10
$\begingroup$

Many posts feature "words words words EDIT: Here are different words, so the old words don't apply anymore...."

I find this problematic for a few reasons:

  • It requires readers to read a post, assemble some understanding/mental model of the ideas presented, then discard those in favor of later words representing changes to the previous ideas. You may think I'm overreacting, but look at edit 4 (!) on this post: it completely changes the constraints on the question.

  • It makes the posts themself look less expertly-done. The hallmark of all Stackexhange sites is the presence of expert users and expert content. Having "Edit:... Edit 2:... Edit 3:..." undercuts that. (This is not hyperbole.) Instead of coming to a site and finding interesting and well-written Q&A, we come to a site and find interesting (to some)/frustrating (to others) streams of consciousness.

  • It's redundant. We have revision histories on all posts. They're really good revision histories. They're not hidden, as clicking on "edited XX ago" is available to all.

I think we should discourage "Edit:" syntax. What do you think?

$\endgroup$
  • $\begingroup$ You're talking about appending a segment beginning "EDIT: Instead of X...," right? Just checking. $\endgroup$ – CAgrippa Mar 26 '16 at 3:44
  • $\begingroup$ @CAgrippa that appears to be nitsua60's intention. $\endgroup$ – Xandar The Zenon Mar 26 '16 at 3:56
9
$\begingroup$

In short, yes. We should discourage it.

That said, I think there is a definite place for this usage when questions are first getting posed and commented upon. "Edit:" passages should be inserted in the question at the relevant points, not all stuck at the end. But such a usage can make it much easier to follow the progress of revision and refinement.

Once a question is effectively "done," i.e., ready to be answered effectively, it would be great to polish the prose and eliminate all such passages. Editing to remove "edit," as it were.

In fact, I'd like to see this be a regular basis for flagging a question for the edit review queue.

The ultimate end-product should of course be clean, coherent questions in tidy prose... and that means no "edit:" usages.

$\endgroup$
9
$\begingroup$

The post should always be a coherent whole, not a sequence of original text + edit 1 (which revises some of that) + edit 2 etc. The "Edit:" style comes from the blog world, I think, where revision histories aren't usually available and authors want to call out a change. Here, the edit timestamp (and post bump) signals that there was a change and the curious can look at the revision. But for people seeing the question for the first time, this style impedes readability.

If an edit is in response to a comment, adding a comment pinging the commenter and pointing out the edit is fine. (Ideally those comments get cleaned up later as obsolete.)

$\endgroup$
2
$\begingroup$

See this Question.

The edit addresses points raised in the Answer and comments, and completely made the Answer not match anymore.

Marking the edit is much friendlier. I pointed out that it is not unreasonable to continue to get input that makes you change your original ideas, and it shouldn't alienate existing answers.

Meanwhile, once you're convinced that a particular point ought to be dropped, you don’t want more answers continuing to focus on that. So an inline edit allows both: existing answers still have context, and the question reflects the improvement.

Now I certainly agree that when comments prompt for more details, this is best added as an improved draft in the proper place in the writing, not as a postscript. That should be encouraged.

$\endgroup$
2
$\begingroup$

No. I would neither discourage nor encourage its use.

It requires readers to read a post, assemble some understanding/mental model of the ideas presented, then discard those in favor of later words representing changes to the previous ideas. You may think I'm overreacting, but look at edit 4 (!) on this post: it completely changes the constraints on the question.

You are making an assumption that all edits are placed at the end of a question, which is simply not the case. Edits at the end of a question do make sense in the case of information that was never in the previous version of the question and are thus not a problem. One (or a percentage) poorly formatted question does not make a certain syntax universally bad.

It makes the posts themself look less expertly-done. The hallmark of all Stackexhange sites is the presence of expert users and expert content. Having "Edit:... Edit 2:... Edit 3:..." undercuts that. (This is not hyperbole.) Instead of coming to a site and finding interesting and well-written Q&A, we come to a site and find interesting (to some)/frustrating (to others) streams of consciousness.

This is opinion. A valid opinion, but still an opinion. A lot of what we do here requires a poster's thoughts to evolve before a final question comes out of it...I do not mind that on the page personally. That is my (also valid) opinion.

It's redundant. We have revision histories on all posts. They're really good revision histories. They're not hidden, as clicking on "edited XX ago" is available to all.

That's great but I can't recall looking at it on more that one or two occasions, and it was mainly because I answered a question, an edit was made and I wasn't sure where it was...which again, is why I do not mind someone pointing their edits out in the question/answer...it's easier.


I prefer to lean toward the benefit of askers and answerers. These are the people truly engaged in the question. People passively reading a question are far less of a concern to me than active participants...after all those passive readers have nothing to read if the participants don't write anything down.

I am not advocating that we encourage this format, I am simply saying leave well enough alone and leave it to the poster.

If you believe that a format change would make the question better, feel free to edit the post. If the poster doesn't like it they can always roll it back.

Worldbuilding is not an exact science...asking a question here is much more complicated than asking a question about coding or chemistry, it should not surprise anyone that edits are common...therefore, while the question is being actively worked, I find the syntax valuable. If we want to mention that once an answer has been accepted it is good form to clean things up that is fine with me.

$\endgroup$
1
$\begingroup$

I believe that as soon as an answer is posed that will suddenly acquire a changed or disabled meaning, that editing the basis of the question is, no pun intended, out of the question.

Whereas with an answer, there is really absolutely no point at all to say you edited it, people can take an answer any way and choose whether to upvote or downvote without knowledge of its edit history.

However, I think that if a basic concept is being changed, if it absolutely must be, a comment that says it has been is always nice.

$\endgroup$
  • $\begingroup$ Code Review has a strict policy of no code edits on questions with answers. Code edits should instead be posted as new questions. However, Code Review also hardly ever closes questions as duplicates unless something gets double posted. We do close as duplicates. So I think that our policy should be less rigorous on that point. Particularly if the person isn't changing their intended question but clarifying it. I.e. many edits are just making it clearer what they wanted to ask. $\endgroup$ – Brythan Mar 31 '16 at 4:56
-1
$\begingroup$

I am wondering out loud whether Edit should be a key word used to track questions with answers. Such as, hypothetically (and a bit dramatically):

Q: How can my spaceship generate gravity by spinning. || Edit 1: Gravity should be the same as on Jupiter

A: It can spin at XYZ speed with ABC distance || Edit 1: It should spin at ZYX speed at CBA distance to achieve the same experience as gravity on Jupiter

Obviously that's a terrible question, but maybe "EDIT" can be used to track changes in Questions and Answers. If I answer a question and then go jogging, but the question has edits; viewers will know that I am answering PRE-Edits and not result in me being down-voted for answering the wrong math (or whatever).

$\endgroup$
  • 2
    $\begingroup$ If the edit is substantial enough to invalidate existing answers to the point of those answers suddenly being downvote-worthy, then the edit should be rolled back because it turns one question into a different question. That new question may be worth posting as a separate question, however. $\endgroup$ – a CVn Mar 30 '16 at 8:46
  • $\begingroup$ @MichaelKjörling - Agreed, I'm just thinking out loud whether there can be a visual tracking mechanism. $\endgroup$ – Mikey Mar 30 '16 at 9:09
-2
$\begingroup$

I see one huge advantage to the "Edit" syntax : it allow answerers to quickly find the new informations without having to read and compare the post to its previous state.

Disallowing the word "Edit" does not prevent the radical change in content and complicate life for those who want to produce and maintain answers.

While this might lower slightly the readability for newcomers, those who already took some pain and some time to answer will have a better experience.

Finally, both answerers and new readers can see the rate of modifications and can have an easier time deciding if it's worth answering a question that dramatically change overnight to suit some whims or if it's a couple important refining following comments.

As a side note, I'm not saying it should be a rule, just that properly used, it's a tool newcomers, answerers and followers can use and maybe even enjoy.

$\endgroup$
  • $\begingroup$ Orders of magnitude more people read a question and its answers than ever answer the question. Would you care to comment on how the convenience afforded answerers compares to the disservice to so many readers (that I claim, in the original post)? $\endgroup$ – nitsua60 Apr 4 '16 at 1:26
  • $\begingroup$ @nitsua60 The convenience depends on when people are reading the question. I'd assume that early readers are more likely than late readers to consider attempting an answer. I'm with James: encourage a refactor after acceptance. $\endgroup$ – Damian Yerrick Apr 6 '16 at 16:51

You must log in to answer this question.

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