# Should we accept edits by non-authors to put Wikipedia links on phrases?

We have had a number of edits recently that just put Wikipedia links on phrases, in posts that the editor was not the author of. Is this a desirable practice?

For example, in https://crypto.stackexchange.com/review/suggested-edits/42748, the question was about oblivious transfer in electronic voting protocols.

• The Wikipedia article on oblivious transfer does not mention electronic voting protocols, but the question is about how oblivious transfer is used in them, suggesting the author had a reference in mind that does connect the two.

• The Wikipedia article on e-voting does not discuss cryptography protocols; it is about the broader sociopolitical context of electronic voting systems, and only briefly mentions cryptography in passing.

Often, a reader who is asking about terminology has found it used in a particular source—such as a textbook or conference paper or journal article, not Wikipedia.

• Surely eminent domain for this is meta.stackexchange.com. Issues around 'linking' people's answers cannot be confined solely to this site... – Paul Uszak Sep 22 '19 at 12:37

Sure, if the links are relevant. Stack Exchange is a collaborative platform. Editing other people's posts to improve them is encouraged.

Quoting the official guidelines on editing:

When should I edit posts?

Any time you see a post that needs improvement and are inclined to suggest an edit, you are welcome to do so. (…) Common reasons for edits include: (…)

If adding a link to Wikipedia improves the post, accept the edit. If the link is wrong or irrelevant, reject the edit. Generally, linking to Wikipedia or some other reference is a good idea in two cases:

• For explanation, when mentioning concepts that are likely to be obvious to some readers, but mysterious to others. (Concepts that most readers are unlikely to understand should be explained in the post, and concepts that most readers would find obvious don't need any explanation.)
• For further information, when mentioning a concept that is peripheral to the post, which doesn't warrant further discussion inside the post but where a curious reader might want to know more.

If the Wikipedia article is relevant but unhelpful or misleading, that's a reason to lead towards rejection. However, the best outcome would be to go and improve the article!

Do not reject a suggested edit or roll back an edit on principle because “it modifies the post”. Edits modify the post by definition. Reviewing suggested edits is a judgement call, that's why we have humans doing it and not computers. But as a human, you're supposed to apply common sense and follow shared values. Stack Exchange has very few firm rules but many guidelines, and for editing in particular there are official guidelines. You should go against the guidelines if a specific case would go against our goal of “build[ing] a library of detailed answers to every question about cryptography and cryptanalysis”. But if you're systematically going against the guidelines, you're doing it wrong.

Different people can have different thresholds for when a suggested edit is acceptable and when it does too much, and that's fine. But if you systematically apply your own rules that go against the guidelines, that's not fine. It disturbs me to see moderators here who completely ignore the official guidelines and make recommendations that go completely against them, given that the job of moderators includes enforcing these rules.

• It's one thing to provide a bibliographic reference to a particular paper that the author of a post was referring to. It's quite another to blindly put a link to Wikipedia on some phrase without independently verifying that the Wikipedia article is a coherent reference on the subject, and provides the definition and explanatory context that the author meant—which can really be assessed only when the editor and reviewer are particularly familiar with the term in question as it is used and confident in what the author was asking. – Squeamish Ossifrage Sep 23 '19 at 20:13
• @SqueamishOssifrage In point of fact, editing is a dialectical process between writer and editor. The discussion is best done face-to-face, and it usually takes at least two iterations. As far as this website goes, we do our best. Your question and response have helped raise the bar of the editing work on this website. – Patriot Sep 25 '19 at 13:19
• @Gilles When you see something good going on among a group of people, that is a reflection of the leadership. I hope this website continues to be an oasis of civility, which is rare to find these days, perhaps even extraordinary. – Patriot Sep 25 '19 at 13:33
• @MaartenBodewes I still think you're twisting my words. But ok, I've edited again and avoided “listing the consequences of continuous breaking of the rules”. I do keep my criticism of the position of some moderators on this site. Objecting to my criticizing moderators is not cricket. – Gilles 'SO- stop being evil' Sep 25 '19 at 22:09
• I don't know what you mean with "cricket" but yes, we were probably out of bounds w.r.t. the rules, and the constructive part of your answer was never questioned. I've also removed that the last comment; let's leave it at this. – Maarten Bodewes Sep 25 '19 at 22:16

No. This is not desirable because it does not necessarily reflect the author's intent, and Wikipedia is a notoriously bad reference on cryptography so it is likely to be misleading at best and change the author's meaning at worst. If the author wants to cite the definition that is currently on Wikipedia, they can do so themselves.

• The questioner is right because Wikipedia is often unreliable. – Patriot Sep 17 '19 at 2:14
• Normally I'd disagree with this, but now that I'm learning more about cryptography, I'm starting to realize just how bad Wikipedia is on that subject... – forest Sep 17 '19 at 5:30
• @forest It's a lot better than most resources on the web, including Stack Overflow. – Gilles 'SO- stop being evil' Sep 17 '19 at 21:02
• @Gilles Asking a professional psychic for cryptography advice is better than using Stack Overflow. – forest Sep 18 '19 at 8:57

Yes it's a desirable practice and we should allow it

## Why?

• Many times the questions don't define these used terms. For the people who see the term first in their life, a direct link could be helpful.
• I created this answer so that users can vote on it as they see fit. I could not come up with a reason to support this answer, but if anyone else thinks of one, feel free to suggest it. – Ella Rose Sep 15 '19 at 19:52
• Many times the questions don't define these used terms. For the people who see the term first in their life, a direct link could be helpful. – kelalaka Sep 15 '19 at 22:40
• @kelalaka Sure, but how do we define when it is ok to insert a link? Should the first occurrence of any given term on every question/answer be a link? That clearly seems like it goes too far to me. But if not every Q/A, then which ones, and why? – Ella Rose Sep 15 '19 at 22:45
• @kelalaka In what circumstances is it appropriate for someone other than the author to provide a reference, which the author may not have even read, for what the author meant? – Squeamish Ossifrage Sep 15 '19 at 22:47
• @SqueamishOssifrage well, maybe a mod can decide that this question has a broad audience so inserting a link to Wikipedia is helpful. This time we put the burden on mods to check that the Wikipedia article is good. IMHO, during the active time of the question is should be handled by anybody interested in the question, not later so that the OP can argue, too. – kelalaka Sep 15 '19 at 22:54
• @kelalaka In what circumstances should that be done by editing the question rather than by commenting on the question and letting the author do the edit if appropriate? – Squeamish Ossifrage Sep 15 '19 at 22:55
• @SqueamishOssifrage IMHO, we shouldn't and anybody can copy and paste to search. If the OP, as the recent RSA question, gave the link is ok, if not, since the context says wiki we can. that's it. – kelalaka Sep 15 '19 at 23:09

No, this is not desirable.

## Why not?

But the reason is not because of the quality of content on wikipedia, it's that edits should be restricted to typographical fixes

There is an exception: removal of thank-yous, signatures, taglines, etc is encouraged.

An edit that modifies the content/subject matter of a question or answer is in general inappropriate.

Editing a link into a question can imply to future readers that the question asker is familiar with the linked content, and that their question is the result of their reading of that content.

If an answer to a particular question should really contain of links to certain resources, then post those links as your own answer (along with an explanation/summary) rather than editing them into someone else's answer.

• I think saying "typographical fixes" (which just means fixing miss-spellings and punctuation?) is a bit too restricting here, as we also commonly improve formatting and style / readability of posts. – SEJPM Sep 15 '19 at 22:10
• Formatting style is part of typography. Typography is not just the study of typos. For example, converting between typewriter math and $\mathrm\TeX$ and fixing markdown is typographical. – Squeamish Ossifrage Sep 15 '19 at 22:19
• “edits should be restricted to typographical fixes” No, this is wrong. There are official guidelines about editing, and they include “add[ing] related resources or hyperlinks”. – Gilles 'SO- stop being evil' Sep 17 '19 at 20:46