7
votes
$\begingroup$

A couple of years ago we voted on our site-specific off-topic close reasons. Since then we have updated them every now and then, improving formulations and the like.

However I have noticed that perhaps our "programming question" close reason isn't as useful anymore now that the mods migrate questions more aggressively. It is my subjective feeling that maybe we want to replace this close reason, e.g. with on relating to our recent homework consensus?

Status Quo

Based on data from the last 90 days, here's our current usage of close reasons and our comments (link requires 10k rep to access):

data

So, what are custom close reasons?

Custom close reasons are hidden behind the "off-topic" close reason. They are specific to the site (to reflect what each site considers on- and off-topic) and can be changed by the moderators as soon as two agree (one proposes, one approves).

How many custom close reasons can we have at the same time?

There are only three (3) custom close reasons for each site on the StackExchange network.

What should I include in my suggestion (= posted answer)?

  1. A TL;DR as a headline. For example, by writing:

    ## Reference recommendation

  2. The formatted text to be displayed in the close reason menu. For example, by writing:
    Requests for reference recommendations are off-topic here. For details, see: Do we want “literature recommendations” and similar “list/subjective questions”?

  3. An explanation of when to use this close reason (may be omitted if obvious)

  4. Links to all relevant meta questions (if any) concerning this close reason. For example, for "Reference Recommendation", you would be expected to add a link to the question Do we want “literature recommendations” and similar “list/subjective questions”? by writing:

    - [Do we want “literature recommendations” and similar “list/subjective questions”?](https://crypto.meta.stackexchange.com/q/1/)
  5. (optional) Remarks / comments, such as "Remark: As of the time of this post, this is an established close reason" or "Remark: This is a merge of close reason A and B"

The above formatting can be witnessed by example using the pre-provided answers for our existing close reasons.

How does this "election" work?

  • Question Score

    In the unlikely case that this question itself scores less than +6 by itself, then the need for a change is considered non-existent and the election is void. If the question score less than -6, I promise not to ever again come up with such a question.

    • Vote +1 on the question to indicate that you want to (regularly) update the close reasons.
    • Vote -1 to indicate that you don't want to see a change with the close reasons at all and generally disapprove the idea of electing custom close reasons.
    • Don’t vote to indicate that you don't care.

  • Answer Scores

    Each answer contains one close reason suggestion. After one month (i.e. at the end of January) we come back to this question and see if there is a strong need to change our exisiting close reasons and replace them with more useful ones. The top three voted close reasons will be our custom close reasons given, that they all score more than +6 (the same threshold as for community ads). If less than three answer score above the threshold and there's a change, then the established close reason with the least votes gets kicked out.

    • Vote +1 on an answer to indicate that you want that reason established.
    • Vote -1 to indicate that you don't want to see that reason.
    • Don’t vote to indicate that you don't care.

  • Deadline for Votes

    The deadline (for this election) is the last day of January (aka the 31st January 2019) and the elected actions will be carried out in early February 2019.

$\endgroup$
1
  • $\begingroup$ The changes have been carried and this question is now locked to provide evidence of the voting results. $\endgroup$
    – SEJPM Mod
    Commented Feb 3, 2019 at 12:48

5 Answers 5

Reset to default
11
votes
$\begingroup$

Programming Questions

Programming questions are off-topic even if you are writing or debugging cryptographic code. Unless your question is specifically about how the cryptographic algorithm, protocol or side-channel (mitigation) works, you should look into asking on Stack Overflow instead.

When to use:
Use this whenever a question should either be migrated to stackoverflow (but say so in a comment and / or flag for a mod to migrate) or when the question is about programming but shouldn't be migrated due to it being a poor fit on SO or it being likely closed there.

Relevant Meta Q&As:

Remarks: This is one of our current custom close reasons.

$\endgroup$
8
votes
$\begingroup$

Ciphertext / Scheme Analysis

Requests for analyzing ciphertext, finding hash preimages, identifying or decoding some code, or even reviewing full cryptographic designs are off-topic, as the results are rarely useful to anyone else and/or would be too long for this site.

When to use:
Use this whenever a question only asks to analyze ciphertext or when it asks to analyze a complex construction which you feel would require an unreasonable amount of work to do, e.g. if a new and not trivially broken symmetric cipher is suggested.

Relevant Meta Q&As:

Remarks: This is one of our existing custom close reasons.

$\endgroup$
9
  • 1
    $\begingroup$ This has my upvote (I can't vote on my own posts) $\endgroup$
    – SEJPM Mod
    Commented Jan 3, 2019 at 14:23
  • 1
    $\begingroup$ FWIW, my impression is that quite many of the questions currently closed using this reason are homework exercises and CTF problems. If those are getting their own close reason, we may want to consider tweaking the wording for this one, or possibly even dropping it from the list entirely. I haven't really seen that many "break my home-made cipher" questions here, and the few that we do get could still be closed with write-in reasons. (Of course, that might just be because this close reason has been really effective at discouraging them...) $\endgroup$
    – Ilmari Karonen
    Commented Jan 3, 2019 at 18:41
  • 1
    $\begingroup$ I think "reviewing full cryptographic designs" should be re-worded, given that there are some questions that are on-topic which could subjectively fit under such an ambiguous close reason. I think a question asking for us to analyze their homebrew crypto can get closed under too broad. What do you think? $\endgroup$
    – forest
    Commented Jan 4, 2019 at 2:09
  • $\begingroup$ No, I don't think we should drop this from the list. I personally don't mind if there are somewhat overlapping reasons (this is actually quite common, sometimes I find questions that tick all of them). This is the most chosen close reason, and there will be plenty questions left that need to be closed. I think "too broad" is too broad a close reason for these kind of questions. And yes, there is a grey area here: how complicated should a scheme get for it to be a review of a design? But I'm afraid that this grey area will persist whatever close reason we think of. $\endgroup$
    – Maarten Bodewes Mod
    Commented Jan 4, 2019 at 8:35
  • $\begingroup$ @MaartenBodewes Well that's why I think "too broad" would fit well, since any well-worded question with a specific problem to analyze is certainly on-topic, despite being disallowed by the current wording of the close reason. If anything, I think we should re-word it somehow so it isn't so ambiguous as to make a substantial portion of high-quality questions off-topic. Right now, the spirit of the law and the word of the law are at odds. $\endgroup$
    – forest
    Commented Jan 4, 2019 at 9:21
  • $\begingroup$ @MaartenBodewes The problem is that the spirit of the law and the word of the law do not distinguish "Does this scheme designed in this way provide security from this threat? Why or why not?" and "Hey guise I made my own cipher pls tell me if it's as secure as AES!!!". The reason I think that it would fit under too broad is because of this reasoning: "So you're actually asking about 100 questions in one.". $\endgroup$
    – forest
    Commented Jan 4, 2019 at 9:26
  • $\begingroup$ I'm opposed to share it under "too broad" because that doesn't provide enough information to the user. I personally am not waiting to get into discussions about this with users, which would be both detrimental to the other users and the person asking the question. "Too broad" is correct maybe, but it is way too generic and users will go into discussion about it. Something that we're currently avoiding most of the time. Avoiding the discussion is why this close vote exists in the first place. $\endgroup$
    – Maarten Bodewes Mod
    Commented Jan 4, 2019 at 9:41
  • $\begingroup$ @MaartenBodewes Can you think of better wording for the current close reason then? Perhaps something about question asking for "broad reviews of homebrew cryptographic schemes"? $\endgroup$
    – forest
    Commented Jan 4, 2019 at 10:19
  • 1
    $\begingroup$ I can certainly think about it (as can you) but I guess we need to do this in a separate Q/A, otherwise we don't know what is voted on. I don't want to make too many changes to answers that people already voted on. SEJPM, shall we make this a TODO after the voting has taken place? $\endgroup$
    – Maarten Bodewes Mod
    Commented Jan 4, 2019 at 10:28
7
votes
$\begingroup$

Homework and CTFs (alternative wording)

Questions asking for solutions to crypto puzzles or homework exercises are off-topic. While we do accept questions about homework problems, such questions must contain more than just a verbatim copy of the assignment, and should preferably ask for general solving techniques rather than just a solution to a specific puzzle or exercise.

When to use?

This is an alternative to Maarten's suggested close reason for homework and capture-the-flag questions. Its intended primary targets are questions consisting entirely of a verbatim copy of a homework assignment, exam question, CTF challenge or some other puzzle or exercise that the OP expects us to solve for them.

Rationale

This proposed close reason is an attempt to address what I feel are some shortcomings in the wording proposed by Maarten.

In particular, I feel that the emphasis on demonstrating effort and on asking about specific parts of an assignment in Maarten's proposed wording may in some ways be counterproductive. Sure, it achieves the goal of telling users not to just copy their entire homework verbatim here, but we can simply tell them so directly just as well.

Meanwhile, my experience from Stack Overflow, which currently has a similar close reason for debugging questions, is that telling users that they "must show some effort" often just leads them to bloat their questions with irrelevant details of all the failed attempts they've made and all the dead ends they've tried, making their already too narrowly scoped question even less generally useful. In the worst case, it can also lead to perfectly good and widely applicable questions getting closed just because the OP asked "How do I do <X>?" without also including a bunch of failed attempts to do <X> in order to "demonstrate effort".

IMO, the real problem with homework (and CTF) questions is that such questions tend to invite answers that merely spoon-feed a solution to the OP, without actually teaching them (or anyone else) much if anything of lasting value. Telling users that they need to demonstrate their efforts to solve the problem on their own first can serve to discourage the worst of such questions and to reduce their number, but it does very little to improve the quality and general usefulness of those remaining homework questions that do still get asked.

In particular, the close reason text is what gets shown to the OP after their question has been closed. As such, two of its most important goals should be 1) to explain to the OP what they should do if they want to improve their question and get it reopened, and 2) to also explain to them what mistakes they should avoid when asking questions in the future. I have tried to phrase the suggested wording above with this in mind.

Also, the part about CTFs in Maarten's version is kind of awkwardly tacked onto the end of the close reason text, and not phrased in a way that would make sense in the context of a post notice. I've tried to fix that in my phrasing above.

Relevant meta Q&A's

Same as for Maarten's proposal, including:

(It might be desirable to include a link from the close reason text to some meta post that more clearly describes our current policy on these kinds of questions, but I'm not entirely convinced that any of the linked threads above are particularly good choices for that. It could be better to just create a new FAQ-style meta Q&A specifically for this purpose.)

$\endgroup$
2
votes
$\begingroup$

Homework and Capture the Flag

Homework assignments are off-topic here. Questions on specific parts of an assignment are on topic as long as an indication is provided that an effort was made to perform the assignment. This reason can also be chosen to close Capture the Flag questions.

When to use:

Use this close reason when there is a clear indication that a homework assignment was dumped on the site without any effort to perform the assignment. If it is not clear that the question contains an assignment then first ask the author in a comment below the question. This close reason also covers capture the flag questions, when the a question is asked to provide the answer rather than to explain part of the problem.

Reference Meta Q&As:

Also tagged using the meta tag.

And for CtF questions:

Note that this close reason kind-of goes against the latest question here on meta. Currently the most upvoted answer to the latest question would possibly mean that the moderators would have to delete correct answers.

$\endgroup$
3
  • 4
    $\begingroup$ While I like the idea of including CTF questions under this close reason, the "This reason can also be..." sentence at the end feels kind of tacked-on and should probably be reworded. The tricky part with these close reasons is that the same text is shown both in the vote-to-close dialog and in the banner on the question itself after it has been closed, so the wording should be chosen so that it makes sense in both contexts. $\endgroup$
    – Ilmari Karonen
    Commented Jan 3, 2019 at 18:30
  • $\begingroup$ Also, my subjective impression is that homework and CTF problems make up a substantial (and IMO the most clearly problematic) subset of the questions currently closed as "requests to analyze ciphertext". If this close reason is added to the list, we might want to reconsider the ciphertext / scheme analysis reason and see if it might be reworded to reduce the overlap, or even removed from the list entirely. $\endgroup$
    – Ilmari Karonen
    Commented Jan 3, 2019 at 18:35
  • $\begingroup$ Ps. I've proposed an alternative wording for a close reason for homework and CTFs. I welcome any suggestions for further improvements. $\endgroup$
    – Ilmari Karonen
    Commented Jan 4, 2019 at 0:29
-2
votes
$\begingroup$

Reference Requests

Requests for literature, software or similar recommendations are off-topic here. For details, see: Do we want "literature recommendations" and similar "list/subjective questions"?

When to use:
Use this close reason whenever a question makes a request for a broad range of literature, e.g. asks for a recommendation for a book or when it asks for similar subjective recommendations such as "what software to use for ".

Reference Meta Q&As:

Remarks: This is one of our existing custom close reasons.

$\endgroup$
3
  • $\begingroup$ Could you give an example of valid reference requests? For example is this valid? $\endgroup$
    – kelalaka
    Commented Jan 3, 2019 at 16:35
  • 5
    $\begingroup$ @kelalaka: A prototypical on-topic reference-request would probably be something like "What was the first publication to introduce <something>?" or "Where can I find the official definition of <something>?". Note the definite articles. As for the question you linked to, I would say it's probably technically off-topic under current policy due to being open-ended. On the other hand, as long as it's doing no harm (i.e. as long as we're not flooded by such questions, and as long as it's not accumulating lots of spammy answers), I'm willing to give it the benefit of the doubt. $\endgroup$
    – Ilmari Karonen
    Commented Jan 3, 2019 at 18:18
  • 1
    $\begingroup$ I've asked in the other mods, and they came up with this link that in itself links to the physics site. They have their own policy on handling ref. requests. Some things are still off topic, of course. $\endgroup$
    – Maarten Bodewes Mod
    Commented Jan 14, 2019 at 23:53

You must log in to answer this question.

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