2
$\begingroup$

I recently found out that on mathoverflow.net, when keying in comments, pressing the return/enter key doesn't submit the comment; rather it does the most sensible thing, i.e., goes to the next line. I am aware of the ruckus this topic caused previously on this site (here and here). But now that mathoverflow.net has this "feature", would it be possible to incorporate this on math.stackexchange as well? If not, will this feature be incorporated in mathoverflow alone, when it gets upgraded to the "new" StackExchange platform.

$\endgroup$
11
  • 6
    $\begingroup$ Comments here intentionally don't allow for paragraphs, I don't see the benefit of this without paragraphs. $\endgroup$
    – user9733
    May 14, 2013 at 8:03
  • $\begingroup$ @MadScientist What do you mean by the statement Comments here intentionally don't allow for paragraphs? I see that the number of characters in a comment both here and on MO is the same $600$ characters. So your statement is unclear to me. $\endgroup$
    – user17762
    May 14, 2013 at 8:20
  • 7
    $\begingroup$ Even if you make a linebreak in a comment the linebreaks are not displayed. You can't create paragraphs in comments, so I don't see how being able to user enter to make linebreaks in comments is anything but misleading. $\endgroup$
    – user9733
    May 14, 2013 at 8:25
  • 4
    $\begingroup$ @MadScientist But line break can $$\text{appear using the math mode}$$ $$\text{Here is a nice little paragraph. If someone wants to write a pragraph he can write it like this, right?}$$ $$\text{What do you think about this? And why are paragraphs not allowed in comments in the first place?}$$ $\endgroup$
    – user17762
    May 14, 2013 at 8:26
  • 3
    $\begingroup$ @user17762 This meta.SO thread was mentioned in comments in one of the questions you linked to: Pressing Enter in comment box unexpectedly submits form $\endgroup$ May 14, 2013 at 12:17
  • 3
    $\begingroup$ @MadScientist: Then read all of the posts of people complaining about the problems that having enter submit a comment have caused them. One benefit of giving enter its normal behavior of making a line break rather than the abnormal behavior of submitting a multiline text field is that it fixes those problems. $\endgroup$
    – user14972
    May 14, 2013 at 14:01
  • 6
    $\begingroup$ @user17762 Your text-in-mathjax paragraph is not nice at all; it breaks the layout of the entire page. If I had my druthers, SE would disable display mode in comments, along with \text... commands. $\endgroup$
    – 75064
    May 14, 2013 at 14:05
  • 1
    $\begingroup$ It seens opinion on stack overflow is currently radically different than here at MSE. $\endgroup$
    – user14972
    May 14, 2013 at 15:12
  • $\begingroup$ @Hurkyl, right, and yet they choose to deploy said change across all sites, even though it would seem only useful for big sites like SO. Oh well, what can we do... $\endgroup$ May 14, 2013 at 15:58
  • 4
    $\begingroup$ Personally I like the current behavior. Comments should be quick, and one thing that makes commenting easy is not having to switch back the the mouse (or use shift-enter). $\endgroup$ May 14, 2013 at 16:50
  • 4
    $\begingroup$ I disagree with @75064 about disabling double-dollar mode. I very often need to write matrix equations in a comment, and without the display mode this produces horrible output and would force odd line breaks. IOW, I very much support the idea of enter-key giving a newline instead of submitting the comment. $\endgroup$ May 20, 2013 at 6:44

2 Answers 2

9
$\begingroup$

Math.SE already has too much valuable content posted in comments. This is a problem for multiple reasons, I'll reiterate a few.

  • Comments are difficult to search.
  • Comments are difficult to port to another site. To display questions and answers, you only need to wrap them in <div> and enable mathjax. Including comments on a page is a headache.
  • Comments have low signal-to-noise ratio, contrary to the philosophy of the network.
  • Back-and-forth discussion is what SE was made to avoid.

Adding formatting options for comments would encourage users to enter more information there.

In my opinion, comments already allow for too much formatting. I would rather see displayed formulas disabled there (converted automatically into inline), and the length of any MathJax code limited to 50 symbols or so.

$\endgroup$
9
  • $\begingroup$ So where do you suggest this valuable content goes? $\endgroup$
    – Lord_Farin
    May 14, 2013 at 14:51
  • 3
    $\begingroup$ @Lord_Farin I guess by "valuable information" I meant "text that essentially answers the question". In which case it should go into the answer box. $\endgroup$
    – 75064
    May 14, 2013 at 14:56
  • 1
    $\begingroup$ Indeed, that is the easy case -- people should really be urged (perhaps after some time) to transform their "answer-comments" into proper answers. But what about comments that provide further perspective, references for further reading or otherwise related material? $\endgroup$
    – Lord_Farin
    May 14, 2013 at 15:07
  • 3
    $\begingroup$ @Lord_Farin If an existing answer lacks references, and you happen to have a reference for the result quoted, you can just edit it in (maybe added at the end, separated by ---, if you feel uncomfortable editing the text.) I think that Math.SE users are much too shy about editing the answers by others, which is an activity explicitly encouraged by SE. ¶ As for further perspectives: if you really want to give OP another perspective on their question, I think you'll need an answer box of your own, not a 600-character message. $\endgroup$
    – 75064
    May 14, 2013 at 15:19
  • 4
    $\begingroup$ Disagree: 1) Comments may be difficult to search, but why should I be bothered by that? 2) Comments may be difficult to port, but why should I be bothered by that? 3) While it is possible to make answers alive, most of the lively interaction here takes place in comments. If you think that lowers the SNR, I say that such exchanges are the spice on an otherwise occasionally sterile site. 4) With many questions that is the only method of teaching that appeals to me. I would rather see the site developed in a direction more supportive of that. $\endgroup$ May 21, 2013 at 7:52
  • 1
    $\begingroup$ @JyrkiLahtonen Perhaps you have no reason to be bothered by the search and portability. But then, symmetrically, SE leaders have no reason to bother with features that do not align with their strategy of content accumulation and monetization. $\endgroup$
    – 75064
    May 21, 2013 at 8:02
  • 1
    $\begingroup$ Quite! I know that I don't speak for all the contributors producing "valuable content". I don't think you speak for all of them either. But you are suggesting changes to the SW that would make continued peaceful coexistence impossible. I fail to see how that would serve the interests of SE either. Two reasons to oppose this. $\endgroup$ May 21, 2013 at 19:28
  • 3
    $\begingroup$ Teaching simply benefits so much from that exchange in a chain of comments. The asker needs the comments to seek clarification. The alternative of continuously editing the answer is undesirable (half a dozen bumps + a broken chain of comments) $\endgroup$ May 21, 2013 at 19:50
  • 3
    $\begingroup$ @Jyrki: It is probably no surprise that I fully agree with your comments here. $\endgroup$ May 23, 2013 at 9:06
7
$\begingroup$

MathOverflow uses the old version of the software. This behavior is the original behavior of he return key in comments.

When MO 2.d occurs, this is likely to be gone.

$\endgroup$
4
  • $\begingroup$ So was math.stackexchange always on Stack Exchange 2.0 platform? I joined the site around $2010$ October, and hence I am not completely clear about the initial stages. $\endgroup$
    – user17762
    May 14, 2013 at 8:24
  • $\begingroup$ @user17762, yes, this site has always been on SE 2.0; in general, all the sites that started out as proposals on Area 51 are SE 2.0 sites. $\endgroup$ May 14, 2013 at 9:24
  • 4
    $\begingroup$ I liked the former typo, @Asaf, but I'm afraid I have to fix it... ;) $\endgroup$ May 14, 2013 at 9:56
  • $\begingroup$ @J.M.: Thanks. That's what you get when you are writing from an iPhone with one eye closed. $\endgroup$
    – Asaf Karagila Mod
    May 14, 2013 at 14:24

You must log in to answer this question.