15
  • @Brick: he meant that it'll persist in the sense that it will no longer be auto deleted by the community user. – Jacobm001 Aug 8 '17 at 1:57
  • @Jacobm001 That really begs the question though, doesn't it? Closing it was the first step toward the community deleting it, but the answer claims that there's no point in closing it now because it won't be deleted. There definitely would be a point in closing it (assuming that it is off-topic, which Ghanima at least seems to agree) because it could ultimately be deleted. Related to the Community user, I'd point out that it came to my attention because Community bumped it to the home page specifically to have the content reviewed. – Brick Aug 8 '17 at 2:06
  • @brick According to this meta.stackexchange.com/q/5221/284388 is a question with a positive score is not auto-deleted - which is what I was referring to that the Q&A would persist on site. It is however possible for the community (not the community-user) to delete it using their votes (as lined out in the helpcenter as you linked - fair enough. To clarify my "no point" argument (which literally is not a "no point" but a "little point"): I was not arguing that it would not serve any good to have it closed (or deleted)... – Ghanima Aug 8 '17 at 18:58
  • ... but instead that the broader question how to handle the issue in general is helped much more with this very discussion and a new community aggreement to the flagging and handling of those flags for new questions right as they pop up. However, deleting existing content (that could be at least a little helpful) to future readers is not a good way (and not how SE usually works). Above linked meta post states What are the criteria for deletion? -> For questions, a post that no longer adds anything to the site should be deleted. ... – Ghanima Aug 8 '17 at 18:59
  • ... So while it was probably not the best decision to not close the question back when it was posted that does not automatically mean it should be deleted now since it is adding anything to the site. But again I think the key issue here is not the deletion of this particular question but to get the practice of closing questions and what is stated in the help center in line. – Ghanima Aug 8 '17 at 19:00
  • At the time that I flagged it, I didn't realized how old it was. The question still (as of right now) has a banner on it saying that "Community" bumped it to the home page to have the content reviewed. It was at the top of the "active" questions lists when I clicked on it for the first time. I agree that scouring past questions for things that should be closed is a waste of time. But if they pop up for one reason or another, they should be handled by whatever rules are in place. Also, I'm not necessarily advocating deleting. I am advocating closing so that it follows the normal review flow. – Brick Aug 8 '17 at 19:47
  • @brick, point taken. I agree with having it follow the normal review process and community moderation procedures. I will close this particular question to cure Jacob's mistake, leave the rest to the community. Further discussion of the underlying issue as pointed out by Jacob's comments to his answer notwithstanding of course. – Ghanima Aug 8 '17 at 20:15
3

I was the moderator that declined your flag. I felt that the question was acceptable because the OP seemed to be focusing on how to tune general software for the RPi and its relatively unique development requirements. I admit, that's an arbitrary line.

You'll see a lot of answers that aren't particularly on-topic here. That's the product of the fact that our community is rather... odd. Many of our users are very new to Linux, embedded things, IOT, web dev, robotics, or insert subject here. At the minimum, they're often dealing with 2-3 new subjects and they honestly don't always know where the lines are.

These new users don't know what's a general programming issue, what's a linux issue, and what's a software issue. Simply closing the questions and trying to send them elsewhere gives people a bad impression and would frankly, leave us with seemingly few questions. Plus, it often just feels mean to me.

Given all that, users are often given a lot of leeway here. What we're left with is a tricky balancing act, and it's something we need to work on and better define.

One of our moderators is currently on a vacation that doesn't include much in the way of internet. When he's back, I'd like to reopen this discussion and see how the community at large feels about the matter.

  • I put some comments on the other answer that also apply here. Specific to your response, my bigger issue is not what is or is not on topic, but the inconsistency between how it's defined in the help and how it's implemented on the questions. If the answer is to change the definition of what's on topic that's fine. If the answer is that we cannot be bothered to make a definition that we believe in and stick to it, that's bad on the mods and the community. – Brick Aug 8 '17 at 0:18
  • @Brick: I agree. It is wrong, and it's been wrong for a while. I think we should have a communal discussion about it. I'm hesitant to do it without goldilocks present since he is both our oldest moderator and one of our highest ranking users. – Jacobm001 Aug 8 '17 at 2:05
  • @Brick: I apologize for the erroneous declination to your user history. It doesn't really mean anything as far as your account standing is concerned, but that doesn't make it more fair. – Jacobm001 Aug 8 '17 at 2:09
  • 1
    I don't care about the one decline. I understand the minimal impact of a single decline and even that mistakes happen. Not the point here. If it's the policy to decline flags that are valid by the site's description, though, that's different and what I hope gets discussed and/or corrected here. The answers so far and the discussion on the link to the previous round of discussion make it seem like people feel like it's ok to be inconsistent, which I think is not good for the site and not congruous with the rest of the SE universe. – Brick Aug 8 '17 at 2:17
  • Looks like goldilocks is back, @Jacobm001. Worth bumping this back for discussion? – Brick Sep 5 '17 at 15:43

You must log in to answer this question.

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