tea.mathoverflow.net - Discussion Feed (Not providing any background in questions)

Kevin Lin comments on "Not providing any background in questions" (7787)

2010-07-30T00:56:27-07:00

VP and Andrew: I really like what you both said! Great points that I will keep in mind from now on.

Andrew Stacey comments on "Not providing any background in questions" (7786)

2010-07-30T00:52:06-07:00

@VP: +1. Notation should always be explained. I'd not be adverse to scattering the odd link to nlab/wikipedia here and there as well. For such a direct question, I don't think extensive background is necessary, but I do think that a sentence explaining why you want to know might help (crucial point in paper? checking references? idle curiosity?). When I give an answer, I like to know how much I'm helping the person.

The other group of people, other than experts that is, that you have to satisfy is the >3k crew. Just looking at the question, on a bad day I could have been swayed to vote-to-close as it superficially looks similar to a "This is exercise 3.2 in X." question. So I would have liked to see something that would give me (a non-expert) some idea of how deep a result it is in the theory.

None of this would have taken more than a couple of sentences. Remember that by asking a question on MO you are asking someone to do something for you with no other incentive than just knowing that they've been helpful. So it pays to maximise that warm glow that they'll get when you leave that magic comment: "Thanks! That's exactly what I was looking for."

Kevin Lin comments on "Not providing any background in questions" (7785)

2010-07-30T00:45:12-07:00

Let me just say that I apologize if I offended anyone with my laziness. Ordinarily I would have elaborated and explained notion, etc. But I was just feeling kind of lazy, and subsequently thought of doing this (hopefully harmless) little mini-experiment to test what people think. :-) I promise to expand the text of the question a bit later.

Harry Gindi comments on "Not providing any background in questions" (7783)

2010-07-30T00:00:05-07:00

@VP: +1

When someone asks a question without establishing notation and makes liberal use of unexplained abbreviations, I always leave the comment:

"Let X be an E.C. with. C.M.", where "with." stands for "without".

VP comments on "Not providing any background in questions" (7782)

2010-07-29T23:31:44-07:00

Let us distinguish between (a) providing background and (b) establishing context and notation. While (a) may or may not be helpful, I think that (b) is absolutely essential to productive discussion, as has been amply demonstrated by earlier questions where the lack of precision resulted in answers all over the place and back-and-forth comments. I don't think it's too much to ask the author to provide unambiguous notation from the get-go, so that answers and comments need not do it over and over again: even if individual people are experts, they may not quite be talking about the same thing. This is covered in "How to write a good MathOverflow question" cited by the FAQ, section "Be precise" ("Of course, the includes stating the hypotheses and goal of your problem"). At the barest minimum, ask yourself whether the statement of the question would be meaningful and acceptable in an abstract of a paper.

In this particular question, one has to rely on the title to guess what HH stands for, and that still isn't sufficient, because the title says "Hochschild (co)homology" (so which is it?), and Mod_A is ambiguous (cf the discussion between Mariano and Kevin in the comments to Mariano's answer).

PS The quoted sentence has a typo (the --> this).

Qiaochu Yuan comments on "Not providing any background in questions" (7780)

2010-07-29T22:51:46-07:00

@AJ: +1. That is more or less my view.

theojf comments on "Not providing any background in questions" (7777)

2010-07-29T20:41:39-07:00

I do a lot of learning on MO. Certainly the largest source is the answers to questions I have posed. But almost as large are (1) taking the time to yet again think about the definitions so that I can gloss them in the question, and (2) listening in on other conversations. So I strongly believe that questions should have enough background: it's beneficial both to OP but also to the audience. In any case, it's clearly an issue of etiquette rather than law.

Kevin Lin comments on "Not providing any background in questions" (7775)

2010-07-29T19:30:45-07:00

@Mariano: The Mythical Book on Hochschild Cohomology!!! The mere thought of it makes me giddy. We eagerly await its arrival... :-D :-D :-D ;-) ;-) ;-)

Mariano comments on "Not providing any background in questions" (7774)

2010-07-29T19:23:51-07:00

As someone who speaks in public about Hochschild cohomology quite often... I doubt background would have helped much here, short of writing first couple of chapters of The Mythical Book on Hochschild Cohomology.

ajtolland comments on "Not providing any background in questions" (7773)

2010-07-29T19:21:55-07:00

There's nothing wrong with not providing background for this sort of question.

Generally speaking, the rule is give what you are hoping to get. If you want a terse answer to a technical question, that's fine. What's obnoxious is asking someone to write an essay in exchange for a sentence or two.

Akhil Mathew comments on "Not providing any background in questions" (7772)

2010-07-29T19:08:46-07:00

I actually agree with Kevin on these kinds of reference requests. I'm not sure how adding a brief explanation of what Hochschild cohomology was would have made it any easier for an expert to answer his question, and I doubt a paragraph or so of explanation would have been helpful to non-experts.

François G. Dorais comments on "Not providing any background in questions" (7771)

2010-07-29T19:04:28-07:00

It certainly shows some level of bad etiquette. In the past I have voted to close some extreme examples of this, but my views were more extreme back then. There is no official requirement to provide background, but it is strongly encouraged since questions without background do not provide as much benefit to the community.

Harry Gindi comments on "Not providing any background in questions" (7770)

2010-07-29T18:31:46-07:00

Yes. It makes it impossible for people who might want to learn what you're talking about to get anything from it. Background isn't for your benefit or your answerer's benefit. It's for the community's benefit.

Kevin Lin comments on "Not providing any background in questions" (7769)

2010-07-29T18:30:16-07:00

I just posed a question in which I didn't bother to give any background whatsoever. My reasoning is: If someone doesn't understand my question from the get-go, then they're probably not going to have an answer, whether or not I provide background. So I might as well spare myself the trouble of typing out the background information.

Do you all consider this bad etiquette?