Do we want list questions?

Question

We have had a few questions recently that have been somewhat broad, essentially asking for a list. Two in particular:
What early home computers have more than one CPU, where both could be used by the programmer? and
What early computers had excellent BASIC (or other language) at bootup?
have between them racked up over 12,000 views so far and have brought many new users to the site.

However, they have also gathered close votes for being too broad.

Up to now, I have deliberately abstained from voting to close as I believe it is for the community, not a moderator, to decide whether we want such questions.

List questions do not make a good fit with the StackExchange model of question and specific answer. Indeed they are explicitly off-topic at some sites, although other sites permit them.

We are a young site and have let the small number of list questions ride so far. They have brought new members to the site and helped it keep growing. We need to consider whether we continue to accept them or become more rigorous at closing as too broad.

Other site's meta discussions include:
Are list questions off topic?
https://scifi.meta.stackexchange.com/q/2638/30512
https://movies.meta.stackexchange.com/q/1060/18237
https://english.meta.stackexchange.com/q/2016/64632
There are others if you wish to look for them.

I seek the views of the community on where we should draw the line on such questions. Please give your views.

Answer 1

My opinion is no, I do not think questions which attract "lists" of answers are constructive and should be considered off-topic.

I will say this: If the community believes these questions are on-topic, then the posts should be locked and a Community Wiki model should be used. See The Definitive C Book Guide and List and The Definitive C++ Book Guide and List for reference.

Answer 2

We should be wary of the argument that a question should stay because it attracted traffic.

Everyone loves link or list heavy reads; there are sites dedicated to the format. But this does not mean it is good for Retro.

List answers are never truly finished. They have a long tail, so whether we lock or not, they will not completely satisfy the question.

List questions rely on multiple answers, with no best that bubbles to the top. Even votes reflect "me too" more than "well said". And answers will often be similar with only a few differences. And even if the me too votes are in comments, we run the risk of actual answers being hidden in comments more than usual. Even in the aggregate, the answers will present duplicate and incomplete information. The is the opposite of what you want from a list of things related to a subject.

List answers basically duplicate what is already out there on specialized wiki and so on. We can never hope to be better than a curated wiki with a QA format. Minimal copy and paste editorializing borrowed from other sites is not good for Retro for many reasons.

We want to collect a body of curated Q&A based on actual knowledge, experience, and research. List base questions give us barely two of those at best, and is usually heavy on the latter.

Ideally, what we should strive for is a body of curated info that those other link and list based sites can use for their content.

Answer 3

TLDR: "Answer is a list" alone is a poor proxy for "the question is bad"; in some cases questions whose answers are lists may produce a new summary of information that did not exist before on the Internet and is unlikely to arise and/or be made widely available outside of a community like this one. Use "answer is a list" as a prompt for further thought, not to automatically generate a close vote.

Disclosure: I am the author of a recent list question, What LISP compilers and interpreters were available for 8-bit machines?, that did seem to be acceptable to the community; after 212 views (including at least two by diamond mods, apparently) it had attracted only one close vote. That voter asked here in meta "Are list questions OK again?", which I am answering here because that question seems to be a duplicate just to stimulate discusssion (though I'm not convinced that's a totally bad thing).

What's bad about "list" questions?

I am in general agreement with the StackExchange question criteria. I also agree that certain types of questions involving lists are not appropriate here, often because there are (or may be) better forums for them:

• Lists based on opinion that cannot be eventually reconciled to fact ("What was the best 8-bit BASIC?"). These almost invariably involve a too-broad or opinion-based question, so it's not clear that the problem here is really the request for a list, per se.
• Lists that are very large and don't rely much or at all on in-depth experience ("What were the games made for the Atari 2600?")?
• Possibly, lists that can easily come from a single or small number of easily-found sources ("What are the addressing modes of the 6809 CPU?") basically duplicate information already easily available. (But I could be swayed on this one.)

I don't think that the argument that one answer can't be chosen as the best holds water if the community is willing to make and support a community wiki answer, as happened with my question above; clearly that answer will end up being the best unless there's not even one person willing to integrate into it appropriate material from other answers.

The multiple-answer format is still useful in this "one community wiki answer" case, though, as demonstrated by the ZIL answer to the question above. It's not clear that ZIL is really a LISP (or wasn't at the time, I am now convinced it is) and whether or not it fit the spirit of the criteria for the question. Indeed, it prompted further criteria being added to the question and I am still consdering, because of that answer, how to further narrow the question.

Can "list" questions be acceptable?

"The answer is a list" is certainly a fairly easy criterion for categorizing a question, and everybody likes easy criteria because they can reduce the amount of thinking needing to be done. But I don't think it equates directly enough to "this is a bad question for Retrocomputing" that we can use it without further thought. (Many answers to similar questions in other exchanges, including the links posted in this question, say essentially the same thing.)

If we look at the one answer in the StackExchange meta (19 votes, accepted) every criterion for bad questions can be avoided by list-based questions; i.e., the problems it raises are all about something other than just the answer being a list:

• "[E]very answer is equally valid": not the case when using a community wiki answer.
• "answer is provided along with the question": not always the case with list questions. (Also, that example can't have a single answer.)
• "[T]here is no actual problem to be solved": particularly in Retrocomputing, going back and digging out information spread out over many historical documents and memories can be a real but also soluble (at least to some large degree) problem.
• "[W]e are being asked an open-ended, hypothetical question," "[I]t is a rant disguised as a question": clearly not the case for some kinds of lists.

I suggest that the criterion "the answer is a list" should not be used alone to decide whether the question warrants a close vote, but instead as a heuristic to prompt asking, "is this really a quality question, and what makes it so or not?" Look at the other criteria for bad questions (including those above), and some critiera for good questions:

• Is the question "finite and well scoped"? An imperfect proxy for this is, "is the resulting list likely to be relatively short?"
• Does it "have a constructive, fair, and impartial tone", "invite sharing experiences over opinions", can it be "be backed up with facts and references", and is it "more than just mindless social fun"?
• Does the question require knowledge likely to be distributed piecemeal across a broad community, as opposed to having authoratative sources (i.e., would the answer become one of the most authoratative sources for this information).
• Can the question be further developed to improve and/or narrow it, and is the poster doing so? (For this, comments on what's wrong with the question, beyond "answer is a list," or how to improve it would be appropriate.)

Answer 4

I agree with what jdv wrote above, and I'd like to add more:

The problem with a list is that it'll never be complete; there will always be new elements found about some obscure device / system, like some Brazilian clone of some device, etc.

This means that the list will require maintenance; also, there will be a bunch of edge cases where devices will kinda fall in a category, but not always depending on how we look at it, etc.

And if we edit the list, how do we make it a current post again for interested people to even know it was updated?

I believe Wikipedia is a much better place for these lists as they do not really bring knowledge nor understanding.

Another problem with a list question would be if someone is asking a question like: "What were the games made for the Atari 2600?". Obviously, the list would be a long duplicate of already existing, and easy to find, lists, but we'd have to discuss why we don't want to include this list, but we included the list of computers with an excellent basic for example.

So, overall, I believe, like most of you from what I can read, that these lists do not have a place here.

Answer 5

As the inquirer of both example questions, perhaps I should say something. Both questions were initially worded loosely, prompting responses outside the scope of what I meant to ask. Folks suggested that the questions be edited for narrower scope, and they were. Unsurprisingly, answers kept coming in beyond what I wanted to know. That did not bother me; the discussion itself was interesting.

See how I snuck the key word in? -- discussion -- For good or ill, most internet users expect to be able to discuss a topic on what they view as a forum. I understand that a design purpose of the SEs is to provide cogent answers to topical questions, and avoid unnecessary discussion. If that is what the community wants, I accept it. However, gird yourselves for the thankless, forever-task of telling new subscribers your question is wrong. No matter how friendly that is worded, the new guy will often feel unwelcome.

Answer 6

This question has been up now and featured for seven days. Thanks to everyone who expressed their views, either in the well reasoned answers or by voting. As I write this there are no downvotes at all on the question or any of the answers.

The consensus is that we don't want list questions.

Don't forget that discussion and requests for lists are acceptable in our chat room.