I have just failed a review audit and was told that this question did not have any problem. To my understanding of the SO asking rules, that question is not actually on an implementation problem, but about a call for discussion about a comment on another question, deserving to me a "too broad" or a "opinion based" closing vote.

Would you mind explain in what that question is legit on SO?

If not, maybe that the "I understand" button of the review audit could be augmented with a "I do not agree"? I guess that it could add some more unwanted discussion...

  • 2
    Blind application of "rules" are invariably the problem and what reviews audit for. Looking at the tags in your profile, you don't have a single one for [haskell] or [clojure]. Are you sure you know enough about those programming languages to tell that the question is "too broad"? The answerer certainly didn't have a problem with it. Best to skip questions about a subject that you know nothing about. Mar 12, 2014 at 13:55
  • 2
    It looks off-topic to me. I expect that the automated review system used it as an audit because it has upvotes and so looks like a good question. Mar 12, 2014 at 13:55
  • 2
    @UphillLuge It's hard to convey tone in text, but your comment comes across as a little harsh. The answers to the question do look like they're matters of opinion "the obvious one is that … next is … typeclasses also …" and "the most fundamental difference is that …". Just because the question is a good question and well written doesn't mean that it's on topic for Stack Overflow. "What are the reasons that protocols and multimethods in Clojure are less powerful for polymorphism than typeclasses in Haskell?" does seem somewhat open ended and broad. I don't think one needs extensive… Mar 12, 2014 at 13:59
  • …experience in haskell and clojure to recognize that. Mar 12, 2014 at 13:59
  • 3
    Sometimes questions that aren't really good review audits get picked. You can cast a close vote on it to ensure that it won't be used as an audit again. Mar 12, 2014 at 14:03
  • @UphillLuge: If I do not use these languages, I can anyway understand when the formulation of a question is open ended or opinion appealing (like with the usage of the "claim" word). And when I am unsure I just skip a review. Moreover, I was filtering on [c] and [asm] tags when I was proposed that review that was falsely tagged [asm] (i checked that the tag was not removed), that makes another question on the review audit. (I do not contest the usefulness of the review audit to test reviewers BTW)
    – Seki
    Mar 12, 2014 at 14:06
  • 2
    @Seki I'm surprised to hear about the fake tag issue, but apparently that's real: Cheated audit with faked tags? Audits bug in the filtered review queue. This answer to the latter explicitly explains that the audit system adds fake tags: "we aren't querying for audits with filters enabled, we just fake the audit post, so that it looks like it matches your filters. While reviewing the faked posts appears like it matches your filters" Mar 12, 2014 at 14:20
  • 1
    @gnat: Thanks for pointing out that former question in the same spirit.
    – Seki
    Mar 12, 2014 at 15:03

1 Answer 1


It seems to me the question is clear and well within the rules of SO. It asks What are the reasons that protocols and multimethods in Clojure are less powerful for polymorphism than typeclasses in Haskell?.

Even if the claim is false, each have a specific implementation that will make one better or not at the claim, polymorphism.

Providing the cons and pros of each for polymorphism is within the rules and one might, or not, be better than the other.

It does look like the question could promote a discussion style answers but it can, and sould, promote explaining the differences between both in proper "SO manner" that answer talks about how they work, are implemented and why one or the other is better for polymorphism.

If you look at the accepted answer, it did that well enough.

  • 1
    "What are the reasons that …" sounds like a good way to have an open ended question. Both languages are Turing-complete, so there's one sense in which neither is less powerful than the other. The pragmatic ways in which one is less powerful is likely to be provided by examples and explanations, some of which could clearly be wrong, but no unique one of which is likely to be the right answer. It's a good, well written question, but that doesn't mean it's necessarily a great fit for Stack Overflow. I think this is probably a border case. Mar 12, 2014 at 14:23
  • The question ask specifically about polymorphism and not which language is more powerful in general, in which case that would really bring opinion based answers. On that particular point, a real, definite answer should be reachable. Mar 12, 2014 at 14:34

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