Why deleting a question (and related answers) with a lot of votes

Question

Today the following question was deleted:

Should I change my Python code if I upgrade my Ubuntu? [closed]

I totally agree with the too broad votes (with the benefit of hindsight) but I'm now a bit puzzled by the deletion not only because of the severe loss of rep.

I consider my answers to this questions good answers (I got many votes) as I provided a way to be notified of build failures with development versions of Ubuntu using a ppa and a few examples where new python versions caused me some troubles, which I still find useful.

I've voted to undelete this question per the following policy:

Before voting to delete, please check that there are no good answers; if so, then the question should be flagged for moderator attention as a potential merge candidate. We don't like to lose great answers!

Again I agree with the closure but ideally I'd have left this question for reference.

What's your opinion?

Answer 1

For the moment, I have gone ahead and undeleted the question. This is so that those AU members who are under 10K can have visibility. Their views are equally valid as those 10K+ members.

That's said - if the consensus is that this Q should be deleted - then we'll slam the door shut again.

---

In my humble opinion - this is a case where a question should NOT have been deleted. The criteria leading me in this direction was that two high rep answers indicating that those answers are informative and contain enough information to warrant the "good answer" clause in the policy:

Before voting to delete, please check that there are no good answers; if so, then the question should be flagged for moderator attention as a potential merge candidate. We don't like to lose great answers!

It has a fair number of views - not massive, but enough to warrant that people are looking at the question - or at least searching on the terms being used in the title/question/answers.

In deleting questions we should all be cognisant of that policy - look at the answers, the vote spread and the views. If the answers have a decent number of votes - this classes the answer as fairly important. The question should therefore not be just eliminated. However the closure is warranted given the broad nauture and incomplete details given.

Answer 2

I'm inclined to think this question does not even need to stay closed.

It's a quite general question, and the OP may personally be helped more if they expand their question to include specific details of the software they're developing and maintaining.

Although it's general, it has good answers that are mostly based on objective considerations. And there's really only a small handful of possible right answers to this question. I could imagine a whole book written about issues related to that question, but I cannot imagine a whole book written just to answer that particular question.

Basically, software developed in a way that might rely on specifics of an earlier minor release (the y in x.y.z) of Python than the system provides can either be tested with the newer version and modified as appropriate, or it can be run with the older version of Python, such as with virtualenv. It's not as though there's an unlimited number of materially different answers to this question. The handful of answers already posted cover the range of possible solutions pretty well.

This is the sort of question that might be benefiting the community as a whole more than the individual who asked it. A definitive answer for what they should do personally cannot be given, but the question can be--and has been--answered by explaining the considerations and the options available. Notably, the answers there that do that aren't even particularly long.

---

Naturally, each of us has somewhat different ideas about what questions ought to be closed and for what reason. A reader may or may not accept my above argument for why this question doesn't need to remain closed. I think it is not actually too broad in the way that's relevant for deciding closure. I expect some others will disagree. Either way, though:

• A reasonable argument can be articulated for why this question is perfectly fine for our site and could even remain open.
• The question is answered with answers that clearly have some value, even if votes and accepts are disregarded.
• The question, and at least one of its answers, have garnered plenty of votes.

Any one of those reasons should motivate special care before deleting (or voting to delete) a question. I think the presence of all three strongly suggests deletion is very unlikely to be an appropriate fate for this question.

Answer 3

This is a typical example of why deleting should be used very sparingly. There are certain users who believe that anything that isn't a perfect example of the platonic ideal of an AU question should be deleted. There have been many, many, many discussions on meta about this and most of them have concluded that deletions should be used only for extreme cases. Despite this, some users insist on deleting anything they don't like.

This question is a perfect example. Yes, the question is not great and should be closed. However, there are two perfectly good answers there, both of which can help future users. Why in the world would we want to delete it? Why lose good content?

Closing the question sends a clear message that it's a bad fit for out site. That serves to educate both the OP and future users. Great. Now, leave it alone so others can benefit from it.

Answer 4

Can we please forget about the fallacy of "many votes" == "good"? Most of the most upvoted things on the entire network, you ask them nowadays and you will get a rain of downvotes/closevotes.

The fact that a question was/is "popular" at some moment of the time, and gathered lots of upvotes, despite being bad questions by today's standards, needs to be removed of the site. That is one of those cases where drive-by bad voters (judging for the quality of the votes) or that was asked when is was considered a good question.

---

Now, moving on. The question is intrinsically bad for the Q&A SE model. Is unclear, too broad and primary opinion based.

Is unclear since it's vague. Is akin to "ok, I have a program, can I upgrade, because I haven't tested anything?" (and please, that's what the question says). It doesn't show any relevant piece of code that would potentially break during the upgrade, nor any detail that would help trying to figure out if his software would even break. This is even more supported by a comment asking more information which OP decided to ignore.

Too broad, since any and every answer would be equally valid. You can say, "if you don't use these functions [puts list of functions] you can upgrade without problems" as you can say any other thing. There isn't a wrong answer.

Won't a recent Ubuntu version affect my software in any way ?

You can gather opinions too! The above question can be answer with "I did with my program, you are going to be fine" or "my program broke terribly, and had to rewrote it". That kind of experience/expertise is ultimately unuseful. If I don't have a program exactly the same as the one providing the answer, is not going to help me. But, hey, someone had the same experience, hence upvoted the answer, and there goes quality.

---

OP haven't made any effort to fix his question at all after it was closed. In fact, each edition just made it worse while it was opened and I highly doubt he would fix it now that he already got the answers he wanted. He got suspended, so I doubt any effort right now would help either, so I just voted delete the question.

---

Now,

Solutions for the good answers. (?)

Again I agree with the closure but ideally I'd have left this question for reference.

If you think that your answer is good but the question doesn't make live up to the goodness of your answer, just re-ask the correct question, with a clear statement, and post again your answer. I've done that too! I didn't get as many upvotes, but hey, I never said I do it for the money imaginary internet points. So, if you were me, and you aren't doing it for the imaginary internet points, I would just ask another question where your answer would be found easily (if you want pointers about how can you ask your question just ask in comments).