How should I change my recent "Counterexamples in quantum information" post, if at all?

Question

Two days ago I created the overall well-received post Is there a comprehensive list of counterexamples in quantum information?. Yesterday I got a comment---which in ratio got a fair amount of upvotes as well so a handful of users seem to agree---that quote "the answers should give the information here, not link to an external resource". For context---as I also wrote in the post itself---I created this list around a month ago and because I got more active in this community since then I felt like it could be a fitting addition to the site.

That being said, I do agree with the quoted comment and I've been thinking about how I can improve the post. As such there are a couple of points I'd like to hear your opinion on, if you don't mind:

1. I've come to the conclusion that, probably, the best course of action is to write out the list in the answer (and keep the google doc linked as a backup should the lists ever be out of sync). I wouldn't mind doing that and things should also stay within the 30,000 character limit for answers; the google doc is well below 20,000 characters for now. Such "overview" style posts are, after all, well established on SE; the one that directly comes to my mind and that I've used a fair bit in the past is Overview of basic results about images and preimages. On the other hand the latter only links to other math.SE questions whereas my list also links to a fair amount of scientific papers; more on this below.

Should the consensus be that my list as a whole doesn't fit the scope of qc.SE I'll of course not oppose to taking the post down altogether. However if porting/incorporating the list into my answer (as plain text + links) is the preferred solution there are some further points I'd like to raise.

2. The majority of the list is in line with topics frequently discussed on qc.SE, such as channels, states, entropies, etc. However, my list features, e.g., a section on quantum thermodynamics which in my opinion doesn't really fit thematically. Should I limit the answer to those topics & questions relevant to this community?
3. Some posts feature additional context such as, e.g., historical remarks or a slightly weaker statement that, unlike the original statement, is true. Should I cut this extra information from the ported answer and keep it minimalistic, i.e. only state the question + link to a counterexample?
4. While many examples feature links to math./phys./qc.SE answers there are also enough counterexamples which link to scientific articles. Ideally I would link to an SE answer as well as to as article (DOI + arXiv), but often there are only "external" (=non-SE) resources I can link to. What should I do with those points?
5. Finally, a total of 8 counterexamples link to notes I wrote myself that are currently stored in my google drive. Again, I agree that this is far from ideal for the ported list. One idea I had was that I could turn these notes into their own qc.SE Q&A post but I fear that this could be seen as spam or upvote farming. The alternative would be to leave them out entirely or to limit this to only 1-2 notes "of greatest common interest" which are turned in to their own post, and the remaining examples stay in the google doc.

Edit (2024/04/28): Thank you for your input everybody! Today I changed my original qc.SE post according to the suggestions you made below.

Answer 1

A few thoughts to complement the other answers:

• As I said elsewhere, I certainly agree that the project would be much better as a (series of) post(s) directly on the site. And this site is in my opinion perfectly suited to host this sort of thing.

  How exactly to go about it, I think is mostly up to you. But in my opinion, the site works ideally when information is built with a "bottom-up" approach, with each post, possibly containing a simple counterexample (or rather, asking about counterexamples on a specific topic), living a life of its own.

  The matter of having a single post containing a "big list" of links or other things is a bit more delicate. Some sites of the network don't like these kinds of posts as they go a bit against the "question&answer" model, among other things. See for example this meta.physics.SE discussion about it. I don't think this issue ever arose on this site though. There is the matter of such big list posts possibly attracting a disproportionate audience and thus upvotes, which might be considered a bit unfair (though I'm personally not that convinced about it). This can be solved by making the post a community wiki though, as you'll notice is the case for the post on math.SE you linked.

• Regarding the point you brought up about making lots of such posts possibly be seen as "rep farming" or "spam": I really wouldn't worry about it. If you make many useful posts that are upvoted by many people, you're greatly contributing to the site, so why shouldn't you be equally rewarded? Of course you'd need to somewhat pace yourself (if you make lots of posts per day it becomes a bit much), but if you even were to make, say, 1 or 2 such posts per day, that certainly wouldn't be an issue. Just make sure to put in the minimal work required to have each post stand on its own, independently from the larger project.

• I'm not saying this to discourage your from posting this here (I think it'd be great), but on a different note: I've seen similar projects maintained on a GitHub repo, which I think is a much better choice than a google docs document.

• Why wouldn't quantum thermodynamics fit thematically? We don't get many posts about it (if any), sure, but I see it as absolutely on-topic on this site.

• In conclusion, I see two possible choices here:

  1. Make a series of posts like Norbert's one on physics. Each post with a title along the lines of "Canonical counterexamples for X", or "What are canonical counterexamples in Y", or something like that. Essentially one post per section in your project. If there's not too many of these, you can simply cross-link them in each post. Otherwise, you could make another post with a list of links to each of these. With this approach, if each post is kept sufficiently focused and specific, I can see such posts remaining non-community-wiki (though I'd still have the summary one as community wiki, as it'd only be a list of links to other posts).

     Another advantage of this approach is that it makes it easier for other people to contribute as well: adding an answer to a post about counterexamples in a specific topic is much less daunting than editing a very large list containing all sorts of things.

  2. Make a single post about "Canonical counterexamples in quantum information", with each answer corresponding to one of the sections in your document. Such a "big-list" question would I think at least warrant being a community wiki, as it fits a bit less in the Q&A model the site is pushing for.

     On this regard, I'd also note that for community wiki posts author attribution is somewhat less obvious. It's still there, but it is by design not highlighted as much. See again the math.SE post you linked to see what I mean. Also, community wikis are designed to make it easier for other people to contribute and edit, and reputation is not affected by votes. I mention this only because this would probably represent a lot of work on your end, so you might want more attribution for it =).

Answer 2

I remember now that this reminded me of the mathoverflow post Common false beliefs in mathematics. You're asking roughly the same question for QI. The linked post has around 300 answers, in my opinion you can do the same here.

Answer 3

My opinion:

1. I don't think it is a problem if it gets too big. Another option is to put one counterexample per answer.
2. I think you should indeed limit the counterexamples to those that fit within the topic here.
3. I can't give a generic answer.
4. References are not only allowed but encouraged. But yes, only permanent resources like DOI or arXiv, we don't want to suffer from link rot.
5. I wouldn't worry about that, the community can always downvote you if they think you're spamming.

Answer 4

I've come to the conclusion that, probably, the best course of action is to write out the list in the answer (and keep the google doc linked as a backup should the lists ever be out of sync). I wouldn't mind doing that and things should also stay within the 30,000 character limit for answers; the google doc is well below 20,000 characters for now. Such "overview" style posts are, after all, well established on SE; the one that directly comes to my mind and that I've used a fair bit in the past is Overview of basic results about images and preimages. On the other hand the latter only links to other math.SE questions whereas my list also links to a fair amount of scientific papers; more on this below.

The doc is already organized in sections/subsections. Putting one section per answer should fit within the character limit. It also has the nice side-effect to be organized according to upvotes, so probably that the most useful counterexamples (from a frequency point of view) would come on top.

The majority of the list is in line with topics frequently discussed on qc.SE, such as channels, states, entropies, etc. However, my list features, e.g., a section on quantum thermodynamics which in my opinion doesn't really fit thematically. Should I limit the answer to those topics & questions relevant to this community?

If you're unsure, put it in. You're never going to be absolutely sure that no one in the field wouldn't be able to use these counterexamples, and this allows to find the answer from the search bar. And worst case scenario, most people don't find this section useful and so they don't upvote it.

Some posts feature additional context such as, e.g., historical remarks or a slightly weaker statement that, unlike the original statement, is true. Should I cut this extra information from the ported answer and keep it minimalistic, i.e. only state the question + link to a counterexample?

If this additional context brings strcitly more information than the link you provide, then sure, go on. As long as it helps the person using the counterexample, it's a good thing.

While many examples feature links to math./phys./qc.SE answers there are also enough counterexamples which link to scientific articles. Ideally I would link to an SE answer as well as to as article (DOI + arXiv), but often there are only "external" (=non-SE) resources I can link to. What should I do with those points?

The problem with non-SE links is that it will be way harder to maintain. I fear that as time goes on, the post will have more and more dead links without you knowing it. To the best of your ability, the ideal scenario is the one where you can directly put the counterexample within the post (just like Norbert Schuch did), and then potentially also add a reference to it if necessary. If it is obvious that the counterexample is, well, a counterexample, I don't think you need to add a reference.

Finally, a total of 8 counterexamples link to notes I wrote myself that are currently stored in my google drive. Again, I agree that this is far from ideal for the ported list. One idea I had was that I could turn these notes into their own qc.SE Q&A post but I fear that this could be seen as spam or upvote farming. The alternative would be to leave them out entirely or to limit this to only 1-2 notes "of greatest common interest" which are turned in to their own post, and the remaining examples stay in the google doc.

SE does encourage you to share your knowledge in a Q&A style. As long as what you're posting is on topic, it'll be fine.