current community

your communities

more stack exchange communities

company blog
Stack Exchange Inbox Reputation and Badges
tour help
up vote 19 down vote favorite

The question

Maybe this is a question of basic Stack Overflow philosophy, but that linked question seems to me to be an example of a little too much "improvement". (Check the revision history.) However, I could be wrong here.

The OP included some sample code, and the accepted answer (from me) included a suggested fix to the particular markup causing grief. The edits (made originally to both my answer and the question itself; I rolled back the edit to my answer before realizing the question was edited too) significantly changed the context.

Now, I can't argue that the edited result is perhaps more to the point, but to me it drains some of the value for the original asker of the question. Maybe my feelings about the importance of that are misguided.

edit — maybe another version of this question (my question, not the SO question in question) is: how valuable is the widely-varied, often error-filled content of questions posed by less-experienced coders, for example as target fodder for future searches by other less-experienced coders?

What would SO be like if every misguided question were edited down to the fundamentals? Isn't the whole point that some misunderstandings are common? (I acknowledge that part of the issue here may be that my instincts might be correct and good but that this particular SO question had little random content of value.)

share
    
The original contents essential arguments with all their specific details and context should be preserved and only what remains (the dressing/formatting so to speak) can be changed and streamlined. – Trilarion 17 hours ago
up vote 13 down vote

There have been a plethora of questions regarding edits to code in questions and answers. It is generally frowned upon. See When should I make edits to code? for a canonical answer.

In this specific case, the edit changed much more than the bulleted list in that canonical answer:

  • Test your edited code to make sure it works (the brand-new code was presumably tested).
  • Fix syntax errors and typos
  • Improve formatting
  • Correct spelling/grammar issues in the answer body

In the question, the change violated one of the "Don'ts":

  • Change the code logic or functionality

The duplicate post explains why the code is showing the error; there was and is no need to edit this question (which could conceivably happen again if someone naively passes a Facebook ID as a number) to make it more clear.

For these reasons, I believe that these specific edits should be rolled back.

share
5  
But that edit doesn't "change the code logic or functionality". The code still behaves the same way as before the edit. – Gothdo 2 days ago
8  
"still behaves the same way"... The original code was in an event handler for a button, which when clicked shows an alert of a server-side written value. The new code logs a hard-coded value. It's the same underlying JavaScript rule which governs it, which is why the duplicate is accurate, but it is not the same functionality as in the OP. As Pointy pointed out in a comment, the point (sorry) is that you've changed the intent of the asker. – Mike McCaughan 2 days ago
1  
I don't think that highlighting the actual problem when it is buried in irrelevant code is a bad thing. I don't believe that doing so changes the intent, as the intent is to solve the problem is it not? If the problem is the same, is the intent not the same? – Tiny Giant 2 days ago
8  
@TinyGiant That ain't "highlighting". Highlighting would be to add formatting or rewording to put emphasis on the problem area. This was rewriting the code completely. The OP did not intend to ask "why does logging this number to the console log a different number". They asked, "why does passing the Facebook ID to a function as a number show a different number?". The duplicate question would let the OP know they have come across the same problem as that evinced by the modified code, but that was not what they asked. – Mike McCaughan 2 days ago
1  
It is still demonstrating that the number is not the same as what was entered. The back story is irrelevant, and the more general we can make things, the more people that information will help. As I said before, the core problem is the same, and the intent is still to solve the problem, I don't see how this deviates from the intent. – Tiny Giant 2 days ago
1  
Then I would suggest adding an answer to the question I linked to, to say that completely changing a question's code is okay. You should also probably hang out on the suggested edit review queue for the inevitable flood of code edits which result... – Mike McCaughan 2 days ago
9  
@TinyGiant the back story is relevant because it's text that can be found by Google search. The edit history is not searchable. Other people with similar head-scratching Facebook ID problems will not be very likely to happen upon the newly edited question. – Pointy 2 days ago
1  
@Pointy, I disagree, but really don't care all that much either way. I think it is irrelevant noise that should be slimmed down to just the core problem. – Tiny Giant 2 days ago
5  
It should also be noted that the edit invalidated a comment: *It's a Facebook ID, I store it as a LONG in C# * which was a good clue as to what the problem was. I know, comments -- who cares? But if you're going to be radically changing question contents, you should be thorough in your cleanup afterwards, IMO. – Mike McCaughan 2 days ago
    
@TinyGiant There's nothing wrong with presenting a modified version of the question's content in an answer. For a developer experienced in the problem area, yes, it is noise. But, the whole point of any question/answer is to help developers gain experience in a particular area. We can't expect each other (or ourselves) to know the "core" of a problem, until we've seen and understood (a) a question that relates to our initial perspective, and (b) an answer which tells us the "core" we've been looking for all along. – jpaugh 16 hours ago
up vote 7 down vote

My view is that the edits go too far, both in relation to this specific question and in relation to the wider question of generifying questions.

This question

The question is a duplicate. The edits generified the duplicate. But we already have the generic version (its dupetarget). Editing it to make it generic when we already have the generic question is actively harmful: It reduces the search surface area, reducing the odds of others finding a solution to their problem. (Lots of people out there trying to do Facebook integration.) Increasing the search surface is why we don't delete duplicate questions. There's no value in a series of generic clone questions. There is value in rich and varied ways of asking the same question, all pointing to a canonical set of answers.

In general

If it weren't a duplicate, I'd see an argument for generifying it so it would be a useful dupetarget. However, if it seemed like the specifics would be useful from a searching perspective (as I believe is this case here), I'd be keen to preserve those in the course of generifying it, not removing them entirely.

share
up vote 4 down vote

What you see here is (too much) code being turned into an MCVE. Nothing else, and I'd like to see more of these edits. Preferably by the OPs themselves.

but to me it drains some of the value for the original asker of the question.

That could be, but it increases the value for future readers of the question by a much larger amount.

share
11  
I appreciate that point of view, but on the other hand if everything on the site were edited down to an MCVE, we'd have a giant documentation database containing stuff that's already written down elsewhere, and which the less-experienced coders have already failed to find and understand. – Pointy 2 days ago
6  
Note that the original question included the (incorrect) assumption that the problem was happening as part of calling the event handler function. We know that's wrong, but that inexperienced coder and maybe others don't. Of course, I have no idea how likely it is for that to come up again for some other random person, so getting ideas is why I posed this meta question. – Pointy 2 days ago
    
I agree with you that that's a very nice edit to make, but the answer would need to be edited as well, otherwise, it is refering to code that isn't present in the question, which makes reading it confusing. – Tunaki 2 days ago
    
@Pointy Since most people seem to agree that my edit was good, could rollback your answer to revision 4, to keep it consistent with the question? – Gothdo 2 days ago
5  
@Gothdo I think the issue is far from settled. How will the next person who is confused that their Facebook ID values are being mysteriously altered stumble across this question now that all mention of that is gone from the question text? Providing searchable content is one of the prime reasons for Stackoverflow's existence. – Pointy 2 days ago
2  
@Gothdo Note that the question could be "cleaned up" much as you have done while still preserving the original text of the question. I don't see how that could possibly make things worse. – Pointy 2 days ago
    
@Gothdo in fact, I have an idea – Pointy 2 days ago
    
"...code being turned into an MCVE..." As long as you don't fix the code along the way, that's probably fine. – Trilarion 17 hours ago
up vote -16 down vote

I have an opinion on this issue that might be controversial. I think that the less editing the better, I'm even in favor of leaving misspelled words, bad English, non-technical terms. There's a reason for that and it has to do with "googling Stackoverflow", namely I sometimes was able to find an answer just because the question had been formulated "badly", similarly to how I would have asked).

What I think are useful edits are

  • code formatting
  • anything that improves readability
  • tagging
  • removal of "thanks"

In this particular case I find the editing excessive, mainly because it removes the whole context (the fact that the number is a Facebook id) that makes the question interesting.

And for polished examples there's always the Documentation.

share
3  
bad English makes it harder for other non-native speakers to understand (particularly those from a different culture to the original posteer.) – Martin Bonner 18 hours ago
1  
"leaving misspelled words, bad English, non-technical terms" =/= "anything that improves readability" – OhBeWise 18 hours ago
1  
I want to edit this and fix the grammar so bad XD – Feathercrown 18 hours ago

You must log in to answer this question.

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