27
$\begingroup$

I have recently asked a question (The original did not turn up in my search): Show that there are infinitely many powers of two starting with the digit 7 which was marked as a duplicate of: Starting digits of $2^n$..

There was much more activity on the duplicate that I have asked and the answers on it seem to be more complete and comprehensive as compared to the original. What would happen to these answers (I assume they will be deleted)?

TL;DR What would be the actions taken if a duplicate has a better set of answers as compared to the original?

$\endgroup$
3
  • 1
    $\begingroup$ Related older post: Original post marked as duplicate. Other questions linked there might be of interest, too. $\endgroup$ – Martin Sleziak Apr 14 '17 at 6:12
  • 3
    $\begingroup$ The most likely reason why your question get so much attention (compared to more-or-less identical question) is that it got into the network-wide hot questions list. For the questions which are duplicates this was not even the possibility - since the title contains MathJax. (Kudos to the users who were able to find older posts with the same - or very similar - question.) $\endgroup$ – Martin Sleziak Apr 14 '17 at 6:16
  • 2
    $\begingroup$ You may not have seen all the reincarnations of the question. Remember that the dupe referrers are shown in the right margin as Linked questions. The 2013 version has 6 linked questions, the root dupe target has eleven. $\endgroup$ – Jyrki Lahtonen Apr 15 '17 at 4:40
15
$\begingroup$

What would happen to these answers (I assume they will be deleted)?

Typically, nothing would happen to these answers. By default, duplicate questions aren't deleted, and unless the question is deleted, the answers won't be deleted either. Poor quality duplicates without high-quality answers are occasionally deleted, but your question has a couple of high-quality answers, so that isn't to be expected.

What would be the actions taken if a duplicate has a better set of answers as compared to the original?

That depends on various details. If the questions are identical, they can be merged, so that all answers appear in one place. That's not an option here, though. Another possibility is to close the older question as a duplicate of the newer question. That could be thought about, but I consider closing the general "starts with an arbitrary string of digits" question as a duplicate of "starts with the digit $7$" not too appealing.

$\endgroup$
2
  • 1
    $\begingroup$ On the same lines as Glorfindel's comment, would editing the question to match the general case help? One can easily see that the existence proof given can be applied to any string of digits. $\endgroup$ – user424290 Apr 13 '17 at 21:14
  • 4
    $\begingroup$ Changing the question to match the general case would only be an option after the answers were edited to match the general case. That would unfortunately be very hard at least for the nice answer using $2^{46}$ as the starting value and multiplication with $2^{10}$ or $2^{53}$ for the step. $\endgroup$ – Daniel Fischer♦ Apr 14 '17 at 11:40
11
$\begingroup$

I guess that my strong antipathy towards duplicates is by now adequately documented. I want to add the following points.

  • Closing a question as a duplicate does not hurt the asker one bit, because they get directed to an answer to their question.
  • Closing a question as a duplicate does not hurt the answerer one bit, because they can post their answer to the original version equally well. This is particularly unproblematic if the answerer is offering a slightly different angle (which is what Jeff Atwood apparently was hoping to achieve when dropping his opposition to duplicates in the cited blog).

So why the opposition? Here's a theory. A bit simplified, but that's how I see it in a bit crude (and possibly also rude) way.

  • Closing a question as a duplicate only hurts the rep farmer answerer who wants to repost the same answer/trick to a fresh question without having to face the competition of older answers that have stood the test of time.

On the other hand, our on-site search engine is not very good at finding duplicates. Therefore I cannot fault anyone too much at accidentally posting an answer to one. The resistance to dupe closing is what baffles me.

$\endgroup$
10
  • 2
    $\begingroup$ Thanks for explaining your perspective. Would you care to respond over here to explain how your reasoning applies there? For the record, this question was closed nearly 4 years after it was asked and 3 1/2 years after I answered it. There are comments from an important researcher in the field and I feel comfortable in saying that both answers are valuable contributions. I think this is exactly the situation Jeff Atwood describes. $\endgroup$ – Mark McClure Apr 15 '17 at 14:36
  • 2
    $\begingroup$ @MarkMcClure I see your point. If it were up to me that question would not be closed. I am targeting cases like whether this and this are dupes or not. Or, whether these are not duplicates of this. IMO they are. $\endgroup$ – Jyrki Lahtonen Apr 15 '17 at 15:53
  • $\begingroup$ Thanks Jyrki. Again, I do think there are plenty of duplicates that should be closed; this certainly applies to the things you linked. But I did appeal via the standard channels to attempt to reopen my linked question without success. That's the way it is and I accept it but I think the site benefit from a softer stance towards users who've made some good contributions. $\endgroup$ – Mark McClure Apr 15 '17 at 16:09
  • 1
    $\begingroup$ Any principle, when applied blindly, will have undesirable consequences. If only I managed to keep that in mind myself :-( $\endgroup$ – Jyrki Lahtonen Apr 15 '17 at 17:32
  • 1
    $\begingroup$ @MarkMcClure an issue is that you continue to fail to make clear what the problem with that dupe closure even is. As far as I can see the sole problem might be with some users' ego. $\endgroup$ – quid♦ Apr 15 '17 at 19:43
  • 2
    $\begingroup$ +1 for the description "rep farmer answerer". $\endgroup$ – Paramanand Singh Apr 16 '17 at 3:57
  • $\begingroup$ @quid True - I find it self-evident. $\endgroup$ – Mark McClure Apr 16 '17 at 15:38
  • $\begingroup$ @MarkMcClure looks like a very cheap cop out to me. If one cannot easily put in words something one finds self-evident it can be prudent to reconsider how evident it actually is. $\endgroup$ – quid♦ Apr 16 '17 at 17:27
  • $\begingroup$ @quid That's ridiculous. When one question is closed over another - a choice has been made and the closed question is diminished. And, yes, I find it insulting. Whether my ego is bruised - I don't know. I'm quite comfortable with my contributions here. I do find it strange, though, that when one Googles "differentiable function with discontinuous derivative" the top hit is this closed question. I think that reflects poorly on our community. While you seem to be a valuable contributor with a lot of outstanding insights, I think we've established that you and I disagree strongly on this issue. $\endgroup$ – Mark McClure Apr 16 '17 at 18:23
  • $\begingroup$ @MarkMcClure indeed, we seem to disagree on various points related to this. (And you chose to ignore my advice on how you could likely, easily resolve the situation. To repeat: edit the question to highlight the difference.) While several details related to your position remain unclear to me, I will leave it at that. As a final thought, if you do not wish to discuss this subject, you might simply refrain from bring it up yet again. $\endgroup$ – quid♦ Apr 16 '17 at 18:45
4
$\begingroup$

Answers to duplicate questions aren't deleted, and neither is the duplicate question itself. From the help center:

Some duplicate questions may eventually be deleted, but often they are left as a signpost pointing people towards the canonical answer to that question.

I agree that the newer question (yours) has better answers, but in this case it's not a good idea to 'reverse' the duplicate (which could be done by reopening your question, and closing the other one), because the older question asks a more generic question. If the questions would be identical, ♦ moderators can merge them. In general, it's certainly possible to close an older question as a duplicate of a newer one.

You could ask the (top) answerers to your question to post a modified version of their answers as an answer to the older question.

$\endgroup$
2
$\begingroup$

Perhaps, neither should be marked as a duplicate?

For the record, some level of duplication is not at all bad. On the contrary, some level of duplication is good, as far as search optimization and expression of multiple viewpoints go.

That's not my opinion, rather it's argued for quite effectively in this blog post by StackExchange co-founder Jeff Atwood. Unfortunately, it takes just a few dedicated enforcers to screw that up.

$\endgroup$
2
  • 4
    $\begingroup$ If you scroll to the end of Atwood's blog post you will see that he is advocating the idea of having 4-5 "live" versions. I once tried to talk an ultra high rep user to post to an "abstract duplicate" thread instead. They replied giving a link to a google search of 1) our site name +2) the username +3) their pet technique. Google gave 264 hits. The difference between 4-5 and 264 is, IMHO, the real problem with duplicates on our site. But even if you try to enforce that difference; there will be no end to the wailing. $\endgroup$ – Jyrki Lahtonen Apr 15 '17 at 4:21
  • 1
    $\begingroup$ @JyrkiLahtonen 4-5 sounds good to me. The problem is that folks here don't tolerate a single duplicate - even if it's a good one, as I learned here. I agree duplication is a problem here. In fact, I searched for duplicates of my post and found many of them that hadn't been closed. That's what makes treatment of a good duplicate that much more insulting - a high quality duplicate is more likely to be closed by virtue of the attention is receives. $\endgroup$ – Mark McClure Apr 15 '17 at 10:48
1
$\begingroup$

I support the current system of marking questions duplicates. I would, however, stop at closing duplicate questions. Rather, the mark of a duplicate merely informs the OP and other users that the same question exists and here are other answers. People may then become aware of duplicate answers and point them out appropriately.

$\endgroup$
3
  • 3
    $\begingroup$ I do not understand this answer. The "current system of marking questions duplicates," which you support, is such that it is impossible to add new answer to questions marked as duplicate; it is not merely informational. A question marked as duplicate currently is a closed question, which you do not want, if I understand you correctly. Thus, I see some incoherence, possibly due to the fact that you misunderstand the current system. (For some things dupes are treated slightly differently from questions closed for other reason, but this is somewhat tangential.) $\endgroup$ – quid♦ Apr 15 '17 at 8:43
  • $\begingroup$ All I am saying is that currently dupes are closed. I wish they would remain open. $\endgroup$ – Ron Gordon Apr 15 '17 at 16:56
  • $\begingroup$ Alright. I still find this answer very hard to parse. $\endgroup$ – quid♦ Apr 15 '17 at 19:41

You must log in to answer this question.