28
$\begingroup$

Inspired by this meta post on RPG.SE: I think we should keep in mind that posts with lots of "Edit:" and "Update:" sections are a bit hard to read from beginning to end, and we should strive to avoid this style (which I'm certainly guilty of using from time to time).

Yes, sometimes we need to change what we've written. But the revision history for the post contains all past versions in case anyone is curious. And most future visitors probably aren't interested in the full history of how a great question or answer got to where it is.

So this is a call for us all to try to make smooth-flowing posts whenever we edit new information into them.

Also, if editing another post for whatever reason, editors should feel free to incorporate updates smoothly into the post. (This last bit might be a bit more controversial here on Physics, where it seems most of us are loath to edit others' posts.)

Is this a good idea? Am I missing some reason why physics posts should read like revision histories?

$\endgroup$
  • 11
    $\begingroup$ In some cases I've had to do this to avoid invalidating existing answers, especially if the edit would make them sound non sequiturish. $\endgroup$ – Manishearth Jun 20 '14 at 19:21
  • $\begingroup$ @Manishearth Do you have an example? $\endgroup$ – Bernhard Jun 21 '14 at 18:39
  • 1
    $\begingroup$ @Bernhard not at the moment, no, but I'll look for one tomorrow or the day after (at a conference, rather busy) $\endgroup$ – Manishearth Jun 21 '14 at 18:48
  • 1
    $\begingroup$ I think sometimes adding the "inline edit history" is useful when the question is misinterpreted, but perhaps we should strive to stop doing that. $\endgroup$ – Kyle Kanos Jun 22 '14 at 2:18
  • 1
    $\begingroup$ @Kyle The reason you should not do that is 1. New readers are confused and read an incomplete question first, and then the correction. 2. If a post has changed, the revision history show exactly what has been changed and where in the text. That is more powerful than a few lines behind the question. Remember this is not a forum. $\endgroup$ – Bernhard Jun 22 '14 at 7:42
  • $\begingroup$ I almost always edit that stuff out and make a smooth flowing post. There are few (if any) cases where it makes sense to leave "edit" tags. $\endgroup$ – DanielSank Nov 26 '15 at 6:01
  • $\begingroup$ At times, when there are only a one edit and it is highly important to mention the change, may be appropriate: meta.stackexchange.com/a/14519 $\endgroup$ – user191954 Oct 23 '18 at 15:17
  • $\begingroup$ Related: meta.stackexchange.com/q/323116 and links therein. $\endgroup$ – user191954 Jan 29 at 15:27
11
$\begingroup$

I believe one reason for using these marked revisions is when the question itself is edited by the OP in a significant way, which necessitates an important change in the answer. The answer may have been a good answer to the original question, and the answerer is reluctant to suppress it altogether : he'll rather add an appendix answering the new question.

Note also that, in such edits, it is often the main text of the question which is edited, not the title: thus the answer remains a good answer for those who've been attracted there by the title, not by the particulars the OP has added since.

Thus a side question would be also, when the original question is relevant and gets answered, should the OP be allowed to modify significantly his question or should he be directed to post another question?

$\endgroup$
5
$\begingroup$

This is not to argue in favour of the practice, but I think one reason people do it is to make clear what the new part is. For example, you post a long answer, but the OP raises a point they didn't understand that needs to be addressed. You could just edit it seamlessly into the answer, but then it would be hard for the OP to find the part that refers to their specific point. Under these circumstances I sometimes put an "update" heading as a sign to the OP that this is the part they should be looking at.

Of course, the OP could just look at the edit history in this situation, but I have a feeling that most people don't do this, and even if they do they have to extract the relevant information from a diff, which isn't always ideal.

Anyway, this isn't a proper answer, it's just one reason why it sometimes happens.

$\endgroup$
3
$\begingroup$

Yes please. Many of these posts are very difficult to read for no good reason.

$\endgroup$
3
$\begingroup$

Sometimes the actual answer changes with time, like BICEP2 more or less retracting their detection of gravitational waves, neutrios found not to be superluminal, the astronomical unit is redefined as a contant, etc.

In these cases I like to see an updated answer (possibly years later) along with the original answer.

What was the major discovery on gravitational waves made March 17th, 2014, in the BICEP2 experiment?

$\endgroup$

You must log in to answer this question.