What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 128 Stack Exchange communities.

So I was doing a bit of cleanup on Stack Overflow. Going through the tag and voting to delete the worst offenders, the ones that asked for “one item per answer”, and got it, and ended up with a jumble of information that's completely unexploitable.

The thing is, many of these questions are locked. So my only course of action is to flag and ask a moderator to delete. Which they are rather reluctant to do, because there's a clear proof of the question's popularity (hundreds of upvotes) against just me, with my confidential flag, that does not show any public support for the deletion.

We clearly need a better way of cleaning up the garbage, that doesn't make the participants afraid. For a start, long-term locking is not the right tool for these old questions. What we're doing here is to frame the garbage.

framed rubbish

So, please unlock all such old questions. Either they're worthwhile after all, in which case they belong open, or they are not, in which case they belong deleted.

share|improve this question
5  
Are you thinking we'd be able to delete it if it weren't locked? Posts need more delete votes based on the score of the posts in the thread; deleting those old posts by vote is near-impossible –  Michael Mrozek Nov 19 '11 at 17:33
5  
@MichaelMrozek We can at least show support. The SO mods are less reluctant to tack on their binding votes to 20 community votes than to throw the sole blow. –  Gilles Nov 19 '11 at 17:46
9  
-1, as someone who flagged many of those to request the locks. It's very possible for non-constructive questions to have value (for example, the list of free programming books is very popular and appreciated), even if they're not remotely appropriate to the Stack Exchange format. I don't think we should have to choose between deletion and unlocking. –  Jeremy Banks Nov 19 '11 at 18:17
4  
@JeremyBanks The list of free programming books is one I didn't flag, because it has a little redeeming value. The ones I flagged were plain one-item-per-answer stuff, with no usefulness whatsoever. The free programming books should perhaps not be deleted, but if not it should be maintained, hence editable. In either case, locking is detrimental. –  Gilles Nov 19 '11 at 18:20
4  
3  
I have a perfectly reasonable solution which this comment is too small to contain. –  Adam Davis Nov 19 '11 at 23:03
1  
Seeing as you're in the mood for tidying up cruft, how about scanning the magento tag and flagging some of the real harm that's going on there. –  Kev Nov 20 '11 at 14:28
add comment

4 Answers 4

up vote 7 down vote accepted
+50

For a brief moment there was a lock reason of "too many answers". But that atomic lock that was removed because:

  • lock should not protect a question from close or delete votes
  • dozens of answers signal a bad question ill-fit for the Q&A system
  • children need to be bleeding from their pants in order for a question to be justifiably locked, as in the case of edit wars or blazing comment trails.

Don't see any real difference between that and the "this fugly piece of dung that everyone has given their juices on needs to stick around because some people equate flies to lasting value" lock reason.

If most of these locked answer mills had any real value, it would have found its way into the relevant tag wikis.

If there are no relevant tag wikis the tips and tricks could be copied into, then it's not really in scope for the site and should be closed and removed.

Locking these questions with "historical significance" is saying that because users have unloaded on the site and have not bothered to clean it up or move it to where it needs to be, we should cherish it for all time as a glorious example. Wrong signal.

share|improve this answer
add comment

That pony moved. I saw it. Quick! Give it another whack...

This is an old, old argument... A quick glance at the comments and revision histories of most of these questions will tell you that there was never a time when they were universally accepted as "appropriate" for the site. The fact that some of these questions are over three years old (and revision log entries for closing and re-opening weren't added until much, much later) should also remind you that there is often a good deal of controversy lost to time...

So in a sense, the fact that they still exist at all puts a lot of weight behind the arguments against deleting them. These may be straddling the edge of what's appropriate, but they've kept their balance while many, many others have failed. No matter how much I might dislike some, I do accord them a sort of grudging respect for that reason alone.

More importantly, it hints at why they're locked in the first place, as they're positioned squarely at the point where community moderation falls apart into bickering, with moderators left to pick up the pieces over, and over, and over again...

Idle mods are Satan's playthings

Stack Overflow moderators are busy people. They'd be busy if they didn't have lives, families, jobs... The sheer quantity of moderation needed on a site the size of Stack Overflow is a lot to ask. So there's something a little bit... thoughtless... about claiming they should head out and crack open dozens of wasps nests just because you missed the chance to stick your nose in them.

Bullets with bat wings

That being said... There is a time and a place for re-opening some of these questions to community moderation. Some of them were hastily locked; others have simply not aged gracefully. There's no harm in periodically requesting that a moderator review them.

Reasons to lock (or keep locked) a "historical" question

  • Community has consistently indicated their support via votes, comments, and re-opening when closed.
  • Hosts answers of lasting value to the Stack Overflow community and The Internet at large
  • Reasonable expectation that no new answers will be needed
  • Ongoing cycle of closing and re-opening are creating a distraction and fostering unhealthy arguments.

Reasons to delete a question

  • Closed for lengthy period of time, no apparent effort to re-open
  • Answers are generally of poor quality, and in total provide nothing of exceptional value to the site or the larger viewing audience.
  • Frequent source of and justification for low-quality "copycat" questions
  • Experienced users have indicated their support for deletion by voting in significant numbers to delete (even if the total number of deletion votes falls well short of what is required to automatically remove the question).

When in doubt, sit on your hands. It keeps them from curling into fists.

So my advice to moderators is, be slow to lock historical questions, and just as slow to unlock. If a moderator doesn't see any compelling reason to lock, unlock, or delete... If they are secure in their conviction that the current state of the question is the one it needs to be in... Then they should do nothing. Dismiss the flag. Future flags will trigger the display of previous flags, allowing future moderators to gauge support over time. Revision history, comments on the question, and the moderator's own best judgement should support the flag. If someone really feels strongly about keeping or ridding the site of a question, a specific discussion here on Meta can go a long way toward backing them up.

The issue of flag weight has been mentioned in comments a few times. I don't really see that these flags need to be declined as a matter of course, but if that thought keeps someone from flagging... Well, good. If you care more about flag weight than you care about seeing action taken on a question, then you don't care very much.

share|improve this answer
6  
I would add this to my list of posts demonstrating that employees think flag weight is stupid, but I've run out of space –  Michael Mrozek Nov 19 '11 at 21:02
add comment

Ok ok so I've been watching all the discussion on this happening in various places in chat. I would like to add my opinion because this is the internet and that's just what you do, isn't it?

Firstly, I'd like to make an observation - SO is now fairly mature. Back in the day, in order to find itself, Stack Overflow had to experiment with illicit questions. They're how we worked out they just do not work. Now Stack Overflow is mature and all grown up, it knows these questions are bad, but it had to realise this for itself.

I think Mark Trapp sums it up best:

Is the internet better off by having those things on Stack Overflow? Do they attract the right people to a site that's supposed to be generating high quality questions and answers? It's not just the content of the question, but where the question is.

That is the determining factor. Are these questions any good? Do they provide value to someone, somewhere, somehow, in the life of a daily working programmer? If so, let's keep them here. It may be necessary to lock/historical significance-note these questions for the benefit of new members and the sanity of the moderator team; I say that's fine. Locks are here for that reason.

The other case is the old stuff that's actually genuinely rubbish. As in, really and utterly pointless.

The problem is finding the difference. So, to that end, let me give some examples from the tag:

  1. What are your favorite Vim tricks? - This question is highly bad for what we understand from the current format. I mean, it's "not constructive" if ever there was a question... but. It's an old question (and therefore a potential wasps nest, in Shog9's words), but I personally have used a trick or two from there. To the average developer who uses vim, this is actually a pretty handy page even if it sucks as a question. It's part of the "those were the days" moments where we remember past foolishness.
  2. Confessions of your worst WTF moment - This question has questionable value anywhere on the internet at all. There are plenty of places on the internet where you can source "WTF" and "fail" should you so desire; plenty of places you can listen to anecdotes, including but not limited to your current co-workers. Votes mean nothing in this context; it's not even like the best trick will float to the top in this case. Yes, I hate fun too. Another example of this sort of question - I mean, the comment left on it was:

    Wow, Digg's redesign looks exactly like Stack Overflow. Oh... wait...

    I don't think I can say it any better myself. These sorts of questions are the things we don't tell our children happened whilst we were fixing the internet.

I think the point Gilles is trying to make here is that by locking the latter set of questions we, the community, can do nothing to delete these questions that really need it. I personally think there's a bigger problem; even if you unlock them, they'll take a lot of downvoting, 5 close votes and many delete votes to burninate and that's assuming 5 random people who thought "oh lolz" don't decide to reopen; every upvote these questions attract makes them harder to remove.

Here is an example of the problem, for those who have not used the post deletion votes:

999 delete votes on the wall, 999 delete votes on the wall and if one of them should accidentally fall...

For reference, at the time of writing the users in various reputation ranges capable of deleting are:

100k+           :42 (42)
50k <= x < 100k :104 (146)
20k <= x < 50k  :494 (640)
15k <= x < 20k  :369 (1009)
10k <= x < 15k  :1417 (1786)
Total           :1786

As such, to remove this post would take approximately 15% of all the 10k users there are.

So I think I'd say this - if you're not prepared to delete them, dear moderators, and lock them deleted, don't unlock them. I think the only way these are realistically going to be removed cleanly (given such high votes, without a to-and-fro in the interval it takes) is by a moderator.

I personally support the deletion of the junk questions and leaving those with some value as is. I also think moving the value content from existing questions into tag wikis is a good idea going forward and is definitely now the right place to host these resources. That said, I see no immediate need to burninate those questions with some value to them just because they don't match the current format.

share|improve this answer
1  
I would be willing to let go of the WTF question... I think it has outlived its usefulness, and I don't think anyone would shed a tear over it. –  Robert Harvey Nov 19 '11 at 23:01
    
The Vim question is not locked. Feel free to hammer it with delete votes. –  Robert Harvey Nov 19 '11 at 23:09
1  
@RobertHarvey I could argue for its staying. I was trying to distinguish between "when we should stir up trouble" and "when it might just be ok to leave them with a historical banner / lock". In my mind, the Vim question could stay under those circumstances - it has value, even if it is not ideal. As for the WTF question - no, I don't think anyone will. –  user142852 Nov 19 '11 at 23:15
1  
The information in the vim question belongs spread out into problem-specific questions. If you want to browse, browse the vim tag. –  Gilles Nov 19 '11 at 23:18
    
@Gilles I don't disagree. I simply argue in its defence for the moment. I see two battles here - 1. The deletion of utter junk and 2. The gradual persuasion towards tag wikis for resources. I say start with the junk that genuinely serves no purpose first and figure out a way that is genuinely acceptable for value questions over time. In the mean time, leave the lock / historical notice. It's not exactly format-conforming, but it isn't totally awful either. –  user142852 Nov 19 '11 at 23:38
add comment

I prefer the lock and place a banner on old questions that just don't belong on Stack Overflow now, in most cases.

They have historical significance, and deleting useful information is a poisonous thing on the Internet (if you don't believe me, then look for Mark Pilgrim's "Dive into" series. All gone).

So yes, it's useful. But it's not appropriate for Stack Overflow.

Why the heck not?

Stack Overflow (today) is about answers to programming problems. The ones that contain code. It's not a site for answers about your feelings, or about common practices (Kanban vs. Agile, FIGHT), or what logging library you should use for a .NET application if the moon is full and MVC is present.

That being the case, we must do something, harumph.

So what do we do? Deletion is out, for the reasons I talked about before. Letting them remain without any indication that they're no longer the types of questions we want to see is also out, because well... Who wants 3000 questions on logging?

So there's locking them. Locking provides a few benefits:

  1. The question is left in its final state. No more upvotes, downvotes, answers, or edits. Nothing to bump it to the front page (thank god!).

  2. A moderator can add the 'historical significance' blurb to it, and no one is going to take it out (especially not the OP, who is probably upset because he stopped getting badges from the question)

  3. The information is still there, and it will be there for eternity*

But what about your best answer, "A List of Free Programming Books?

Update: It's 2012, and we now have instances of people migrating my best answer to tag wikis. I support this, so long as Tag Wikis are discoverable through site searching and google searching. Best estimate is: 6-8 weeks. If this were to happen, then the question should be locked and the historical blurb placed on it.

One of the answers that I have given (in fact, it's my most popular answer) is both a failed experiment and something that should remain unlocked.

It's a failed experiment because when I created it, I just wanted to compile all the answers into one answer (a la Joel Spolsky's recommendation), and I hoped that other people would edit my answer and make it the most complete answer.

It didn't work.

People still added their own answers (even after pages of answers they hadn't read), and I was forced to spend a lot of time looking at all the answers and adding them to my answer.

But I still think it should be unlocked. It's still relevant, useful, and there's nowhere else on the Internet where you can find a list that complete and that can still be edited at any time.

So should it be locked and deleted? Of course not.

Questions that have no present usefulness, or will have no 'new' answers should be locked. Things like, "What is your favorite programmer cartoon."

* or until the thing that replaces the Internet comes along and we forget to migrate our stuff to it.

share|improve this answer
1  
The compilation is valuable; it means I don't have to sift through all of the answers to find the gold. I hope you didn't add it just so people would stop posting new answers. –  Robert Harvey Nov 19 '11 at 22:57
2  
That list of free programming books is why I haven't flagged this particular question: there's some useful information there. It seems to me that this is not the best venue though; a straight wiki would fare better than a Q&A platform. Your anti-deletion does not hold water: I'm not saying they don't belong on the Internet, I'm saying they don't belong on Stack Overflow. As for your pro-locking arguments, I don't understand them: why would you want content to remain on the site if it can't be improved? –  Gilles Nov 19 '11 at 23:13
1  
@Gilles: even when you lock the question, you can still edit the answers... IOW, the "best books" questions could be locked forever and still kept up-to-date by folks making (or suggesting) revisions to the top CW answer. Which is the downside of collecting answers into the question (see: the C# version of "Hidden features") since now if it's locked it's guaranteed to become stale. –  Shog9 Nov 19 '11 at 23:31
    
@Shog - have you just suggested the way forward with these questions? Instead of flagging the question, users need to check if the individual answers have been added to the main CW compilation answer, and then flag the answers as appropriate? –  slugster Nov 20 '11 at 1:07
1  
@slugster: for some questions - those with a consolidated mega-answer - this is completely appropriate. –  Shog9 Nov 20 '11 at 1:27
add comment

You must log in to answer this question.

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