The reason that we don't allow reposts is because the same answers will be copied from the first challenge and the original can still be answered. But what about answer-chaining challenges? Those are inherently different if they were run multiple times. Imagine if the challenge Add a language to a polyglot started with LISP, it could have ended up being a completely different challenge. (that might not have been the best example). The point is, do you lot think it could be acceptable to repost such challenges, provided we don't spam them over and over? It would enable languages that at this point can no longer be reasonably inserted to compete, and could provide different kinds of challenges.
10\$\begingroup\$ I'm torn on this. On the one hand, different initial conditions would lead to different, interesting variations on the challenge. On the other hand, we don't want dozens of challenges that are all the same except for their initial conditions. \$\endgroup\$– user45941Jan 1, 2017 at 21:20
Yes, under specific circumstances:
I think that DJMcMayhem made some really good points as to why we don't like duplicates. I'll address the two he felt were still relevant to answer-chaining:
- A challenge becomes less interesting when it's been done before. Even challenges around a theme become boring after being repeated too many times. (For example, flag challenges, "Print the alphabet in this pattern" challenges)
- Taking credit for the creativity of another user who wrote the original challenge is not OK.
I believe that the following measures would solve both point 3 and 4:
Require that reposts be posted by the challenge author, or with their permission.
Require that a repost be posted at least 6 months after the original post, maximum 3 times.
Both measures would act as a natural limiter to ensure that reposts are high-quality. It doesn't eliminate the problem of low-quality reposts, but I think that the benefit outweighs the cost.
\$\begingroup\$ If the original author no longer used the site, can we say that if they don't respond in a month, they give permission? \$\endgroup\$– PavelJan 3, 2017 at 2:43
\$\begingroup\$ @Pavel I'm on the fence on that. I'll leave that up to another meta post if it becomes an issue \$\endgroup\$ Jan 3, 2017 at 2:44
7\$\begingroup\$ 1 month seems a bit short, especially if you're limiting it to three times. Make it a year or at least half a year. That also allows more interesting languages to be created that can change how the challenge plays out. \$\endgroup\$ Jan 5, 2017 at 10:08
\$\begingroup\$ @MartinEnder the challenge in question (Add a language to a polyglot) is just about to turn 1 month old, and has pretty much finished. If you make it much longer, I think you'd lose interest in it. However, I'll take upvotes on your suggestion into consideration, and change it if enough support it. \$\endgroup\$ Jan 5, 2017 at 14:04
1\$\begingroup\$ @NathanMerrill I don't think losing interest is relevant. It'll get attention when it's reposted and if the time between is longer then it's actually likely that it will get more interest. \$\endgroup\$ Jan 5, 2017 at 14:09
\$\begingroup\$ Random idea: instead of having a set time limit and max number of reposts, we could have a thread similar to the sandbox where there could be "repost requests", and in order for a challenge to get the thumbs up for reposting, it must reach (insert ratio here) upvotes to downvotes \$\endgroup\$ Jan 8, 2017 at 1:36
\$\begingroup\$ @ConorO'Brien that's not a bad idea. Perhaps post it as a separate idea here? \$\endgroup\$ Jan 8, 2017 at 2:25
\$\begingroup\$ @NathanMerrill Done. \$\endgroup\$ Jan 8, 2017 at 2:38
\$\begingroup\$ Why have a maximum of 3 times? Have a good, long wait period sure, but PPCGers 100 years from now may be mad that they can't redo historic challenges. \$\endgroup\$ Feb 26, 2017 at 4:53
\$\begingroup\$ Also, I'd say an author's permission is not required if the don't respond to questions about their challenge within a month. Not all authors still frequent SE. \$\endgroup\$ Feb 26, 2017 at 4:56
Yes, but the reposting of a challenge must be voted on
Instead of having a set time limit and max number of reposts, we could have a thread similar to the sandbox where there could be "repost requests", and in order for a challenge to get the thumbs up for reposting, it must reach a certain ratio of upvotes to downvotes. A request might look like this:
# [challenge](link/to/it) <Citation of eligibility for the user wishing to repost> <Any additional information>
3\$\begingroup\$ This could also include reposting old challenges with unclear specifications. \$\endgroup\$– PavelJan 8, 2017 at 4:09
4\$\begingroup\$ @Pavel I do like that idea, but that's beyond the scope of this question \$\endgroup\$ Jan 8, 2017 at 4:26
\$\begingroup\$ Do we need a separate thread, or could this use the existing sandbox? \$\endgroup\$– DLoscJan 9, 2017 at 23:01
1\$\begingroup\$ @DLosc I think this would warrant a separate thread. \$\endgroup\$ Jan 9, 2017 at 23:02
This should not be allowed, but we could lower the threshold for these challenges to be considered a "duplicate".
The reason that we don't allow reposts is because the same answers will be copied from the first challenge and the original can still be answered.
That's certainly not the only reason. There are four reasons I can think of to not allow duplicate challenges.
Like you mentioned, answers from the first can be copied and posted on the new one,
The old challenge can usually still accept answers,
A challenge becomes less interesting when it's been done before. Even challenges around a theme become boring after being repeated too many times. (For example, flag challenges, "Print the alphabet in this pattern" challenges)
Taking credit for the creativity of another user who wrote the original challenge is not OK.
Now you are right, answer-chaining challenges address most of these points. 1 is moot because a tiny change in the first post could drastically change future answers. 2 is addressed somewhat, because lots of answer-chaining challenges have a deadline, or are impossibly hard to add a new answer to. (For example, 95 movie-quotes)
But that still leaves points 3 and 4. Reposting answer-chaining challenges is still taking credit for someone else's creative challenge idea, and still gets old if done too often.
So I think that we should not allow directly copying an old answer-chaining challenge, but I think it's perfectly reasonable to be slightly more lenient about what we consider to be a duplicate.
2\$\begingroup\$ Side note: That particular challenge has an incredible amount of participation, and the community's response has been overwhelmingly positive, with lots of fantastic answers. Personally, I wouldn't mind making an exception and reposting it much later once the opportunity for new answers is all gone. It would probably need to be a community wiki, and we'd definitely want full community consensus before posting it. Of course, this is completely tangential to my answer. \$\endgroup\$ Jan 2, 2017 at 19:24
4\$\begingroup\$ You said "be more lenient", but you haven't said how. \$\endgroup\$ Jan 3, 2017 at 2:12
Please note that, even if this technically works, it actually lets you post dupes one after the other. That loophole should be eventually fixed. Any feedback for that instead of just downvotes is appreciated.
Yes, allow reposting, but:
- Make sure that you cannot solve the original challenge or any of its already existing reposts. If you can, there is no reason to repost, as solving at least one of them isn't impossible.
- Adding to the above bullet-point, if an answer is posted to a repost older than yours or the original challenge, and your challenge didn't have answers at the time the new answer was posted, your repost is subject to closing as a dupe of that repost. Please keep up with the spirit of answer-chaining. Answering such a repost is extremely discouraged, although there is no easy way of determining whether a newer answer to an older repost or the original challenge has been posted. The only way is to look through each older repost and see if an answer has been posted there later than the posting of the repost question. Note that if the question poster is the only answer poster, it still counts as a dupe.
- Credit the original author, not subsequent reposters, and link to the original challenge. Linking to the original challenge will add the repost to the original challenge's "Linked questions", thus marking the existence of your repost. If your repost is closed as a dupe, the link is subject to be removed, or replaced with a no-ad URL-shortened link (such as goo.gl or bit.ly), so as to not clutter the original question's "Linked" list.
- Self-contain the original challenge AS-IS, do not just give a link to the old challenge which might have been deleted, so that users without moderator privileges can still read the challenge's spec.
- If the original challenge has been closed, do not repost it yourself. Doing such a thing means your challenge will most probably be closed too, as the content is AS-IS.
- Follow the exact formatting described below.
<original title> (Repost <repost number>). Replace
<original title> with the original post's title, and
<repost number> with the number of your repost (first repost is
1, second is
**Note that this is a repost of the original challenge linked [here](<link>).** --- <original challenge>
<link> with a link to the original challenge (not URL-shortened unless repost is closed) and
<original challenge> with the contents of the last revision of the original challenge where no answer was accepted. The link must be in the form of
<id> is the original question's ID.
Tags must be the same as in the original challenge. No added or removed tags please.
\$\begingroup\$ I don't really like this answer because it doesn't address the problems with duplicates that I mentioned. IMO Nathan's answer is a better
Yes this is OKanswer because it doesn't have any loopholes, addresses the issues with duplicates, and is a little more restrictive about when it would be OK to repost a challenge. \$\endgroup\$ Jan 4, 2017 at 0:12
\$\begingroup\$ @DJMcMayhem Nathan's answer has a too restrictive "at most 3 times", which I generally disagree with. About the loophole, it does not really have a good fix, unless someone goes through all of the older reposts of the original challenge and the original challenge itself, and checks whether an answer has been posted there at least a second after the repost was posted, and that the repost has no answers or only has answers posted by OP. Also, if OP has a second account, they might post a starting answer with that, and then the challenge goes on. That's why I have stated such a loophole. \$\endgroup\$ Jan 4, 2017 at 9:36
\$\begingroup\$ I really debated about the "3 times", but I think its a good limit before the community gets sick of a particular style of challenge (I believe we started getting sick of alphabet challenges at around 3-4 of them). \$\endgroup\$ Jan 4, 2017 at 14:34
\$\begingroup\$ That said, my biggest problem with this method is that its really hard to determine if a challenge can't be answered anymore. The entire point of answer-chaining problems is that they get progressively more difficult, and just because I can't figure out the next link in the chain doesn't mean nobody can. Either the requirement will scare reposters away, or they will just ignore it. \$\endgroup\$ Jan 4, 2017 at 14:36
\$\begingroup\$ @NathanMerrill I don't think there is any method making it easier. Maybe we should agree with DJMcMayhem's answer then. \$\endgroup\$ Jan 4, 2017 at 14:37
\$\begingroup\$ Right, but I'm personally ok if there are two "active" answer-chaining challenges. The likelyhood of it happening with my method is low, but even if it is occurring, that's fine to me. They'd be drastically different challenges after a month. \$\endgroup\$ Jan 4, 2017 at 14:38
\$\begingroup\$ @NathanMerrill Well, I was actually of the opinion that the answer-chaining tag should be obsoleted once and for all, and I can't think of a better way of solving answer-chaining's issue. Yes, the biggest problem is the tag's spirit, which is that the challenge itself is to try to create an answer based on the challenge's rules and the existing last answer. \$\endgroup\$ Jan 4, 2017 at 14:43