12

The explanation for should be closed in the new flag dialogue for questions reads (boldface mine):

This question is completely unclear, incomplete, overly-broad, primarily opinion-based or is not about [site’s topic] as described in the help center, and it is unlikely to be fixed via editing.

The explanation for very low quality in the flag dialogue reads:

This [question/answer] has severe formatting or content problems. This [question/answer] is unlikely to be salvageable through editing, and might need to be removed.

I consider these wordings problematic, as many such posts can actually be fixed – by the poster. However, if I am not very much mistaken, one should not wait with such flags until it has become apparent that the poster has vanished or no intention of fixing their post.

Moreover, often somebody other than the author can fix these posts by applying an inappropriate edit. While it is clear to most high-reputation users that this not taken into account here, it can cause confusion amongst low-reputation users (remember that flagging is available at 15 reputation), who may get the idea that it is appropriate to radically change the post themselves.

I thus propose to change these wordings, e.g., to something like:

This question is completely unclear, incomplete, overly-broad, primarily opinion-based or is not about [site’s topic] as described in the help center. It needs to be fixed by its author.

This [question/answer] has severe formatting or content problems. It has to be salvaged by its author, and might need to be removed otherwise.

Note I did not explicitly include something like

can only be salvaged by its author (without making an inappropriate edit)

as I considered this to be too complicated. I am open for suggestions of better wordings that address this issue.

  • Note that while most new questions on SO (since the Triage cutover) can't be flagged VLQ, old questions still can, and presumably all questions elsewhere. So "This answer" isn't appropriate. – Nathan Tuggy Apr 24 '15 at 17:43
  • @NathanTuggy: You seem to have been commenting on the wrong post. – Wrzlprmft Apr 24 '15 at 18:30
  • Pardon? "This answer has severe formatting or content problems. This answer has to be salvaged by its author, and might need to be removed otherwise." is still misleading/awkward. That was my point. (There is no central question to handle this rather sudden revamp, or I'd have commented there.) – Nathan Tuggy Apr 24 '15 at 18:32
  • I think I understand you now. That was a mistake by me. See the updated question. – Wrzlprmft Apr 24 '15 at 18:39
1

The author/editor distinction is a red herring. There's no legion of skilled and willing editors wishing to fix every low-quality or off-topic post, prevented from doing so only by arbitrary rules... There are quite simply more people posting bad questions and bad answers than there are people to fix them, and in the vast majority of cases, trying to make sense of such posts is a cruel waste of time.

Yes, there are cases where a question could be answered were it not for the lack of some critical bit of information. The classic "I'm encountering an error when..." question which lacks any information on the error is as infuriating as it is frustrating...

...but unanswerable questions are actually rather easy to get rid of: the system deletes them automatically.

...and let's be honest: Very Low Quality on answers usually implies that the author cannot form a cogent sentence, so any phrasing which states that only an author edit can salvage the post badly misidentifies the problem, putting it into the lap of those cowed editors again.

The hard problem here is the questions that shouldn't be answered, and the answers that should never have been posted. The ones that encourage you to guess at their meaning, wasting the time of those who would answer and then many times over the time of future searchers who suppose their problem matches only to find a dead end. Answers that are unreadable; questions that imply a problem other than that which is eventually answered.

If every person who wrote an answer did so only after verifying that the question they answered was on-topic and reasonably-scoped, closing would rarely be necessary. If every person who wrote an answer took the time to re-write the question they answer such that it clearly stated the problem they'd just solved, closing would rarely be necessary...

...But neither of these happens, and so we have closing as a way of marking questions, "should not be answered".

And of course, VLQ exists simply because any time there's a textbox on The Internet, someone's gonna start typing garbage into it every few minutes and a few of 'em will manage to find the "submit" button.

For a post to be edited, someone - be that the author or anyone else - has to see enough value in it to devote the time to editing it. This is the qualifier that the flag dialog text attempts to capture, that here is a post so unloved, it will not be edited. If you can think of a better way to capture that...

  • That’s all very true – but we are talking about guidance for flaggers here, i.e., which gives them a good and easily applicable criterion when to use this flag. And flaggers cannot possibly make the connection from “unlikely to be salvageable through editing” to “here is a post so unloved, it will not be edited”. Also, for the purpose of flagging, it does not and should not matter whether the author returns to edit the post. The crucial criterion is the current state of the post. – Wrzlprmft Feb 21 '17 at 8:14
  • As this is a guidance for flaggers, what do they have to decide? 1) The post is of no use whatsoever in its current form (or a question that should not be answered). 2) A third person cannot fix the post with reasonable effort. — That’s admittedly a little bit different from your criteria, but I cannot imagine a case where the results would be different. And that’s more or less what my proposed wording captures. What confusion, wrongly flagged posts or other problems would arise by using it? – Wrzlprmft Feb 21 '17 at 8:22
  • I agree with your first comment; why do you contradict it in your second? The flagger shouldn't care about whether the author will return... So why are you trying to make an understanding of the author's potential actions an additional requirement for flagging? – Shog9 Feb 21 '17 at 14:35
  • So why are you trying to make an understanding of the author's potential actions an additional requirement for flagging? – Where do I do this? – Wrzlprmft Feb 21 '17 at 18:08
  • Both of your suggested changes add - in bold - the requirement that the post be salvageable only by its author. – Shog9 Feb 21 '17 at 18:13
  • If you so wish, yes. But being salvageable by its author is not really a strong requirement, as this applies to every post. But how does this “make an understanding of the author's potential actions an additional requirement for flagging”: It is completely irrelevant how likely the author is to actually fix their post. It only matters that nobody can fix this post – other than the author. – Wrzlprmft Feb 21 '17 at 18:26
  • Again, why are you hung up on the author, @Wrzlprmft? In every possible scenario where this matters, the author has already failed to fix the post. Their ability is the one known in this equation; you can only guess at who else might see the post or what they might know, but you know the author has seen it, has sat with the editor in front of them, and has left it in the state it appears. You wish a flagger to predict that only the author - who has already failed once - can possibly succeed in a task, when what matters is the flagger themselves. – Shog9 Feb 21 '17 at 18:35
  • You wish a flagger to predict that only the author […] can possibly succeed in a task – Yes, but that’s not a prediction about the author; it’s a prediction about everybody else. You could equivalently write “if anyone can fix this, it’s the author” or “unlikely to be salvageable through editing by a third person”. And that’s the point of this flag: The community cannot fix this. — Also note that I suggest “it needs to be fixed by its author” not “the author will likely fix this”. The flagger is not required to predict how likely it is that the author actually fixes something. – Wrzlprmft Feb 21 '17 at 18:51
  • Which brings us back to the critical problem that these flags exist to capture: terrible posts that are not going to be fixed. It doesn't matter who isn't going to fix them; a terrible post that won't be fixed by you is as much a problem as a terrible post that won't be fixed by its author. Hence, the "red herring" I referenced in my opening line. – Shog9 Feb 21 '17 at 19:37
  • terrible posts that are not going to be fixed – But that’s not what the current guidance (at least for very low quality) is about. It is not a good criterion for flaggers either because it requires flaggers to estimate how likely it is that anybody edits the post, which was your critique on my suggestion. — Anyway, if you are so opposed to mentioning the author, how about something along the lines of “This post cannot be salvaged by the community [with reasonable effort].”? – Wrzlprmft Feb 22 '17 at 12:49
  • Nothing wrong with that, @Wrzlprmft. How is it preferable to what we have now? – Shog9 Feb 22 '17 at 18:29
  • Well, see the problems mentioned in the original questions, most importantly that edits by the author are not excluded. Right now, somebody could see a post and think for example: “This question is incredibly broad without some specific information. Adding this information is something the author may do. Hence I do not think that unlikely to be fixed via editing applies. Hence I do not flag.” – Wrzlprmft Feb 22 '17 at 18:45
  • This is something you do? Or something you've seen/heard people doing? – Shog9 Feb 22 '17 at 18:52
  • It is something that confused me in my early days (though the wording was slightly different then, IIRC). I am also quite confident that it came up in discussions about interpreting the VLQ flag (something along the lines of “What does ‘unsalvagable by editing’ even mean? Everything is salvagable by editing.”), but I do not have a link to any example now – if I could go back two years on my network-wide activity tab, I probably could find the example that caused me to ask this question. – Wrzlprmft Feb 22 '17 at 19:04
  • Ah. Well, I introduced that wording to the main flag dialog a couple of years ago to combat the confusion we were seeing at the time (primarily on Stack Overflow, where correctly differentiating between "bad" and "no hope" translates to a huge difference in flagger and moderator work on a daily basis). The results were good at the time, but I'm certainly open to changing it again if we can nail down a desired outcome - my primary metric is % helpful. – Shog9 Feb 22 '17 at 19:11

You must log in to answer this question.

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