Kuro5hin.org: technology and culture, from the trenches
create account | help/FAQ | contact | links | search | IRC | site news
[ Everything | Diaries | Technology | Science | Culture | Politics | Media | News | Internet | Op-Ed | Fiction | Meta | MLP ]
We need your support: buy an ad | premium membership

[P]
The 24 hour edit queue and the spam button

By Rogerborg in Meta
Fri Mar 21, 2003 at 12:32:25 PM EST
Tags: Kuro5hin.org (all tags)
Kuro5hin.org

We've had a 24 hour edit queue with a spam button for long enough to get a feel for it, perhaps long enough so that new members might not even know that the prior method had a fixed 2 hour time limit, and no spam button.

There has been a smattering of opinion on it scattered among other articles, and I suggest that to help rusty out, it's time to collect some comprehensive feedback on whether we feel that this has improved K5, and whether it could be improved further.


Suggested issues:


  • Given that there's been no site news about it, do you even know how it works?
  • Has it improved K5?  Could it be improved further?
  • What is the purpose of the edit queue, and is it fulfulling that purpose?
  • Is the 24 hour edit queue becoming an ersatz publication medium?
  • Are we losing too many good comments attached to bad articles?
  • If the articles posted to section and FP are still of high quality, then the edit queue is demonstrably doing its job, so is it worth worrying about?

Some necessarily simple poll options are presented.  My own opinions are expressed in a reply to this article.

Sponsors

Voxel dot net
o Managed Hosting
o VoxCAST Content Delivery
o Raw Infrastructure

Login

Poll
24 hour edit and spam button
o It's fine as it is. 18%
o Lower the spam button threshold. 17%
o Raise the spam button threshold. 2%
o Cut back the absolute maximum time allowed in edit. 17%
o Raise the absolute maximum time allowed in edit. 1%
o Make it easier to send articles to vote if they do not receive regular edits. 23%
o I care about this issue, and have expressed my views in a write in. 4%
o I don't care about this issue. 14%

Votes: 111
Results | Other Polls

Related Links
o site news
o Also by Rogerborg


Display: Sort:
The 24 hour edit queue and the spam button | 155 comments (128 topical, 27 editorial, 0 hidden)
Write in (5.00 / 20) (#1)
by Rogerborg on Fri Mar 21, 2003 at 07:05:34 AM EST

Overall, the 24 hour plus spam button edit queue is an improvement, and copious thanks to rusty for implementing it.  It allows exceptional articles time to be improved by incorporating factual references and different viewpoints.

However, it is also allowing a lot of mediocre articles to be abandoned in edit, and I believe that's a bad thing, and that it could be improved.  Here's why.

  • An edit queue has one function: to allow editing of stories before they go to vote.  Nothing else.
  • As I understand it, the spam button doesn't register a negative vote, it just contributes to hitting a threshold that sends the story to the vote queue with the same 0 vote starting point as any other story.  
  • It appears to me that the spam button is working well at dumping the worst offenders, but that it is not helping to push mediocre-to-decent articles that have simply been abandoned in edit on to voting and publication.
  • This bothers me because if we become used to authors not responding to edit suggestions, then we'll become less inclined to make them.  I've noticed this in myself.  I now hit the spam button by default, unless the article has just been posted recently (minutes, not hours), or the author is clearly responding to editorial suggestions by making edits (as distinct from posting replies explaining why none are necessary).
  • I believe that we are losing good topical comments attached to bad stories.  I'd rather see us get stories to FP or section faster and then comment on them topically once they're not going to disappear.
  • When I come back and re-read a story that's been in edit for hours, I do so to read the improvements.  If there are none, then I've wasted my time suggesting them, and re-reading.  That irks me.  I know that I don't have to keep coming back to re-read it for changes, but if it's not being changed, then why is it still in edit?
  • It seems clear to me that the 24 hour edit queue is being abused as a publication medium, and often by J. Random Newbie who makes no further contribution to K5.  I have absolutely no objection to posting decent articles by newbies, but I don't want them cheapening the perceived value of the edit process while they're about it.
  • I like the 24 hour limit, and in fact would be happy to see it raised.  If an author wants to post an article in good faith and then wait several hours before coming back to read the comments and perform edits, she should be able to choose to do that (a courtesy post to that effect is a great help).  But most articles can go through edit in minutes, not hours.
  • If no editing is being performed on a story, it should be easier to push to vote.
For all this ranting, I suggest only a slight tweak to the current system.  Instead of having a fixed spam button threshold, just reduce the threshold the longer an article has gone without receiving an edit.  Bearing in mind that we did well enough with a 2 hour fixed cutoff, I'd be inclined to be aggressive about the criteria.

The current threshold seems about right for dumping blatant abuse quickly, so if we start from there and say:

Time to last edit       Spam button threshold
 0-1 hour                == current threshold
 1-2 hours               0.75 x current
 2-8 hours               0.5 x current
 8+ hours                0.25 x current

Does that sounds reasonable?  Further to that, I'd actually suggest removing the 24 hour limit altogether, and in tandem renaming the "spam" button as "ready for voting".  I have no objection to seeing exceptional articles sitting in edit for days, so long as they're receiving active edits, and we're happy to keep improving them there.  I'd just like to see a few more of the abandoned articles be pushed to voting faster.

Let's reclaim the edit queue for those giving, receiving and acting on editorial advice.

"Exterminate all rational thought." - W.S. Burroughs

Also, I should make clear (4.33 / 3) (#13)
by Rogerborg on Fri Mar 21, 2003 at 07:56:03 AM EST

That despite the above, I do think that the current scheme is substantially sound.

I'd be quite happy if we got a consensus that it's fine as it is, just so as long as we give rusty some kind of reasoned feedback about how he's running the site (e.g. "substantially sound")

"Exterminate all rational thought." - W.S. Burroughs
[ Parent ]

now that you got me thinking about it (none / 0) (#154)
by martingale on Sun Apr 20, 2003 at 11:18:47 AM EST

I think I definitely prefer having a two hour limit. I can see theoretical reasons for a 24 hour period, such as allowing everyone around the world to offer editing suggestions, but the thing that bothers me is I've lost the guarantee that the author is sitting next to his story for two hours, waiting to respond to editorial comments.

What if I make a suggestion, and someone else objects/amends it? Within two hours I can respond, but if I have to wait 24 hours to read somebody's objection to my editorial comment, will they wait 24 hours for my reply to their objection? Should the story go to vote after 48 hours in that case?

Interactive editing is just not compatible with long periods. Right now, I'm inclined to go back to two hour editing. Or actually, here's an idea: we could have temporarily two edit queues, one 24 hour and one 2 hour. Authors could choose which queue to submit to, and we could compare the quality of editorial comments.

[ Parent ]

I think you're wrong about what it does. (4.66 / 6) (#32)
by dark on Fri Mar 21, 2003 at 09:19:18 AM EST

Sometimes, after I hit the spam button, I find that the story is now in voting and my vote has been registered as -1.

Some actual documentation would be nice :-)

[ Parent ]

I've heard that (3.50 / 2) (#59)
by Rogerborg on Fri Mar 21, 2003 at 12:42:06 PM EST

But if the spam button threshold is (e.g.) 10 clicks, wouldn't that start those stories at -10 votes?  I don't see that happening.

"Exterminate all rational thought." - W.S. Burroughs
[ Parent ]

The code is your friend (5.00 / 5) (#65)
by cdyer on Fri Mar 21, 2003 at 02:54:34 PM EST

The relevent code is in scoop-0.9/lib/Scoop/Admin/AdminStories.pm.  grep spam AdminStories.pm

It looks like it registers a count to go toward the spam threshold (currently set at 100 votes), and also toward a spam percentage (25% by default).  Once a story receives more than $spam_threshold spam votes, it checks the ratio of spam votes to page views, and if it is greater than the $spam_percent variable, it goes to voting.  It also seems to register a vote when you click the spam button, which I assume is -1, but I didn't trace the code back to where it actually happens.

Also, I'm not sure if it checks for multiple votes from one person.

That said, some more transparency in the interface would be nice.  Like a little message saying "Your spam vote has been received.  Spam ratio is at 7%.  Threshold has not yet been reached."

Cheers,
Cliff

[ Parent ]

By crikey, so it is (3.50 / 2) (#74)
by Rogerborg on Fri Mar 21, 2003 at 03:50:45 PM EST

100 votes and at least 25% of views resulting a spam vote.  So, when I refresh the page (to see edits have actually been performed), I help to preserve the article.  Ouch.

No wonder it's hard to dump stories after they've been in the queue for a while.

Also, when the story gets kicked to vote, it appears to reset the votes for it, so it seems like the claims that hitting spam sometimes registers a -1 are spurious.

Step 1 is to dump that 25% criteria.  I don't want to be discouraged from checking for edits.

"Exterminate all rational thought." - W.S. Burroughs
[ Parent ]

no you're not... (none / 0) (#97)
by janra on Fri Mar 21, 2003 at 08:43:56 PM EST

I believe it doesn't count page views but number of accounts who've viewed it. If you read it one time or ten, you count as one view.

That's how it was explained to me, anyway. It uses the viewed_stories table (the same one that keeps track of which comments you've read) to see how many people have read the article, instead of a counter to see how many times the article was read.


--
Discuss the art and craft of writing
That's the problem with world domination... Nobody is willing to wait for it anymore, work slowly towards it, drink more and enjoy the ride more.
[ Parent ]
-1 if already in voting (4.50 / 2) (#79)
by nxor on Fri Mar 21, 2003 at 05:21:36 PM EST

I think what's happening is that if someone else pushes the story into voting and then you press the "Spam" button, it registers you as a -1. I don't have the Scoop code with me though.

[ Parent ]
That looks right (3.00 / 1) (#81)
by Rogerborg on Fri Mar 21, 2003 at 05:27:42 PM EST

It uses the same counter for spam and votes, but only collates spam votes up to the point where it goes to vote.  However, if someone hits spam after it's just gone to vote, it will indeed register it as a negative vote.  Nice spot.

"Exterminate all rational thought." - W.S. Burroughs
[ Parent ]

Man (4.33 / 3) (#101)
by trane on Fri Mar 21, 2003 at 10:00:34 PM EST

sounds like a ridiculously complicated and arbitrary algorithm to me.

[ Parent ]
Not that complicated. (none / 0) (#130)
by cdyer on Sat Mar 22, 2003 at 03:10:18 PM EST

It's really not so bad if you think about it. Basically all its saying is that if a certain percentage of people think it's spam, it gets booted to voting. The threshold is there because otherwise, if the first person who views the page says it's spam the percentage shoots to 100, and it gets sent to voting instantly with no chance for anyone else to veto that choice. Cliff

[ Parent ]
err... (4.83 / 6) (#43)
by pb on Fri Mar 21, 2003 at 10:28:42 AM EST

What's the point of lowering the threshold if there are no decent editorial suggestions? I've found that I get tons of editorial suggestions (and perhaps the most useful or damning ones) the minute I move a story to voting, and not before. And because of this, I don't feel bad going to sleep for 8 hours, and then coming back and fixing up the few trivial things people comment about.

That is to say, if more kuro5hin users took the Edit Queue seriously, I'd take it more seriously as an author as well.

Also, I see potential for abuse in removing the limit altogether; perhaps the Spam button should have a short memory, or be reset after each edit, so that a story that has been in the queue for an overlong period of time actually has a chance of being voted out. Then again, the 24 hour limit isn't so bad either; I don't mind having some sort of (large) upper limit to the time a story can be in Edit.
---
"See what the drooling, ravening, flesh-eating hordes^W^W^W^WKuro5hin.org readers have to say."
-- pwhysall
[ Parent ]

Perhaps a symptom of the problem (4.50 / 4) (#48)
by Rogerborg on Fri Mar 21, 2003 at 10:48:59 AM EST

People are wary of giving editorial comments because there of articles malingering in edit?  It's not considered worthwhile to comment on an article until it shows its mettle by going to vote?

I know that's how I feel sometimes, and it's a feeling that I'd like to experience less of.

However, if it's a sound article, and especially if you've posted a comment or two thanking people for editorial suggestions and committing to performing more before submitting, then I still think that it's unlikely to get pushed to vote even if you leave it for an extended period.

I'm suggesting just a gentle nudge towards handing the reins over to the readers a little sooner.  Baby steps.

"Exterminate all rational thought." - W.S. Burroughs
[ Parent ]

I second Rogerborg's take (none / 0) (#153)
by martingale on Sun Apr 20, 2003 at 11:08:11 AM EST

I find that for all intents and purposes, I don't make a practical distinction between stories in edit or in the voting queue, as far as reading and commenting goes. HOWEVER, I've practically stopped giving editorial comments, even though I used to a lot more. When it's in voting, it's too late and when it's in editing, there's no point.

It used to be that when I caught stories in editing, the nature of the two hour time limit meant that my suggestions would be acted upon (one way or the other), and if needed a small comment exchange would be performed. Now if I see a story in edit, I have no idea if it's been there since the middle of last night, or since the last five minutes. I don't know if the author is reading what I have to say, so I don't really bother to say much. If I do say something, the story might go to voting in the middle of the following night, and I haven't seen the latest version until it's too late to offer further edits.

Now that I'm thinking about this whole topic, the original two hour limit was definitely better imo.

[ Parent ]

Sorry (4.00 / 4) (#2)
by tmenezes on Fri Mar 21, 2003 at 07:12:46 AM EST

I couldn't resist pressing the SPAM button on this one.

Fine by me (4.66 / 3) (#3)
by Rogerborg on Fri Mar 21, 2003 at 07:20:58 AM EST

That button shouldn't be seen as topically judgemental, just an editorial "ready for vote" button.  Go hog wild!

"Exterminate all rational thought." - W.S. Burroughs
[ Parent ]

Was just a pun (4.00 / 1) (#36)
by tmenezes on Fri Mar 21, 2003 at 09:46:28 AM EST

Nothing against your article. Feeling silly today. Must be the weather.

[ Parent ]
Do you feel lightheaded? (5.00 / 2) (#40)
by Rogerborg on Fri Mar 21, 2003 at 10:10:08 AM EST

Is your vision dimming?

Got duct tape?

"Exterminate all rational thought." - W.S. Burroughs
[ Parent ]

Huh? (none / 0) (#139)
by woem on Sat Mar 22, 2003 at 10:13:19 PM EST

I think the button should be topically judgemental. If it means "ready for vote", it should be labelled as such. Labelling it "spam" implies that the button should be clicked if the article is spam.
:woemeowoemeowoemeowoemeowoemeowoemeowoem:
i either +1fp or -1. no exceptions. i ♥ turmeric.
the only class that should be discriminated against is the stupid.

[ Parent ]
The spam button. (4.57 / 7) (#5)
by Arkayne on Fri Mar 21, 2003 at 07:28:41 AM EST

Was a survey done here in Toronto a few years back about those intersection buttons.. you know the ones, that supposidly change the light in your direction if you press them.

Turns out they did nothing at all, and yet people swore that they changed the light faster and felt better using them.

I propose that the spam button does nothing at all - there's no wires connected behind that button, but that it's still effective because the people using it presume it's doing something.

Oh my! Really? (4.00 / 1) (#7)
by TheModerate on Fri Mar 21, 2003 at 07:37:02 AM EST

To think how many times I have hit the button since the light was so slow to turn.

"What a man has in himself is, then, the chief element in his happiness." -- Schopenhauer
[ Parent ]

They work (4.83 / 6) (#10)
by BadDoggie on Fri Mar 21, 2003 at 07:49:27 AM EST

The pedestrian buttons normally do work, but you would rarely notice if one was broken. They have little to no effect during daytime and peak traffic, depending on location, but at night, you they have clear effect.

Most lights have the night cycle programmed to a multiple cycle which has skip-steps built in. The light skips the secondary road change once or twice unless triggered by a car detector or a pedestrian button. It will always change once every four to five minutes, in case the buttons and sensors aren't working.

Also, many crossing signs don't change to "Walk" or whatever symbol your town uses unless the button is pressed. Again, during the day, most lights are programed to assume there are people there.

Now, about those "Door Close" buttons in the elevators...

woof.

"The line between genius and stupidity is very fine indeed, but you're so far away from the line that it doesn't matter." -- Parent ]

Door Close (4.83 / 6) (#14)
by tps12 on Fri Mar 21, 2003 at 07:57:34 AM EST

The worst thing about the Door Close buttons in elevators is the stupid etiquette that they create. If you're the one closest to the buttons in the elevator, then not leaning on Door Close is going to get you dirty looks from your fellow elevatees.

[ Parent ]
Better way. (5.00 / 10) (#18)
by BadDoggie on Fri Mar 21, 2003 at 08:06:04 AM EST

I like to amaze people by waiting for the right moment and then pointing at the Door Close button in a very menacing way. Sometimes I might even say something to the doors or panel. It almost always works.

Parents, of course, hate me for this because the little squirrels will forever be trying to do it themselves and having temper-tantrums if someone dares actually press the button. This is funnier still, because these are the same brats who whine, "But I wanted to push the button!"

woof.

"The line between genius and stupidity is very fine indeed, but you're so far away from the line that it doesn't matter." -- Parent ]

Touche'. (4.00 / 1) (#17)
by Arkayne on Fri Mar 21, 2003 at 08:02:29 AM EST

Alright, I submit to your obvious expertise with the whole street-crossing button thing. However my point still stands.

Even if the spam button did nothing from a coding point of view, it's still effective because people feel satisfied that they expressed an opinion by pressing it. This, in turn, does help to prevent responses that immediately nose-dive toward the "Hidden Comments" section.

[ Parent ]

Good point (4.00 / 2) (#20)
by Rogerborg on Fri Mar 21, 2003 at 08:22:40 AM EST

Would it be better if it said "Ready for voting" though?

"Spam" is pejorative, which might leave people more frustrated if they want to press it but don't because they're not sure if it might contribute to canning an article that they would otherewise abstain on.

"Exterminate all rational thought." - W.S. Burroughs
[ Parent ]

Actually... (4.00 / 2) (#49)
by BadDoggie on Fri Mar 21, 2003 at 10:54:41 AM EST

I think it would be best with "Send to Voting". Many things in Edit should be sent to voting, ready or not (I'm thinking of an ochre-coloured spice here).

OT: You've been rather prolific lately, Rogerborg. Better job? No job? All in all, I've liked most of your comments in the past few weeks; many are better than the articles to which they're attached. Keep it up!

woof.

"The line between genius and stupidity is very fine indeed, but you're so far away from the line that it doesn't matter." -- Parent ]

Many, many compiles to do (4.00 / 3) (#52)
by Rogerborg on Fri Mar 21, 2003 at 11:05:41 AM EST

That, and I'm positioning myself as the natural successor to rusty when the revolution comes.  Imagine getting paid to surf Kuro5hin.  Mmmm.

"Exterminate all rational thought." - W.S. Burroughs
[ Parent ]

Honestly. (none / 0) (#116)
by Arkayne on Sat Mar 22, 2003 at 06:32:29 AM EST

To be honest, I never read what the hell the spam button really does, so I can provide some insight into those seeing it for the first time.

I'm beginning to suspect otherwise, but I have always assumed that the button meant "This story is nothing more then mindless spam, please destroy and DO NOT send to voting."

I know I should read the help files, but I've never actually used the button. To me, that's the definition of "spam" and how I have interpeted how it works. If this ISN'T what it does, then yes, I fully agree that it needs to be renamed.

[ Parent ]

Further to my other reply (4.80 / 5) (#22)
by Rogerborg on Fri Mar 21, 2003 at 08:25:08 AM EST

What would you prefer on an elevator button?
  • Close door
  • Door is stuck open
"Close door" ("Ready for vote") is an affirmative action.  "Door is stuck open" ("Spam") is a negative criticism.   Let's help people to feel good about contributing.

"Exterminate all rational thought." - W.S. Burroughs
[ Parent ]

Yes (none / 0) (#85)
by celeriac on Fri Mar 21, 2003 at 05:56:34 PM EST

However in most cases if you walk up to an intersection while the light is green in your direction and press the walk button, you won't get a go-ahead to walk until the next green light. Some of them are better designed and allow you to walk immediately if the light will remain green for long enough.

[ Parent ]
Not so (5.00 / 15) (#16)
by Rogerborg on Fri Mar 21, 2003 at 08:01:05 AM EST

The spam button at the very least demonstrably remembers that you've pressed it.  Clearly rusty is collating data on who's voting down the submissions from his "turmeric" account, for when the revolution comes.

"Exterminate all rational thought." - W.S. Burroughs
[ Parent ]

First Time! (3.33 / 3) (#96)
by Agent1 on Fri Mar 21, 2003 at 08:15:41 PM EST

I actually laughed at a comment to a meta story ;)


-Agent1
"Thats the whole point of the internet, to slander people anonymously." - Anonymous
[ Parent ]
Heh (none / 0) (#126)
by theantix on Sat Mar 22, 2003 at 12:37:11 PM EST

When I wrote the code for the spam button, I actually proposed that the button do nothing at all because it is easy (duh) and the main function is accomplished (people stop whining).  However rusty came up with a really cool method for getting the spam button to work, the % viewed and vote threshold combination, so it actually does something.  Of course if you set the % high enough it still does nothing and works as you suggest, but I'm not sure how he has it set up exactly.

-- YOUR TERROR ALERT IS: red

--
You sir, are worse than Hitler!
[ Parent ]

Rename it (4.95 / 20) (#6)
by AndrewH on Fri Mar 21, 2003 at 07:34:22 AM EST

The name “Spam” discourages people from using it on all but the most worthless stories. Call it “Send to voting” or something like that.
John Wilkes Booth, Lee Harvey Oswald, John Hinckley Jr — where are you now that we need you?
Rename indeed... (4.60 / 5) (#30)
by cvou on Fri Mar 21, 2003 at 09:10:51 AM EST

Can we call it the turmeric button? :-/

[ Parent ]
"Ready for voting"? (n/t) (4.00 / 1) (#51)
by Rogerborg on Fri Mar 21, 2003 at 11:03:51 AM EST


"Exterminate all rational thought." - W.S. Burroughs
[ Parent ]

I say send. (4.66 / 3) (#66)
by cdyer on Fri Mar 21, 2003 at 03:05:29 PM EST

As someone said above, SPAM is a value judgement on the story, while a better label would describe the action undertaken. The elevator analogy was "The door is stuck open" vs. "Close Door." As such, I think "Send to voting" is better than "ready for voting." Though it lacks the pejorative meaning of spam, it is still a comment on the state of the article (like a button that says "No one is standing in the elevator doorway now.") It still doesn't explain what the button does. The reasons to send to voting could be many. Spam, crappy story, no editorial problems, and so on. But it's not always because the story is ready. So I say "SEND!" Cheers, Cliff

[ Parent ]
Spam means advertisement (4.66 / 3) (#57)
by lauraw on Fri Mar 21, 2003 at 12:31:38 PM EST

I agree: rename it or at least document it better. I've never pressed the Spam button, because without clear documentation I've been assuming that it means "This post is a commercial ad" (or other sort of worthless promotion). I certainly haven't assumed that it means "Ready for voting" or even "This story sucks out loud", because that's just not what the word "Spam" means to me, and to most net users I suspect.
-- Laura

[ Parent ]
I concur (4.66 / 3) (#70)
by MajorMajor on Fri Mar 21, 2003 at 03:29:11 PM EST

I that the biggest problem with K5 is the latency between initial post and getting to the front page/section, especially if the article is about current events. Often the article is already quite good after having been sat in the edit queue, readers should be able to vote for it to be pushed into the voting queue. They should also be allowed to vote for/against it being posted to the section/FP as well during editing, as whether the substance of the article is worthy should be independent of its presentation.

I find that by the time an article has reached the front page, I have already read 90% of the comments on it, and usually the changes are very minor. I guess this is an indication of the general quality of initial posts.

[ Parent ]

Hurrah (4.00 / 1) (#108)
by Rogerborg on Sat Mar 22, 2003 at 02:46:14 AM EST

I haven't said this yet, but it bothers me as well.  Shouldn't we as a community aim to be commenting topically on articles mostly after they have been published?  Let's push those suckers through.

"Exterminate all rational thought." - W.S. Burroughs
[ Parent ]

Bad articles deserve to fail (4.33 / 6) (#8)
by HidingMyName on Fri Mar 21, 2003 at 07:42:37 AM EST

Are we losing too many good comments attached to bad articles?
This sounds like the road to mediocrity to me. If an article is really bad, don't waste your time making elaborate comments on it, say something brief and vote it down. This idea of accepting shoddy articles because the comments are so good is weak. If the comments are that good, make them articles in their own right.

Sorry, I was unclear (4.75 / 4) (#11)
by Rogerborg on Fri Mar 21, 2003 at 07:51:41 AM EST

(and I guess I should have waited for this comment and made edits based on it ;-) )

I didn't mean to imply that we should keep poor articles just because they have good comments (even though that's what autopost does).  I just meant that we could perhaps looks at ways to get articles into (and therefore through) voting in a more timely fashion, so that there's less opportunity to get involved in a discussion attached to a doomed article.

"Exterminate all rational thought." - W.S. Burroughs
[ Parent ]

If good discussin is occuring (4.00 / 1) (#73)
by Mysidia on Fri Mar 21, 2003 at 03:39:15 PM EST

Then regardless of the quality of the article, the good discussion really shouldn't be hidden.

In an ideal world, articles of such a nature that would be voted down from the section pages would be posted to some other section like "Diaries" or "Misc Discussions" or something

The problem is there's no real way to measure the quality of a discussion: flame wars/crapfloods can generate as many posts as worthwhile discussions can, yet be utter garbage deserving to be headed to oblivion.



[ Parent ]
That's not the problem (4.00 / 1) (#77)
by DrJohnEvans on Fri Mar 21, 2003 at 05:11:38 PM EST

How many worthless articles will produce good discussion? Would the percentage be high enough to bother implementing such a feature?
The problem is there's no real way to measure the quality of a discussion
K5 has a full-featured comment rating system in place, the purpose of which is to do just that. The problem is people's reluctance to rate comments. Convention seems to dictate that people either give 5s or 1s with nothing in between, and only to choice comments, rendering the rating system largely useless. But that's all been said before. My point is that the system of quality measurement does exist; we, as a community, aren't taking advantage of it.

Indeed, Scoop takes comment ratings into account when calculating the postability of a story coming out of the voting queue. It wouldn't be difficult to implement something similar for the editorial queue; but the problem of preserving the comments without the article remains.

--
Proud member of the K5 Axis of Evil since 2002.
[ Parent ]

K5 Problems Skew Ratings (none / 0) (#94)
by HidingMyName on Fri Mar 21, 2003 at 07:23:38 PM EST

Indeed, Scoop takes comment ratings into account when calculating the postability of a story coming out of the voting queue. It wouldn't be difficult to implement something similar for the editorial queue; but the problem of preserving the comments without the article remains.
There are many users (abusers) with multiple accounts, who use them to self aggrandize (by modding up their own comments) and can use that to game the system to make borderline stories pass (giving all comments on a story a +5 rating to try to push it over the edge). This however is a problem with K5 administration/structure.

Another problem with K5 is that it lacks a really good way to navigate old content, so it tends to have a very short term memory, which is tending to make it a write-only forum for long term archival purposes.

[ Parent ]

you've some good points (none / 0) (#156)
by martingale on Sun Apr 20, 2003 at 11:48:24 AM EST

The short term memory issue is imho crucial to much of the current k5 failings. I'm not sure how to fix it, but it's definitely a searchability/comment awareness issue.

k5 sucks for managing your own comments. If somebody responds to one of your comments tomorrow, you might miss it. Even if you don't miss it, you might not be aware that an older comment of yours, whose reply you've already read, may have spawned an interesting subthread after you thought you'd read it all.

The end results of the personal comment management difficulties are that it's best to end discussions relatively quickly, and everyone except newbies knows that commenting on old stories is a waste of time.

Now suppose k5 had a sophisticated personal comment management system. For example, your comments could be sorted by activity within the thread it lives in. So if the next day, there's an active discussion related to the topic you commented upon, k5 highlights that comment. Now you might go back to that discussion and comment some more on it. Other participants would see this when they are managing their own comments, and so on. This sort of help would go a long way towards fixing timezone problems and would give the site a much longer memory. This would truly make k5 different from that other site.

[ Parent ]

Good comments (4.00 / 4) (#12)
by Slothrop on Fri Mar 21, 2003 at 07:53:24 AM EST

But I think that it needs to be reduced all around.  12 hours for the edit queue, a lower spam threshold, and the spam button renamed to send to voting. To be honest, I'm more concerned with the 'using the edit queue to publish and then dissapear for ever' issue than any of the others.  Your solution sounds ok, but seems over complicated.  Easier to tweak the parameters of  the current algorithm than to define a new one and test it.
__________________________
Provide, provide!
Umm (4.00 / 1) (#19)
by Rogerborg on Fri Mar 21, 2003 at 08:06:22 AM EST

Well, I substantially agree with:

A lower spam threshold
Renaming the spam button.

My concern with an unconditional lower threshold is that it punishes good faith articles that are making regular but perhaps not obvious edits.  Of course, it's still open to abuse by turmeric people making faux edits to their articles, but under no circumstances is it worse than the current method, and I'm interested in not punishing active contributors along with the abandoners.

"Exterminate all rational thought." - W.S. Burroughs
[ Parent ]

What I would do (5.00 / 18) (#31)
by Ranieri on Fri Mar 21, 2003 at 09:14:57 AM EST

I would like to propose two extremely simple modifications:
  1. Change the label on the button
    Labelling a thing as "Spam" has a rather high threshold for many people. I feel the label should reflect the actual function of the button. I therefore suggest "Push to Vote" or something similar.
  2. Show the elapsed time since the last edit
    People can then make their own decision regarding whether the article is being edited, without having to go through the trouble of checking comment timestamps. Of course it could be readily abused by doing and undoing insignificant edits, a strategy we shall call "The Penelope Exploit" in honour of Ulysses' wife. Such an abuse would however readily be noted by our vigilant queue watchdogs.
The first modification is what I feel would have the biggest effect, and could be implemented immediately without code changes. There is no reason for not doing it.
--
Taste cold steel, feeble cannon restraint rope!
Can I suggest (4.75 / 8) (#35)
by Rogerborg on Fri Mar 21, 2003 at 09:32:49 AM EST

"Ready for vote" ?  If we're going to be positive rather than pejorative, then let's do it with gusto.

"Exterminate all rational thought." - W.S. Burroughs
[ Parent ]

Sure, sounds neat. (4.25 / 4) (#37)
by Ranieri on Fri Mar 21, 2003 at 09:48:02 AM EST

"Ready for vote" is exactly the sort of label we need.
--
Taste cold steel, feeble cannon restraint rope!
[ Parent ]
I like that. (4.00 / 3) (#64)
by porkchop_d_clown on Fri Mar 21, 2003 at 02:04:10 PM EST

It would help a lot to know whether the author is actually making changes or just dumped something in the q and went to bed.


--
You can lead a horse to water, but you can't make him go off the high dive.


[ Parent ]
Totally agree (4.50 / 4) (#100)
by rhyax on Fri Mar 21, 2003 at 09:39:42 PM EST

since this feature was never explained I personally would not use the spam button. I don't consider spam and trolling as the same. To me spam is advertising, for profit crap. if someone leaves a very opinionated story in the queue forever knowing that it would not get voted up i would not call that spam. i might however reload the page a few times to see if the person gets yelled at.

This is why not explaining how a system works is a bad idea. By not understanding the system I have inadvertently voted for the story I think is poor.

I think the button should definitely be renamed to "move to voting" or something along those lines, I would also like page-views not to be an issue. Many people reload pages, especially in the queue, those should not count as votes.

[ Parent ]

Same old same old. (4.00 / 2) (#33)
by DarkZero on Fri Mar 21, 2003 at 09:25:42 AM EST

I really haven't noticed a difference since the Spam Button was put in. Despite the fact that there's no anti-Spam Button to cancel out the spam votes like there is with the comment rating system, it still suffers from pretty much the same problem: very little gets hidden. You practically have to post ASCII art of bestiality porn for people to press the Spam Button. If it skirts the line between what we commonly think of as spam and just an article that is definitely going to go from 0 to -20 in .06 seconds when it hits the moderation queue regardless of edits, it sticks around for 24 hours. And now that I think of it, that actually goes against the entire point of the moderation queue. If people don't want something on the site, i.e. they will downvote it the second it hits the moderation queue no matter what, it shouldn't be on the site.

Personally, I would bring the amount of time in the edit queue down to twelve or fourteen hours, lower the Spam Button threshold a little, and change the name of the Spam Button so that people don't think that they're breaking the rules by sending a hopeless article to voting. I should qualify that, though, by saying that I think it's all pretty minor. I don't see the edit queue as a big problem or anything. The comment rating system, on the other hand... now that I could tear apart like a rabid dog at a moment's notice. ;)

Sounds about right (5.00 / 1) (#44)
by Rogerborg on Fri Mar 21, 2003 at 10:30:27 AM EST

I think we probably agree that the aim isn't to drop more articles, just to decide on them faster and stop them hiding in edit as a substitute for voting.  I find myself spending more and more time in the edit queue and less and less on the front page and sections, and that can't be right.

And yes, I agree that it's pretty minor.  I don't want anyone to think that this keeps me up at night.  It's just polite to give rusty some feedback and recognition of a job well done every so often, bless him.

"Exterminate all rational thought." - W.S. Burroughs
[ Parent ]

I agree completely. (nt) (none / 0) (#91)
by DarkZero on Fri Mar 21, 2003 at 07:00:44 PM EST



[ Parent ]
I find the comment rating ironic (4.00 / 1) (#93)
by godix on Fri Mar 21, 2003 at 07:19:44 PM EST

I've seen comments with a couple 0's, tons of 1's, and one person rating 5. Those 1 ratings pull the average up enough that the single 5 vote keeps the comment from being hidden. Whenever a non-TU's voting something as crap keeps the comment visible I have to laugh.


Love - A temporary insanity curable by marriage.
- Ambrose Bierce, The Devil's Dictionary
[ Parent ]
Please do your math (none / 0) (#149)
by tetsuwan on Thu Mar 27, 2003 at 11:16:46 AM EST

A "1"-vote will not affect the average to the advantage of the one posting the comment. If rounding is left out of the picture, a comment with six zeroes, a hundred ones and one 5 will still have an average below 1. Twos on the other hand, are supporting the visibility of a comment.

[ Parent ]
Move to vote (4.83 / 6) (#47)
by krek on Fri Mar 21, 2003 at 10:47:06 AM EST

I, personally, would only push the 'Spam' button in cases of dengenerate crap, while I would just let a poor article sit.

Perhaps 'Spam' is a bit harsh of a name, I have come to understand that pushing the 'Spam' button does not actually register a vote of any kind, it just pushes the article closer to the edit queue, in that spirit, should it not have been named 'Move To Vote'. This would let people use it who feel that an article is done in the edit queue and ready for voting, regardless of whether or it is a good article or a crap article.

I Concur. (4.00 / 3) (#53)
by porkchop_d_clown on Fri Mar 21, 2003 at 11:19:00 AM EST

The name of the button is unfortunate, "Push into Voting" or "Ready for Voting" would be much better.


--
You can lead a horse to water, but you can't make him go off the high dive.


[ Parent ]
The "SPAM" buttom (4.66 / 6) (#54)
by corian on Fri Mar 21, 2003 at 11:20:45 AM EST

I admit not to reading all the 'rules' closely, but it looks like I've been completely in the dark as to the meaning of the "SPAM" button. I've assumed that it's a way to flag articles that are, somehow, inappropriate -- advertising, random text, pornography -- things that would never get approved as an article, and thus should be discarded rather than voted on. You know, spam-like articles. If it's not a incator for spam, then it shouldn't be labelled "SPAM."

Good point (3.00 / 1) (#55)
by Rogerborg on Fri Mar 21, 2003 at 12:14:37 PM EST

That's not how I understand it, but I'm not in the non-existent cabal, so what do I know?  I'm glad someone raised this as a possibility.

"Exterminate all rational thought." - W.S. Burroughs
[ Parent ]

Ditto ...... n/t (2.00 / 1) (#67)
by ageing hippie on Fri Mar 21, 2003 at 03:10:35 PM EST


------------------------
Fool me once shame on you, Fool me twice shame on me
[ Parent ]
I prefer the new method (4.40 / 5) (#56)
by bluefusion on Fri Mar 21, 2003 at 12:25:10 PM EST

Being new to K5, the 24-hour queue is one of the major things I love about it. Like others have stated, being able to go to sleep with something in the queue is very beneficial. It also keeps a story hanging around longer, collecting more intelligent feedback. Two hours? That's nowhere near enough. So what if the articles aren't posted immediately--this is a news site in some aspects, but we don't need everything to be timely. Give me well-written articles over three-word headlines and two paragraphs of utter crap any day.
--------------------------------------

"Real? What is real? If you are talking simply about what you can see, taste, touch, hear, then 'real' is simply electrical signals interpreted by your brain."

Remember (4.66 / 3) (#60)
by Rogerborg on Fri Mar 21, 2003 at 12:47:20 PM EST

That we did just fine with a 2 hour hard limit for quite some time.  I'm sure real old timers will remember it being 15 minutes or 7 days or something.

You learn to live with the limits.  When we had a 2 hour limit, article authors had to do their own research, and if they were interested in doing edits, they had to do so fairly rapidly.  Also, there was more of a sense of urgency to provide editorial comments.

I'm not advocating that we go back to that, just documenting that it worked well, but in a different way.

"Exterminate all rational thought." - W.S. Burroughs
[ Parent ]

Well (4.00 / 1) (#71)
by Mysidia on Fri Mar 21, 2003 at 03:32:29 PM EST

We did just fine before K5 had an Edit Queue too, but having it _did_ help, even if it held crap in the mod queue longer; perhaps, it reasons that being able to have more time in edit will help some too.



[ Parent ]
It does help (none / 0) (#84)
by Rogerborg on Fri Mar 21, 2003 at 05:50:12 PM EST

And longer would help more, for those authors that use it.  However, we may be creeping towards the point where it's being used as a dumping ground, which (IMHO) cheapens the process.

No biggie, just perhaps there's room for a little more tweaking.

"Exterminate all rational thought." - W.S. Burroughs
[ Parent ]

Time zones and edit queue time limits (4.66 / 3) (#72)
by MajorMajor on Fri Mar 21, 2003 at 03:38:12 PM EST

I think we should be careful about setting timeouts on the edit queue shorter than 24 hours. If someone posts an interesting article from a TZ where there are few interested K5 readers, they may not get sufficient feedback in time.

On a separate note, I sometimes find it frustrating when I get to an article which has been posted maybe 12-18 hours ago, post a comment in the full knowledge that nobody will read it because the article has been left for dead.

Which one is it going to be then? (none / 0) (#110)
by Rogerborg on Sat Mar 22, 2003 at 02:52:39 AM EST

I agree with both of your points, but I don't see how we can have it both ways.  The number of articles that are still being actively edited after 24 hours seems very small compared to those that are just abandoned, and every one of those that we read disinclines us to trust the good ones.

"Exterminate all rational thought." - W.S. Burroughs
[ Parent ]

why don't we put all rejected stories in the diary (4.58 / 12) (#75)
by circletimessquare on Fri Mar 21, 2003 at 04:23:52 PM EST

why don't we put all rejected stories in the poster's diary?

no really... that way, if it becomes part of your "permanent record" then you will think twice about posting spunk

also, since most comments on bad submissions are "put it in your diary", wouldn't this be putting our money where our mouth is?

i have posted stories i then put in my diary when they were rejected (well, only one) but the point is that most anyone who has emotional attachment to their submission would appreciate this feature

seems win-win all around

The tigers of wrath are wiser than the horses of instruction.

I agree (4.50 / 2) (#98)
by jonathanwilson on Fri Mar 21, 2003 at 08:48:05 PM EST

My recent story "Persecution of Christians" which was dropped, had a lot of great comments and discussion.  I put the story in my diary, but how in the world do you get all the comments to stay?

[ Parent ]
there is a way to do this (none / 0) (#102)
by trane on Fri Mar 21, 2003 at 10:10:28 PM EST

but I've forgotten exactly. I think you just substitute "comments" for "story" in the original story's url, and you'll get all the comments (but not the story).

[ Parent ]
everyone look at the above comment (4.00 / 1) (#136)
by circletimessquare on Sat Mar 22, 2003 at 05:43:46 PM EST

IT IS A VERY GOOD POINT

all of those comments and discussion are the heart of kuro5hin... why throw it all away on a good story where a lot people talked about it, just because the story went belly up at 48 hours with 79 votes or something like that?

The tigers of wrath are wiser than the horses of instruction.

[ Parent ]

My opinion... (4.50 / 2) (#76)
by Menard on Fri Mar 21, 2003 at 04:53:52 PM EST

I think that the edit queue is fine as is. People complain of long times with no edit, but things like that could just as easily be people wh post a story to editing, and go away to sleep/do errands/etc. for a few hours, and then edit in response to comments when they return. I've seen this happen a few times.

Along with this, your last point is very relevant. The quality of articles that actually get posted has not gone down, and, maybe more importantly, it's not like there are so many stories in the edit queue that it becomes difficult to sort through them all. I've never seen more than ten stories in the edit queue at one time, and often less than that.

Finally, a case could be made that the extra editing time has contributed to better grammar/spelling, as both seem to be very good in posted stories, which is kind of unexpected for user-created content. I'm not really trying to make that case, as it's one that would be difficult to conclusively prove, but I would suggest that, if we were going to go back to a straight time-limited queue, that the limit be longer than two hours, say six, or possibly even eight.

Oh, one final thing, the 24 hour queue allows the user to pick an optimum posting time for their story to voting, which is nice.

The Spam button should be for indicating spam (4.00 / 3) (#80)
by Mysidia on Fri Mar 21, 2003 at 05:25:50 PM EST

Not for saying you want to force the author to move it to vote now (though they could just 'cancel' the article and repost it as edit if they saw it had gone to vote before they wanted)

IMO The spam button shouldn't cause it to go to err, vote if threshold is reached, but 'kill the story'

And it shouldn't be 'button pushes' to 'page views' , but rather 'button pushes' to 'page views by unique users'

So viewing a page counts as a default abstain vote and makes it harder for it to be considered spam unless you hit that push button.

Voting history should then be active for pages: if you hit 'spam' (Meaning, hide the story, it's spam), then you should appear as having voted that (accountability). Perhaps there should be a "not spam" button too (beside it)

Oh, this raises an interface question for ordinary voting: when voting for articles, why have a drop-down list rather than just three push buttons?

It seems like it would be faster to "Just click the +1 button" than bring down the drop down list, and it would be more consistent with the interface for the "Spam button" (single button click for vote)

Deciding what to vote takes longer, but it occurs that bringing down a drop down list, clicking an entry, and _then_ hitting a button involves more actions and takes longer (Ie: it's "harder" or "more of an effort" than simply hitting a push-button-vote)

Based on the principle that interfaces ought to be consistent where that's reasonable, the Spam button ought to be a drop down list defaulting to "Not Spam" and having a "Spam" entry, beside vote button, OR the normal vote submission form should be 3 push-buttons side by side, etc, etc, regularity :)



-Mysidia the insane @k5
The only problem with the system as it stands (4.40 / 5) (#82)
by RobotSlave on Fri Mar 21, 2003 at 05:35:08 PM EST

...is that the "spam" button is horribly mislabeled. It should be changed to "Move to Voting."

The word "spam" carries a hell of a lot of baggage. Those who don't quite understand how this newfangled system works are reluctant to render a verdict of "Spam" for trifles like an article which lingers too long in the edit queue, or suffers the burden of indifferent prose.

If everyone realized that the button simply meant "Move To Voting Queue" rather than "Brand Author As A Blackguard And Most Despicable Cur," the system would probably be much more effective.

What the code says (4.20 / 5) (#83)
by Rogerborg on Fri Mar 21, 2003 at 05:35:42 PM EST

Stories are pushed from edit to vote if:
  1. They receive at least 100 spam votes.
  2. and at least 25% of page views have resulted in a spam vote.
Spam and voting use the same counter, but a spam vote shouldn't count as a -1 because the count is reset when it goes to vote.  The exception is when someone is viewing the story while it's still in edit, but hits "Spam" after it's been pushed to vote, in which case a -1 vote is registered.  That explains the occasional report of a spam button click being shown as a -1 vote.

Interesting implication of the above: the more page views that a story gets, the harder it is to hit the 25% threshold and push it to vote.  So coming back to check for edits on an abandoned article helps to preserve it.  That explains why articles seem to either be dumped very quickly, or go all the way to 24 hours.

Thanks to cdyer and nxor for pointing this out.  I've verified it in the CVS scoop code as of 21-March-2003.

"Exterminate all rational thought." - W.S. Burroughs

And incidentally (none / 0) (#88)
by Rogerborg on Fri Mar 21, 2003 at 06:23:02 PM EST

That means that articles pushed to vote are technically socred at -100 just before being reset, and then require another net -20 to be dumped proper.

"Exterminate all rational thought." - W.S. Burroughs
[ Parent ]

Not really (none / 0) (#123)
by cep on Sat Mar 22, 2003 at 12:07:53 PM EST

... because the people who have pushed the spam button can vote again.

[ Parent ]
Sorry, clarification (none / 0) (#146)
by Rogerborg on Wed Mar 26, 2003 at 11:16:22 AM EST

"can", yes, but also "have to".

"Exterminate all rational thought." - W.S. Burroughs
[ Parent ]

Not quite page views (4.00 / 1) (#128)
by panner on Sat Mar 22, 2003 at 02:26:35 PM EST

Interesting implication of the above: the more page views that a story gets, the harder it is to hit the 25% threshold and push it to vote.

Not true, because page views isn't exactly what gets measured. The spam button uses the story_views data, which is the same used to do new comments. So the button is looking for some percent of users that viewed the story to push the button, not for some percent of page views to have button pushes. Once you view the story once, repeated viewing will have no effect on the count.

I'd say the behavior you noted, with stories either getting pushed quickly or taking the full time, is related to the mechanism working. The true crap gets pushed out quickly, and the stuff that's not crap sits until the author pushes it out, or the time limit is reached.



--
Keith Smiley
Get it right, for God's sake. Pigs can work out how to use a joysti
[ Parent ]
The Unwritten Rules (4.85 / 7) (#87)
by coderlemming on Fri Mar 21, 2003 at 06:16:35 PM EST

Ok, so I'm seeing a lot of little rules going around that everyone (but me) seems to know, but that aren't in the FAQ. It's really not fair to scream and shout at an author because they've broken rules that they didn't even know existed. Let's collect them in one place and either add this to the help section, or do something else useful with it. Please feel free to add your own guidelines as replies. Remember what it was like when you were brand new to the site, and try to list all the things you've learned that aren't in the FAQ.

I should note that I, myself, have been interested in submitting, but I've been discouraged by not knowing the way things work yet. People will RTFM, but we have to give them a manual!

Guidelines I've picked up so far:
  1. The "spam" button would more descriptively be called the "move to voting" button.
  2. When posting a story in the edit queue, either watch it like a hawk for comments, or post a comment with your intention to edit after you go to sleep, after class, or whatever the case might be.
  3. Fiction should be fully complete work, the edit queue is not a writing workshop. (This may seem obvious, but it's not stated anywhere...)
  4. There is a 24 hour limit on the edit queue (I didn't know this!)
  5. You should respond to acknowledge all editorial suggestions/corrections when you put them into the story. (right?)
Anything else? The more out in the open these guidelines become, the more easily new members will be able to join the community.


--
Go be impersonally used as an organic semen collector!  (porkchop_d_clown)
Seems good. (5.00 / 2) (#89)
by Menard on Fri Mar 21, 2003 at 06:33:03 PM EST

I personally disagree with number three, editing is editing, and I see nothing wrong with someone requesting help with a story there.

As for five, I think this is definately something you should do, credit where credit is due and all that. Beyond just giving credit, it could help your story in voting. If I'm on the fence about voting for something or killing it, I'll usually check the editorial comments to see if the authour responded to editing suggestions. This doesn't mean implementing everything, but rather implementing things you agree with, and also stating why you disagree with the things you don't use. I don't know where other people look at this in voting, but I'd assume that at least a few do so. Personally, I also consider it good form to give a five to someone whose edit you used.

[ Parent ]

number 5 (none / 0) (#104)
by coderlemming on Sat Mar 22, 2003 at 01:12:40 AM EST

Ok, well, that was me paraphrasing an editorial comment on this story that I read. So obviously, not only are the rules unwritten, they're not even agreed upon. What with the opinion floating around that fiction doesn't even belong on k5... it would seem that, if your fiction piece is not at least well thought out and reviewed/revised a bit before hitting k5, it might get dumped because people don't want to try to save it.


--
Go be impersonally used as an organic semen collector!  (porkchop_d_clown)
[ Parent ]
I agree that there's no rule about this (5.00 / 1) (#107)
by Rogerborg on Sat Mar 22, 2003 at 02:42:15 AM EST

Except the realpolitik rule that submitting a work that's not substantially completed will:
  1. Get you dumped.
  2. Prejudice more people against the fiction section.
That's fair enough from the point of view of the author (especially the Johnny Noposts that tend to wander by and braindump into fiction) but I've got a soft spot for the section, and would like to see more completed fiction get voted up there.

"Exterminate all rational thought." - W.S. Burroughs
[ Parent ]

A few more (4.66 / 3) (#90)
by godix on Fri Mar 21, 2003 at 06:57:58 PM EST

don't scream at people giving you editorial advice. The person was nice enough to try and give you help, don't make them regret it. I'll vote down a story where the author bitchs at editorial comments even quicker than I'll vote down goatse.cx MLP's.

Anything that is not the story itself belongs in a comment. The time you'll be away and not making edits, any reference to story voting or comment rating at all (unless it's a Meta story), saying hi to your mother, whatever. If it isn't part of the story it belongs in a comment.

Articles that have been pulled and resubmitted should have two things: A link to comments in the first submission and enough changes to actually warrent pulling the story to begin with.


Love - A temporary insanity curable by marriage.
- Ambrose Bierce, The Devil's Dictionary
[ Parent ]

pulled stories (none / 0) (#103)
by coderlemming on Sat Mar 22, 2003 at 01:07:39 AM EST

Ok, while we're on the subject... how exactly does pulling a story affect the comments and the story's existence? Is it still in the DB, just unlinked? How about stories that are dumped, are they still around, or are they gone forever?


--
Go be impersonally used as an organic semen collector!  (porkchop_d_clown)
[ Parent ]
The comments persist (none / 0) (#106)
by Rogerborg on Sat Mar 22, 2003 at 02:38:50 AM EST

And can be retrieved with the story URL, with "story" replaced with "comment".  E.g. for this story:

http://www.kuro5hin.org/story/2003/3/21/7140/67328 is the story

http://www.kuro5hin.org/comments/2003/3/21/7140/67328 is the comments.

If you dump and resubmit, it's only polite to put a link to the old comments.  OTOH, the people doing a dump-and-resubmit are generally the ones least likely to know this, so it's usually up to one of the comment submitters to provide the link.

"Exterminate all rational thought." - W.S. Burroughs
[ Parent ]

the link... (none / 0) (#124)
by coderlemming on Sat Mar 22, 2003 at 12:13:19 PM EST

How would a comment poster get the link if it was dumped? Sorry to be so picky, but these are the kinds of questions newbies might really want to know, but never find the answer to (i.e. me).


--
Go be impersonally used as an organic semen collector!  (porkchop_d_clown)
[ Parent ]
After it was dumped (5.00 / 1) (#137)
by godix on Sat Mar 22, 2003 at 07:13:44 PM EST

How would a comment poster get the link if it was dumped?

I can think of three ways:

  1. A poster noted the URL before the article was dumped.
  2. The author can still view dumped stories, they can provide a link to the comments (which is what I was suggesting).
  3. You go look up a specific posters comment in a story. For example, if I recall you made a comment in a dumped story I could go look at your posting history. When I see the comment I remember I just slightly modify the URL of it to get the rest of the comments. The modification is to remove the (number here)#(number here) part at the end of the URL.
Examples:
www.kuro5hin.org/comments/2003/3/21/7140/67328/124#124 is your comment
www.kuro5hin.org/comments/2003/3/21/7140/67328 is for all comments
www.kuro5hin.org/story/2003/3/21/7140/67328 is the story itself.

Notice the only changes are if it's Story or Comments and if there's a bit at the end of the URL referencing a specific post or not.


Love - A temporary insanity curable by marriage.
- Ambrose Bierce, The Devil's Dictionary
[ Parent ]

Doesn't the original poster... (none / 0) (#140)
by roiem on Sun Mar 23, 2003 at 03:28:40 AM EST

...get a link to the story? I seem to recall that the original poster can see the story even after it was dumped. Am I hallucinating? (Never been a submitter myself, so I don't know this first hand, but I think it was somewhere in the FAQ)
90% of all projects out there are basically glorified interfaces to relational databases.
[
Parent ]
candidate for ui change? (none / 0) (#155)
by martingale on Sun Apr 20, 2003 at 11:28:16 AM EST

Sounds to me like this is a candidate for a slight UI change when dumping stories and resubmitting. Maybe a button [link to old comments] ?

[ Parent ]
That's half the fun, though... (4.00 / 1) (#113)
by kpaul on Sat Mar 22, 2003 at 02:59:01 AM EST

finding out the 'rules' as you go along.

heck, some nights i even make up my own rules. ;)


2014 Halloween Costumes
[ Parent ]
Oh yes (none / 0) (#118)
by Rogerborg on Sat Mar 22, 2003 at 09:39:12 AM EST

We all remember Nekkid Webcam Twister Tuesday.  If anyone doesn't remember that, seek therapy.  You are experiencing deep repression.

"Exterminate all rational thought." - W.S. Burroughs
[ Parent ]

before my time... (none / 0) (#121)
by kpaul on Sat Mar 22, 2003 at 10:51:15 AM EST

Alas, that incident must've happened before I got here. Any details?


2014 Halloween Costumes
[ Parent ]
Hey, it was YOUR idea (none / 0) (#133)
by Rogerborg on Sat Mar 22, 2003 at 03:54:03 PM EST

Don't go eschewing it now, just because of that nasty little court case.  As far as I'm concerned, you were a victim of politics.  That judge had been bought.

"Exterminate all rational thought." - W.S. Burroughs
[ Parent ]

not really (5.00 / 1) (#125)
by coderlemming on Sat Mar 22, 2003 at 12:15:25 PM EST

If I have to put that much work into just figuring out how this place works, it really makes me not that interested in contributing. And if new people are discouraged from contributing, this place will stagnate with a bunch of posts from the same core of authors.


--
Go be impersonally used as an organic semen collector!  (porkchop_d_clown)
[ Parent ]
Or... (none / 0) (#127)
by kpaul on Sat Mar 22, 2003 at 01:39:28 PM EST

only a dedicated handful will put enough into it and K5 will end up with a lot of good stories instead of floods of mediocrity (which I'm guilty of a lot...) in the queue.


2014 Halloween Costumes
[ Parent ]
Minor suggestion (4.60 / 5) (#92)
by godix on Fri Mar 21, 2003 at 07:13:22 PM EST

I think we need two buttons in edit mode. A spam button (renamed as others suggested) AND a dump it button. I personally only hit spam when I don't want to see the article again but the way things are now I must go look at the article at least once more to actually vote against it. My goal isn't to move the article to voting, it's to get rid of the damned thing. Leave a 'move to voting' button for people who have skill in English (in other words, not me) to use on articles without errors. Add a 'kill this damned thing' button for those who think the article is beyond redemption.


Love - A temporary insanity curable by marriage.
- Ambrose Bierce, The Devil's Dictionary
Which makes sense because (4.00 / 1) (#109)
by Rogerborg on Sat Mar 22, 2003 at 02:50:11 AM EST

When an article is pushed to vote, it's because it's received a score of at least -100 (which is then reset to 0 on entering voting) from people judging it to be spam.

Granted that's an unbalanced gross, but why are we throwing away all those opinions when it enters vote proper?

"Exterminate all rational thought." - W.S. Burroughs
[ Parent ]

so a pre-voting voting? (none / 0) (#151)
by Vellmont on Sat Mar 29, 2003 at 04:04:26 AM EST

Hmmm.. well I agree the spam button is miss-named, but I don't really see much need for two buttons. The spam button should maybe be re-named to spam/troll/garbage button, which seems to be the purpose it now serves. Furthermore if enough people hit this button, it should go bye-bye, and not to voting.

It seems to me the purpose of the editing queue is for people to get help in making an article better.The article sucking isn't a valid reason to kick it out of the queue. The only valid reason for kicking something out of the edit queue is if it's abusing the queue. i.e., it isn't a real attempt at an article. Yah, there's a lot of crappy articles that get into the edit queue, but no one is forcing you to read the edit queue. If you don't want to help out, don't read the edit queue.

As an alternative, maybe there should be a "hide this piece of crap article from my edit queue" button. This being for people who don't like their edit queue to be messy.

[ Parent ]
Edit Queue (4.33 / 3) (#95)
by stoothman on Fri Mar 21, 2003 at 08:11:54 PM EST

I think rogerborg is absolutely correct, the spam button needs to be renamed.  I think I am one of the people he was referring to.  The other afternoon I was working on a story.  I had to leave it and go run errands and did not get back to it until after 9 pm.  By then it had already been pushed to voting and was dumped.  So I never did get a chance to see if there was anything else I needed to fix.  Now this is my fault, since I did not post a note saying I would be back later to work on it.

I would not be opposed to either shortening the time to maybe 12 hrs in edit queue or even lowering the spam threshold.  However I think if this is done, it needs to be balanced out by an equal time in the voting queue instead of the vote only threshold.  I think along with this a dumped story should be placed in the author's diary.  Not so that every post made by an author is captured, but so that the lively discussion on some stories that are dumped can continue.

In addition, I think the addition to the FAQ is a great idea, so potential authors have some idea of the ground rules and the customs.

New Features (4.66 / 3) (#99)
by rhyax on Fri Mar 21, 2003 at 09:24:26 PM EST

I would just like to say that i don't like having new features added without any idea how they work. I have searched occasionally for how this button works and the best i can ever find is an occasional comment here and there. Honestly, when something like this happens that's what site news is for.

But surely you understand perl! (4.00 / 1) (#112)
by Rogerborg on Sat Mar 22, 2003 at 02:58:59 AM EST

It's right there in the source!

Sorry, sorry, I agree.  I'm surprised that rusty hasn't explained how this works, but I believe it's because he's still waiting to see how it does work in practice.  This article is an attempt to get him some feedback for that.

"Exterminate all rational thought." - W.S. Burroughs
[ Parent ]

How it works (4.50 / 4) (#117)
by rusty on Sat Mar 22, 2003 at 08:02:45 AM EST

I was sure I have explained how it works. It may only have been in a comment somewhere, in which case this one isn't going to do much good either. But if you're still reading, it's like this:

When a user views a story in the edit queue, we have a database row that indicates that (this row is related to the code that tracks new comments). When you hit the spam button, we've also got a row that indicates that. So we have two key pieces of information: how many people have read a story in editing, and how many of those have hit 'spam'.

From this we can get a ratio of reads-to-spams. The code waits for a minimum number of 'spam' votes (currently 25), and then starts calculating this ratio. If it goes over a threshold (currently 20%, or 1 spam vote per 5 views) it gets kicked into voting.

Both of the "knobs" above, threshold and ratio, can be changed. I observed the numbers pretty closely for the first couple of weeks, and there was a pretty clear break-point around those values. Most stories would either never get 25 spam votes, or if they did it was only after thousands of views. A few would get 25 spam votes within about an hour or two and generally were just above 20%. If you don't believe this is the case anymore, I can do another active survey, but unfortunately not for a week or so (I'm on the road this week).

____
Not the real rusty
[ Parent ]

I get 100 votes, 25% (4.00 / 1) (#119)
by Rogerborg on Sat Mar 22, 2003 at 09:56:37 AM EST

Based on CVS on 21-March-2003


*@laptop:~/downloads/scoop/scoop> find . -iname *.sql -exec grep -H -s spam_vote {} ';'  
./struct/patch-files/0.8-0.9/patch-37-MoreEditQueue.sql:INSERT INTO vars (name, value, description, type, category) VALUES ('spam_votes_percentage', '0.25', 'Percentage of spam votes required to forced a story into the normal moderation queue', 'num', 'Stories, Comments');
./struct/patch-files/0.8-0.9/patch-37-MoreEditQueue.sql:INSERT INTO vars (name, value, description, type, category) VALUES ('spam_votes_threshold', '100', 'Number of spam votes required before the % calculation can take place.', 'num', 'Stories, Comments');
./struct/scoop.sql:INSERT INTO vars VALUES ('spam_votes_percentage','0.25','Percentage of spam votes required to forced a story into the normal moderation queue','num','Stories, Comments');
./struct/scoop.sql:INSERT INTO vars VALUES ('spam_votes_threshold','100','Number of spam votes required before the % calculation can take place.','num','Stories, Comments');

OTOH, perl and sql are for filthy socialist hippies, so what do I know?

It's working well, but perhaps a slight relaxing on the percentage might be in order.

Also, does the percentage definitely only count one view per account, not all views in total?  I don't help to preserve an article simply by re-viewing it, right?

"Exterminate all rational thought." - W.S. Burroughs
[ Parent ]

Those numbers are adjustable (4.66 / 3) (#120)
by rusty on Sat Mar 22, 2003 at 10:48:28 AM EST

The actual values of number of votes and percentage are just SQL values. They are not the scoop default values here, they're based on my observations of what the real usage looked like. And I checked before wriitng the above comment that they are in fact 25 votes, 20%.

No, viewing a story more than once doesn't count as more than one view. There is one database row that says "Rogerborg viewed this story last at X date, and it had Y comments." Next time you view, that's updated but remains a single row.

I think if anything were to be relaxed, it would be the 25 votes threshold. By the time the threshold is reached, it's rare that major changes happen in the ratio. That tends to be a representative sample point. I will take another look as soon as I can though.

____
Not the real rusty
[ Parent ]

This is what I mean (4.00 / 1) (#122)
by kpaul on Sat Mar 22, 2003 at 11:06:43 AM EST

by learning the 'rules' as you go. I feel as if I've just discovered a treasure chest and gained 200xp ;)


2014 Halloween Costumes
[ Parent ]
Rogerborg (5.00 / 2) (#131)
by Rogerborg on Sat Mar 22, 2003 at 03:45:48 PM EST

needs food, badly.

"Exterminate all rational thought." - W.S. Burroughs
[ Parent ]

Mad love (none / 0) (#132)
by Rogerborg on Sat Mar 22, 2003 at 03:46:30 PM EST

You filthy socialist hippy. ;-)

"Exterminate all rational thought." - W.S. Burroughs
[ Parent ]

representative sample (none / 0) (#152)
by simul on Mon Mar 31, 2003 at 12:28:13 AM EST

typically, only 30 votes are necessary for a sample that's significant to 1 vote.

assuming people don't gang up in cliques or clumps, at 30 views only 7.5+1 or 8.5 spam or 9 votes would be needed for a 96% assurance of spam.  

clumping works by getting a story hidden or spammed before enough people have seen it.  to reduce the effects of "clumping", you need to skew by time

when it first goes in, it requres 25 votes/20%, but then each hour that goes by, it requres fewer votes, bottoming out at 10/20% after 6 hours or so... safely above a margin of error, and long after any organized "attack" could have been committed by a group.

i would say that time/skewing to reduce clumping should also work on the main story submissions... ie: a story could quickly go to -25 within 3 minutes, but because it happened so quickly, it could be given time to work it's way back up.  something like -50 skewed down to -25 over a half-hour period.... or something like that

again, this reduces the power of voting cliques.


Read this book - first 24 pages are free to browse - it rocks
[ Parent ]

(yet another) idea... (4.00 / 1) (#105)
by kpaul on Sat Mar 22, 2003 at 02:17:14 AM EST

if someone has a story 'spammed' X number of times, his/her stories automatically go into the queue for voting or stays in 'edit' for a shorter amount of time.


2014 Halloween Costumes
That will meet a lot of resistance (5.00 / 2) (#111)
by Rogerborg on Sat Mar 22, 2003 at 02:57:30 AM EST

Even though there's very little wrong with it in theory.  But we tend to shy away from punishing accounts because it's so trivial to just create a new one.

Better the turmeric you know than the 4yu8934hrwe you don't.

"Exterminate all rational thought." - W.S. Burroughs
[ Parent ]

Yeah, you're right... (3.00 / 1) (#114)
by kpaul on Sat Mar 22, 2003 at 03:01:09 AM EST

the cliques would become even more maddening with people blacklisting each other ;)


2014 Halloween Costumes
[ Parent ]
And what is that third number (none / 0) (#129)
by exile1974 on Sat Mar 22, 2003 at 02:44:55 PM EST

On the moderate submissions link. I figured out the vote and edit but still
don't know about the last.
"A sucking chest wound is Nature's way of telling you to stay out of a firefight." --Mary Gentle
Third number (none / 0) (#134)
by cep on Sat Mar 22, 2003 at 04:11:32 PM EST

The third number is the number of submissions in the queues.

[ Parent ]
Eehhhh? (none / 0) (#135)
by exile1974 on Sat Mar 22, 2003 at 04:46:33 PM EST

0/6/11 doesn't add up it would have to be 5/6/11 to make sense.

exile1974

"A sucking chest wound is Nature's way of telling you to stay out of a firefight." --Mary Gentle
[ Parent ]

Because the first number is unvoted (none / 0) (#138)
by ZeuS572 on Sat Mar 22, 2003 at 07:41:15 PM EST

That's because the first number is the number of submissions in the voting queue that you have not voted on. So for example, 0/6/11 means that there are 6 in the edit queue, 11 total, and 0 that you haven't yet voted on (implying that there are 5 in the voting queue that you have voted on). So 3/6/11 means 5 in the voting queue, 2 of which you HAVE voted on, and 6 in the edit queue. I didn't immediately figure this out, and I had to see how the numbers changed as I voted to see the pattern.

[ Parent ]
Made a lot more sense when there was no edit queue (none / 0) (#141)
by roiem on Sun Mar 23, 2003 at 03:34:00 AM EST

Back then it just said Unvoted/Total, which was pretty obvious.
90% of all projects out there are basically glorified interfaces to relational databases.
[
Parent ]
What exactly does the spam button do ? (MT) (none / 0) (#142)
by OldCoder on Sun Mar 23, 2003 at 03:33:37 PM EST



--
By reading this signature, you have agreed.
Copyright © 2003 OldCoder
It increments a counter (none / 0) (#148)
by Rogerborg on Wed Mar 26, 2003 at 11:22:18 AM EST

When the counter hits at least 100 and at least 25% of those who viewed the story have hit Spam, the story gets sent to vote with a 0 starting score.

There's a slight implementation hiccup in that the standard voting mechanism is re-used, and a Spam vote is counted as a -1.  The vote count gets reset when it goes to vote, but if you're viewing a story in edit and hit Spam after it's been moved to vote (either by the author, the 24 hour timer, or spam votes) then your spam click gets counted as a -1 vote.

"Exterminate all rational thought." - W.S. Burroughs
[ Parent ]

Edit Queue Contradiction (none / 0) (#143)
by OldCoder on Sun Mar 23, 2003 at 03:42:36 PM EST

Should one be permitted to vote down an issue to which one has posted a Topical comment?
In other words, if I actually post a topical comment in the Edit queue, which means I'm discussing it, isn't it a contradiction for me to vote down the article once it's in the Vote queue?

Also, shouldn't voting be limited to people who are over 18?
ok, that's a troll™ — well, not completely.

--
By reading this signature, you have agreed.
Copyright © 2003 OldCoder

No, it's not. (4.00 / 1) (#145)
by tiamat on Mon Mar 24, 2003 at 07:46:33 AM EST

There's any number of reasons to vote down an article other than just discussion merit.

[ Parent ]
No! That's stupid, because (none / 0) (#147)
by Rogerborg on Wed Mar 26, 2003 at 11:18:00 AM EST

...uh, because, ummm...

You know, that actually makes a lot of sense.  Interesting suggestion.

"Exterminate all rational thought." - W.S. Burroughs
[ Parent ]

Let people post whatever comments they like. (5.00 / 1) (#150)
by Vellmont on Sat Mar 29, 2003 at 03:44:20 AM EST

There's many valid reasons for voting an article down, and also posting topical comments. You may believe it's inevidable that the article will be posted, but you believe it's important to clear up some important flaws in the article.

[ Parent ]
And another thing... (5.00 / 3) (#144)
by MajorMajor on Sun Mar 23, 2003 at 04:54:11 PM EST

Can we have a preference to set whether a comment is editorial or topical by default? I'm fed up of posting comments and then realising I've forgotten to set them to "Topical" instead of "Editorial".

The 24 hour edit queue and the spam button | 155 comments (128 topical, 27 editorial, 0 hidden)
Display: Sort:

kuro5hin.org

[XML]
All trademarks and copyrights on this page are owned by their respective companies. The Rest 2000 - Present Kuro5hin.org Inc.
See our legalese page for copyright policies. Please also read our Privacy Policy.
Kuro5hin.org is powered by Free Software, including Apache, Perl, and Linux, The Scoop Engine that runs this site is freely available, under the terms of the GPL.
Need some help? Email help@kuro5hin.org.
My heart's the long stairs.

Powered by Scoop create account | help/FAQ | mission | links | search | IRC | YOU choose the stories!