12
$\begingroup$

Many questions ask for 'What are some examples of ...?' (Including the questions). But I would expect 'What are some examples of <thing>?' with no little/elaboration to be closed as lacking details but others are well received.

What should be the quality standards and guidelines for asking such questions?

$\endgroup$

2 Answers 2

11
$\begingroup$

"What are some examples of ..." is not a good question format, and should just be left off entirely. A number of questions with different phrasing also fall into that trap, but not all of them.

A good question would identify a real problem with constraints, not just solicit an undistinguished list of examples, and would expect analysis of the models presented as applied to that area. Probably this would amount to multiple questions encompassed within each of those "examples" questions, but each tailored to a different domain or goal.

Many of the answers to these questions to date are just statements of what some language does and don't provide substantive insight. With an actual design scenario in play, they could give analysis on the actual trade-offs made by that approach. They would still have many possible answers, but the answers would be set up to succeed in a way that a bare "what are some ..." doesn't give them.

As-is, they're easy to pump out answers to, but very hard to give a good answer to.

$\endgroup$
1
  • $\begingroup$ "As-is, they're easy to pump out answers to, but very hard to give a good answer to." - and I personally have a feeling that a lot of the "good answers" will fall back to some core principles that will just end up getting repeated a lot. Hence my attempt to refactor some of the effort I put into this out to this. $\endgroup$
    – starball
    Commented May 22, 2023 at 9:26
7
$\begingroup$

I don't know a lot about PL design, but I'm puzzled (concerned?) to see so many of these questions popping up so quickly.

These would probably be closed on Stack Overflow quite quickly for being too broad / needing focus. What's annoying about these is that it's so easy to fire off a short answer about how one language does something without getting into anything deeper or doing any actual comparison. I'm really not sure if that's what we want. Ideally (I think) we'd get a few answers with a lot of depth and breadth to them instead of a bunch of interspersed answers with fragments of information.

It's not that this isn't possible to achieve. I've seen it done really well. It's just that it takes a really wide and solid coverage of domain knowledge... And when actually done well ends up being quite lengthy. Ex.

And we'll need to be very careful if we get questions about comparing languages. I assume we'll close such questions as too broad (unless it's really a comparison of specific language-feature designs and makes a significant effort to be well-scoped). I defer to this ISO C++ wiki section on the subject of programming-language comparisons.


Note that there are Stack Exchange sites that allow Q&A in this style, but those are pretty intentional choices on their parts, and I'm pretty sure they have norms around how those Q&A should work. For example, Code Golf (which is a pretty exceptional case if you ask me), and Math's big-list tag ("Ask only when the topic is compelling, and please do not use this as the only tag for a question.").


Here's what I think might be useful to expect of such Q&A:

  • For questions that lend towards answers that are big lists (where there are a lot of easily found examples and where the question is not looking for a lot of depth): What I think we should expect if we want these types of question to stay is that the answers should get better over time more complete and more thorough (which is probably the opposite of what's likely to happen if we don't try to control the culture around this). Community Wiki answers sound like a good fit here.

  • For questions that may lend towards answers per example (where there are fewer examples, and where the question is looking for relatively more depth), I think it's fine as long as depth in answers is a goal, and the question is well-defined.

In any case, the questions should make sure they do a very good job of specifying their constraints.

Proposal: We can do a Meta Sandbox Q&A for these

It might be useful to have a sandbox area for these questions (Ex. on Puzzling and on Code Golf) to vet and improve them before they get posted to the main site. In fact, I think I'd go as far as to strongly suggest this.

$\endgroup$
3
  • 3
    $\begingroup$ Note that this site is inherently more subjective and our questions are inherently more broad than say SO. We are more comparable to UX.SE $\endgroup$
    – Seggan
    Commented May 19, 2023 at 13:01
  • $\begingroup$ I like this sandbox idea if it results in higer quality questions but as a counterpoint why is there not such an area for regular SO? (I like the idea that those who are will to put more effort into a question can be guided to avoid falling into traps and getting a question closed. Repeated edits are regularly not enough to get a question re-opened) $\endgroup$ Commented May 21, 2023 at 21:57
  • 1
    $\begingroup$ "why is there not such an area for regular SO?" - how much do you follow the news on MSO? meta.stackoverflow.com/… $\endgroup$
    – starball
    Commented May 21, 2023 at 21:59

You must log in to answer this question.

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