18

So it's summer-time, and while you may be thinking that means warm temperatures and blue skies (sorry, Southern Hemisphere!), the Stack Exchange network has caught the clean-up bug.

Our sister sites like OnStartups and Ask Ubuntu have proposed cleaning up old questions to spruce up the place, and we thought it'd be a good idea to explore here on Programmers.SE, as well.

Stack Exchange has an automatic deletion system in place (see Jeff's answer for the details), but we've still amassed 2,690 (and counting) closed questions so far. The idea would be to sweep up the hundreds of questions collecting dust because they weren't great fits for our site and were closed or abandoned.

It'd also be a way to get everyone involved in moderating the site, and we moderators would rely heavily on you guys to go through and take a look at what needs to go or what needs to be reviewed. Plus, it's a chance for more flag weight increases than you can possibly imagine!

Criteria for deletion

Our thinking right now is to follow the leads of our sister sites and delete questions that have one or more of the following criteria:

  • Have a vote total of -2 or less (Edit: mostly complete)
  • Have a very low view count
  • Have no answers, or no upvoted answers

With the additional criterion that the question must've been closed, ideally for a decent amount of time to allow the original poster to revise the question and try to get it reopened (let's say, a month).

Additionally, based on Joel Spolsky and Robert Cartaino's suggestions, any questions with have not been answered and haven't had any activity in several months (i.e. the "forgotten" questions).

Getting involved

This is where you guys come in. If you see a question that's worth saving but exhibits the qualities above, fight for it. If a question is sitting around that really doesn't belong here, tell us.

What does that mean?

  • Identifying questions that could be improved. Some questions were closed because they were bad fits for the as written, but nevertheless had some really great answers. If you think you can get the question into a state that'd make it a fit for our site without invalidating the answers, edit the question or suggest an edit.

  • Voting on questions. If you have the ability to vote for deletion or re-open, use your powers for good. If you see a question that's wrongfully closed, vote to re-open it. If you see a question that should be deleted, vote to delete it.

  • Flagging questions that need moderator's attention. If you find a closed question that needs one of us to take a look at it, flag it and let us know what your concern is. If you don't have the ability to vote to re-open or delete but feel strongly about it, let us know.

Finding candidate questions

Programmers.SE search has a few operators to help locate candidate questions:

Questions and comments?

So what do you guys think? Are there any other parts of the site we should be looking at for some clean-up? Do you have any suggestions or ideas about how to handle this?

  • 1
    Isn't most of this done automatically? Or does it take too long to kick in? "We already auto-remove negatively voted unanswered old questions automatically after 30 days, network wide, with no human intervention required." – Ivo Flipse Jun 26 '11 at 9:52
  • Instead of just a cleanup, I think we should identify all questions that are questionable and perhaps work on cleaning up the FAQ and what's on/off topic here. I know it's been getting a lot better in the past few months, but there's still some work to do. Just go through and pick out questions that someone is like "what about this?" and see what the community thinks about it (even if the question is highly rated and viewed), posting each one as an "answer" so that it can be commented on and discussed? – Thomas Owens Jun 26 '11 at 12:45
  • @Ivo We've amassed close to 2700 questions in less than a year, which is a lot of broken windows. The goal, like the other sites, would be to speed the process along. – user8 Jun 26 '11 at 21:29
  • 1
    @Thomas similar to my response to Ivo: we have close to 2,700 closed questions. Re-evalutating even a sample of those (not sure I understand the value of doing that in the first place) would be really difficult. Nevertheless, if you think there's a value in doing that, I suggest creating a new question with your proposal. – user8 Jun 26 '11 at 21:30
5

For reference, the automatic rules.

Enable automatic deletion of old, unanswered zero-score questions after a year?

Just to formally document the exact policies we have in place to remove old abandoned / dead questions:

If the question is more than 30 days old, and ...

  • has -1 or lower score
  • has no answers
  • is not locked

... it will be automatically deleted.

If the question is more than 365 days old, and ...

  • has a score of 0 or a score of 1 with a deleted owner
  • has no answers
  • is not locked
  • has a viewcount <= the age of the question in days times 1.5
  • has 1 or 0 comments

... it will be automatically deleted.

This check is run every week across all sites.

  • I think in our case, only a small percentage of questions would fit those criteria (and not for at least another 3 months in the case of the second set). Is the intention we keep our closed questions in perpetuity if they don't meet these criteria? Right now, that's over 20% of questions on the site that have no future or were determined to be of little value. – user8 Jun 26 '11 at 21:33
  • @mark just making sure you know what's automatic. If you have suggestions for improving these automatic rules, we're all ears. Well, not literally because that would be disgusting, but you know what I mean. – Jeff Atwood Jun 26 '11 at 21:34
  • shudder I'll have to think about it (the suggestion, not the ears). But thanks for pointing this out: I think the moderators were aware of auto-deletion, but everyone should know it anyway. – user8 Jun 26 '11 at 21:39
  • 4
    The auto-deletion is pretty good for tracking down questions that the entire community assigns no value to (good on most sites). The more pressing (IMO) issue on this site is tracking down and deleting questions that certain elements of the community have wrongly assigned value to. This is going to be like the programming cartoon deletion debate on SO/MSO, times fifty. – Aaronaught Jun 27 '11 at 18:32
  • @aaron sure, I understand, just again -- making sure people know what happens automatically as a baseline. – Jeff Atwood Jun 27 '11 at 20:59
4

I'd suggest starting your search in . I accidentally happened upon that tag page while searching for a duplicate, and... wow. Most of the questions there seem to have no redeeming value, and several are still open!

Almost every question in the list is some variant of:

  • What's the best language?
  • What's the worst language?
  • What's the best language feature?
  • What's the worst language feature?
  • What will be the best/worst language/feature?
  • I love/hate X, do you love/hate X?
  • Let's make a list of languages fitting $hopelessly_generic_criteria!

I'd almost go so far as to say that is effectively a placeholder tag for questions that need community and/or moderator attention. The ratio of on-topic/constructive to absolute crap is shockingly low.

I'm going to start voting to close the blatant polls that are still open; I'd suggest quick and ruthless deletion of the ones that already are closed, and I hope that a few 3k+ community members will chip in and cast their own close votes where appropriate so the moderators don't have to do all the dirty work.

  • I disagree, some of those questions have very good answers, were extremely popular, and/or fall under the "kept for historical reasons" category. Only questions that attracted little attention should be pruned – TheLQ Jul 6 '11 at 17:20
  • 2
    @TheLQ: Popular does not mean good, and "very good answers" is not exactly a qualified statement here. Several of us have been going through hundreds of questions in this tag and finding out that most of them are in fact crap, despite high upvotes (often, but not always, as a result of Stack Overflow users dumping their trash here and not thinking before migrating). If you see close/delete votes and don't agree then go ahead and defend it, but don't think that vote totals or "historical relevance" is going to work as a defense. – Aaronaught Jul 6 '11 at 17:41
4

One thing I would caution against is getting too mired down in consensus building and (what we call) the Mother Teresa Complex (i.e. everything can be saved). Getting a community to crowd-source this sort of thing has turned out to be exceedingly difficult.

There are a whole slew of posts that should unquestionably be deleted. Just delete them; outline the criteria and just delete them.

We're only talking about posts which are taking up space and cluttering up this site: closed-as-off-topic posts, chronically down-voted and unanswered posts… that sort of thing.

The idea of these "Cleanup Days" is to set aside a few days to rid yourself of the worst of the useless clutter. Setting this time aside gives everyone fair notice that this is not a normal part of the site operation.

@Joel talked about these posts on Answer.OnStartups

…this site has an unusual number of really old questions that never got an answer. If a question has been languishing for 6 months without a single answer, I'd say just delete it mercilessly. If it's important, someone will ask again....

"Someone can ask it again" is an important consideration. These posts currently have no useful content.

Start with Joel's suggested "forgotten posts" that have been either abandoned, no longer relevant, or "nobody cares." That will get you 90% of the way there — Then you can circle around and rally everyone around more general question-improvement activities.

  • I think you have a great point, although checking for "forgotten" posts only yields a handful. But taking out the closed questions with less than 100 views and the questions with negative vote score yields several hundred, and would be a nice dent. – user8 Jun 29 '11 at 1:29
  • I guess this is a novel approach to increasing the "percentage answered" statistic: just delete the unanswered ones :) – Benjol Jul 29 '11 at 8:43
  • @Benjol: Kidding aside, removing "unanswerable" posts which should have never passed community scrutiny is a far cry from deleting legitimate content just to fix the numbers... That would not be looked upon kindly. – Robert Cartaino Jul 29 '11 at 14:35
3

To keep the ball rolling on the cleanup, I've deleted all questions prior to July 1st that met the guidelines in the question:

  • Closed questions that had -2 score or lower
  • Closed questions with less than 100 views
  • Closed questions with no answers that weren't duplicates of non-closed questions (duplicates left for SEO)

And all questions, open and closed, with no answers and less than 100 views prior to February 1st.

It's hard to say how many that was (I did it in bursts), but I'd estimate it was about 700 questions total, which is:

  • 20% of all closed questions
  • 4.4% of all questions asked
  • 30% of all questions deleted

This leaves us with 2,832 questions still left closed out of 13,295: 21%.

I'd like to figure out a way to fairly clean up a larger portion of the closed questions, but I could use help with coming up with rules to apply.

You must log in to answer this question.