10

I just noticed a community wiki question that I edited in the past shows me as the posting user. Per Tim and Bill's comments below, this is because I am responsible for the largest percentage of the edits.

Why do community wiki posts show the last editing user user responsible for the largest percentage of the edits as the OP instead of the actual original poster?

enter image description here

Improve this question
5
  • 2
    They don't show the last editor, but the most prolific editor. However, I don't think that you're responsible for 87% of the content of the post...I wonder if the calculation excludes the original revision, and if that's intentional.
    – Tim Stone
    Commented Dec 13, 2012 at 16:00
  • 1
    @TimStone That looks right. I think Rachel is responsible for 87% of the diff, not 87% of the total content.
    – Bill the Lizard Mod
    Commented Dec 13, 2012 at 16:01
  • @TimStone and Bill - Thanks, I've updated the question to reflect that information
    – Rachel
    Commented Dec 13, 2012 at 16:07
  • I thought there was a SO blog post that explained why CW posts emphasize the largest contributor over the OP, can't find it though.
    – Servy
    Commented Dec 13, 2012 at 16:10
  • "Current diff algorithm makes CW posts look like contributors have been abducted by aliens..."
    – gnat
    Commented Dec 13, 2012 at 16:27

2 Answers 2

Reset to default
12

To understand this, you have to recall the intent of Community Wiki was to be... Well, a wiki: a collaboratively-edited document with no real owner. This is in contrast to normal questions and answers, which have an explicit owner (the original author) who retains certain rights over the post even if none of their original work remains in it!

In the early days of Stack Overflow, there was a lot of debate as to how "ownership" would be handled in the case of multiple authors; at one time, there was even hope that a system could be devised for splitting reputation gained from votes between the editors who contributed to a post. Although this would be ripe for abuse, the "strong owner" model also has serious down-sides: those contributing substantially beneficial edits get little recognition and no extrinsic reward for doing so, which can be discouraging (this is also why posts that've been extensively edited by several people used to automatically be converted to CW).

Community Wiki is (was) intended to be a compromise there: no one gets reputation and the owner is de-emphasized — at one time, no names were displayed below the post once it had been edited by someone other than the original author! This isn't as strange as it sounds — when's the last time you saw an author name on a Wikipedia article (outside of the talk or history pages where it's still more of a social convention than anything)?

However, folks accustom to the "strong owner" system elsewhere on SO weren't particularly comfortable with seeing no names on a post, and there were several requests for doing exactly what you describe: showing the original author always, with the last editor separately noted as on normal posts.

Instead, the current system for attribution was established: the user contributing the most content to the post is named, with the % of their contributions and number of other editors noted as well. This provides some attribution, without giving one person undue credit for what is supposed to be a collaborative work.

Although as gnat notes, the algorithm used for calculating contribution percentages isn't exactly perfect.

Improve this answer
6
  • It seems like a bit of an understatement to say the algorithm "isn't perfect" :) Now that it's been a few years and things have changed a bit, do you think it would be OK to post a feature-request to go back to not displaying the Owner at all for community wiki posts in cases where there is more than one significant editor?
    – Rachel
    Commented Dec 13, 2012 at 17:17
  • I don't see the point of it, @Rachel. For starters, CW really shouldn't be used very often - and when it is, there should almost always be multiple authors. But more importantly, the simplistic algorithm actually works reasonably well once the document is well-formatted. Yes, it's crap when someone writes a three-page screed with no line-breaks, or typos on every single line and you go in to make it readable - but after that, it should work pretty well.
    – Shog9 Mod
    Commented Dec 13, 2012 at 17:53
  • I suppose I'm thinking more of questions that get made community wiki due to the number of answers instead of ones that are made community wiki due to the number of edits. Thanks though :)
    – Rachel
    Commented Dec 13, 2012 at 18:33
  • your reasoning in favor of line breaks count is quite compelling; in my experience these are typically really important for readability. But the fact that algorithm counts invisible consecutive breaks and especially that it counts completely irrelevant breaks in rolled back revisions makes me feel that it is a bit too err straightforward to be reasonable, isn't it?
    – gnat
    Commented Dec 13, 2012 at 20:37
  • @gnat: it's a really dumb algorithm, but it's fast and... better than nothing. That said, I certainly wouldn't mind seeing it get better.
    – Shog9 Mod
    Commented Dec 13, 2012 at 20:50
  • well, fast, hm let's see... Data access wise, ignoring rolled back revisions would be at worst only twice slower: one could do just one additional pass through post history to find out which revisions are valid. Determining valid revisions is trickier; brute force solution is O(N^2) where N is number of revisions. On the other hand, this looks like a known graph problem (I think I even have seen it in Skiena manual) and it's quite likely that there's O(N*logN) or even O(N) solution
    – gnat
    Commented Dec 14, 2012 at 6:30
7

Because community Wiki posts are community-owned. That's all. The person who contributes the most to the post isn't the OP, but merely the one who made the most significant contribution to it.

In many cases (but not all) the OP turns out to be the most prolific editor (having contributed the entire content of the original post).

Improve this answer
5
  • If they're community owned, shouldn't they show no OP instead of the most prolific editor? It seems incorrect to attribute a post to someone who isn't actually the person that posted the question, and may not even be the person who wrote the majority of the content. (In my example, all my edits did was rephrase a few parts of the question. None changed the actual content in any substantial way)
    – Rachel
    Commented Dec 13, 2012 at 16:21
  • Also per Bill's comment above, it looks like it determines who to attribute the post to based on percentage of edits, not percentage of the total content, so I think your final line is incorrect.
    – Rachel
    Commented Dec 13, 2012 at 16:34
  • 1
    @Rachel CW attribution is based on the count of line breaks
    – gnat
    Commented Dec 13, 2012 at 16:37
  • @Rachel The issue there is that it's incorrectly attributing the majority of contributions to you, not that the most significant contributor is the named on the post. I'm also doubting that the OP doesn't count; I've seen lots of CW posts where the OP is the primary contributor and they had no edits.
    – Servy
    Commented Dec 13, 2012 at 16:37
  • @gnat Thanks for that link, it seems to explain why I'm considered the most contributing editor. Maybe it's time for a [feature-request] to remove the name entirely from community wiki posts since their algorithm for determining the most contributing user is so flawed, and because many users are used to associating the name in that spot with the OP :)
    – Rachel
    Commented Dec 13, 2012 at 16:46

You must log in to answer this question.

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