Should we allow "summary of information" posts?

Question

The question List of useful external tools has triggered a discussion about the kinds of questions that should be allowed on the site. Specifically, how do we handle questions that are intended to just accumulate a list of tool/editors/commands/packages/whatever? I think we can all agree that some people would find them useful, but is it really a good idea to have these giant lists? It seems to me that it would be better to keep questions tightly focused. Anything that would be a good answer to one of the list questions will probably also make an appearance on a more focused question where it is more relevant.

Anyway, what do you all think?

Answer 1

We could copy the [big-list] tag from Math Overflow; my summary from math.SE:

1. Questions that don't look like they will have a single, correct answer should be tagged either [soft-question] if that's because the question isn't precise enough, or [big-list] if that is because we expect the question to invite many different, incomparable answers.
2. These questions should be made Community Wiki.

Answer 2

I'm fine with "lists of somethings" as long as they are CW and the "somethings" is somewhat specific: e.g. graphic packages, or sites with BibTeX citations. However asking for a list of external tools? Anything could fit in there!

In some list-of questions it might also be useful to revise the answers from time to time and include the seemingly best solutions into the question itself, so that people googling can quickly find the information they're looking for.

Answer 3

I agree with the sentiment that the SE software is not a great place for these lists. In addition to the other reasons given, I'd add that the lists so compiled are never complete. They attract a lot of answers when they first appear, but then very few people go through and answer them later on, meaning also that they quickly get out of date. The voting system also doesn't work with these questions.

However, given that these are inevitable, I would like to suggest a Third Way (I am British, after all). I think that "Let's have a big list" can easily be turned into an acceptable question, along the lines of:

Can someone give me a link to a list of graphics packages for LaTeX, preferably one that lists the main features (and disadvantages) of each? If not, is there an obvious place where could I start one?

(tag suggestion: reference-request)

Then if the answer to the first is "No", the person can start a list and include the link in the question in a later edit. Inevitably, some answers will contain actual graphics packages, but then the questioner can copy those across to the list. Note that the questioner - and only the questioner - gets notified when a new answer is posted so it's easy to do this.

Answer 4

I don't see what is wrong with questions which are attempting to get a list of tools/packages, as long as they're classified as community wiki. I have learned a fair amount from such questions being posted on StackOverflow and gauging what people think based on the up-votes of the answers.

I recently asked what graphics package people use in an attempt to be made aware of anything out there that is better than TikZ, or that might be good for particular applications, and it was commented upon negatively as "I think we should discourage very open ended questions like this one" and it had a few votes to close it.

I'd personally like to see questions in the community wiki like 'What latex resources do you find indispensible?' or 'What packages could you not live without?' which both have counterparts on StackOverflow. Of course, the community might decide they don't like these types of questions and they will simply not get up-voted.

Answer 5

Over at Chiphacker.com, we have a number of list questions. They're all community wiki, and they've been useful. Meta SOFU sites also have lists frequently. I agree that the list has to be manageable (Less than, say, 50 answers expected), and "useful external tools" certainly does not fit that description. However, there are many good questions that would be well served by a community wiki list.

See the meta.SO questions what are community wiki posts and when to mark the community wiki checkbox on a new question for the official stance on CW posts - The MO community takes a slightly different approach here, which is OK, but we need to decide for this site.

Relevant to this question are the following excerpts (But you should read the whole thing):

Community Wiki posts help enhance the wiki aspect of the site.

The site is supposed to have a wiki type aspect - Every question does not have a single correct answer!

They make the post easier to edit and maintain by a wider group of users

CW posts can require maintenance, as a list would.

The FAQ says, "this is a place for questions that can be answered", but it doesn't forbid questions which have more than one answer.

List questions inherently have more than one answer.

Personally, I think that we should allow wiki-type (aka "summary of information") posts, because "wiki" is an element of StackExchange:
Venn diagram w/ blog, wiki, forum, reddit http://sstatic.net/stackoverflow/img/venn-diagram.png
* = StackExchange

Answer 6

I'm in general alignment with much of the spirit of the question. I do, however, bristle a bit at "should we allow". I vote for "allow, but use the voting mechanism, and comments to strongly discourage."

If not allowing would involve deleting, I'm strongly opposed. Since there is a clear constituency for such "list all X" questions, leaving some such questions on the site with downvotes and rationales will serve to convey to (at least some of the) newcomers who want such questions that that's not quite how the site works.

Edit: In light of @ShreevatsaR's useful comment, I change to

I vote for "allow, but use the voting mechanism, the ability to close, and comments to strongly discourage."

Answer 7

I find it a really bad sign that we didn't even make it a day into the beta before the [subjective] tag was created. We should try to decide earlier rather than later what the stance should be on compilation questions like this; they were popular in SO's early days but generally aren't tolerated now, which leads to:

• People closing questions that are years old
• Other people complaining that those questions should be open because they're years old
• People posting more questions along the same lines and citing the earlier questions as justification

I'm personally of the opinion that they absolutely shouldn't exist; they're not what the SE engine was designed for. They don't answer a specific question, they aren't helpful for people searching for answers to a given problem, and they tend to vastly skew metrics (viewing the top-voted SO posts (even in a given tag) is useless, because they're all poll questions). I thoroughly enjoy these questions, just not on SE sites; they belong on sites like Reddit that are formatted for that sort of thing, and people tend to have trouble differentiating "this question is good" and "this question is good on this site"

I probably sound a bit intense, but I'd very much like to avoid another issue like SO's Hidden Features insanity; for a while we couldn't go a day without those questions coming up