« first day (2141 days earlier)      last day (1790 days later) » 
00:00 - 14:0014:00 - 00:00

2:00 PM
I just don't think that in the sea of OPs out for easy quick cheese (answers that work but are ultimately detrimental to their program or knowledge) that they should be choosing the best one. In a scenario of "I give you some quick code to copy paste" and "I explain to you why you need to re-think your approach for a given scenario and do something better instead" the first getting accepted is an objective loss for future readers
In fact, I think the loss to future readers is easily quantifiable given the number of views on questions with an answer voted higher than the accepted answer.
And the OPs that know what they're doing and accept properly, usually (I have no data but experience says this) choose the highest voted (== judged best by a jury of experts) answer anyways
trying to judge an answer's helpfulness by its metrics is a pretty meaningless task... votes and accepts are just people's opinion
Maybe we can keep the checkmark and just remove its sticky-power
@Quill A meaningless task? It's the task that is fundamental to how SO (and other content sites) works as a platform.
2:03 PM
@bjb568isnotapebble And, again, what about new answers that are now more accurate and have had the accepted answer changed to that new answer?
@bjb568isnotapebble at the end of the day, that's all just opinion. just because the jquery answer has 300 more votes than the vanilla js answer doesn't make it better just because 300 people thought so
@Andy This is pretty rare. For a) old answer b) out of date c) new answer d) OP's still here e) OP takes action to change accept f) accepted answer still doesn't have as many votes
individually we all need to make decisions based on our own opinions instead of hiveminding onto the idea that other people's votes are an indication of quality
@Quill Of course not, that's why there are multiple answers, but if 300 more people think the jQuery answer is helpful than the vanilla answer, it should be on top so as to not waste new visitors' time.
@Quill Thats by and large the only way that makes some sense to determine that
2:06 PM
Ok. What about the answers from years ago that recommend bad practices (looking at PHP here). Many of those have huge numbers of upvotes. Removing the accept doesn't do anything in those cases. New answers will take years to be equal in score and that's assuming the old ones get no new upvotes
Unless you're an expert in the field yourself, in which case that answer does bupkuss for you
@Quill Most people currently do and always will skim the top answer. We need to optimize to them.
@Andy No algorithm of sorting answers is going to make PHP suck less.
That's ignoring the question.
Every language can and does have bad practices
If good practices doesn't get upvotes, how will the system know to show them at the top?
Without voting as a signal, the system can't work. Accepts don't solve that.
@bjb568isnotapebble Voting doesn't solve that either. At least that's how I'm reading your argument. If a post has so many votes that a new answer can't over take it, so it's never read, what's the point of a user wasting time posting such an answer?
2:09 PM
Answers are to solve questions, if the question isn't solved then the answer is moot, even if 5 or 50 people vote it up
@Andy That's why it should be sort by hot, not top.
@bjb568isnotapebble Like voting velocity?
The system should take into consideration post time and votes.
Not accepts.
Somewhat doing that'd do much, honestly.
@Undo Yeah. You could get even fancier with looking at the sequence of votes with edits, new answers, comments, and traffic.
2:12 PM
Interesting idea
Not sure how well it'd work to solve the answer sorting 'problem'
sorting answers by votes isn't about the question anymore. it's about the problem all the voters think the question is about, which is the same as forcing duplicate closing
@bjb568isnotapebble Do you think that's really going to help? GoogleJoe wanders in to a post and sees one post with 50 upvotes and one with 5. GJ reads the 50 votes post, thinks it's reasonable and upvotes. They don't move to the 5 vote one because they can copy paste the answer and it works. Now the scores are 51 to 5
@Andy Reddit discounts pile-on votes (after a lot of votes are already there) so they have less effect.
Quick question: If I vote to close a question as duplicate, and the dupe target is later deleted (before the question gets closed), what happens to my pending close vote ?
@bjb as bluefeet said, any of these would be too confusing for users. Seeing answer with score of 5 above answer with score of 30 is confusing and will make people think "this is bad", without bothering to check why it's like that.
2:14 PM
Honestly, the only time I really scroll down to the next few answers is when the first answer wants me to use a jet turbine to sharpen a pencil. Sometimes there's an actual pencil sharpener down further.
If anything, maybe let votes "expire" somehow.
@bjb568isnotapebble Which is valid, for their community. Now you are suggesting two very big changes to this network. 1.) Answers are sorted by a new algorithm and 2.) If you upvote an already upvoted answer, your vote counts for less.
@ShadowWizard if you were a new user (not familiar with the network) would that be intuitive?
@Undo You know you want to throw a pencil in a turbine...
@ShadowWizard I don't really think users who are just copy-pasting will care that much. Reddit users certainly don't care. But if need be, post scores can be hidden for users under some rep level (with an option to enable them even if you're under).
2:16 PM
@bluefeet There's an argument there for "if implemented right, it doesn't need to be intuitive"... but it'd be insanely hard to implement 'right'
@bluefeet It can be simple enough with a tooltip for when their vote disappears.
@bluefeet Just not giving the accept sticky-power would be intuitive
@Andy Sure. SO needs some big changes — it has big problems.
more intuitive then now, anyways
@Undo At SO's scale they can throw A/B tests at everything. I don't think it should be an engineering challenge.
2:18 PM
I have a hunch though that this will just be another tsunami in the water glass and nothing will come of it in the end
@bjb568isnotapebble I'm going to disagree with you on that one.
vote expiration sounds unnecessarily complicated.... and then you add rep on top of that, old question votes expiring back to zero on all answers
@bjb568isnotapebble A/B testing something that central is going to be ugly
for all its advantages and there are many the SE system of self governance is very attracted to gridlock
@bluefeet you mean new user just looking on answer? I won't be able to know votes expired. If you mean as a user whose post score just dropped... well, that's a problem indeed.
2:18 PM
@Undo Eh, well, maybe not A/B tests, but SO is so huge that they have the tools to do things like this.
Not following
@Undo but it does have to be intuitive. Most people coming to the site aren't going to be familiar with "how things work" we wouldn't want to implement something that was difficult to explain, etc
@Andy Even if you don't fix everything, any one improvement should help, somewhat.
@blue the most intuitive, straightforward design would be "Highest voted answer first, no exception"
And then having a green checkmark on the one that the asker chose
@Undo They can do usability testing. I can't. They're big. I'm small.
2:20 PM
@bjb568isnotapebble since score is one of Stack Overflow most famous "marks" (along with reputation), it won't happen.
It gives both signals very quickly. the current system means you have to look at both checkmark and score and the score of the following answers to know
@bluefeet Agree, but theoretically this could be a black-box thing that no one needs to know about. For example, I don't need to know anything about Reddit's algorithm (still don't) for it to pull up interesting stuff. I click on the interesting stuff and vote accordingly, and I have no idea what actually happens with those votes.
@ShadowWizard Eh, fine. DevDoodle's hidden post score from the beginning.
@Magisch it stops being about the question being asked and just becomes about how helpful the answers are to others
With a change like that I could look immediately at the top scored anser and decide if I want that, and if I want the asker's perspective instead I can go for the green checkmark
2:20 PM
@bjb568isnotapebble well, DD never aspired to be like SO.... or did it? :)
But... yeah, the idea of changing votes to something other than 1:1 mapping is scary
the answers aren't there for the hell of it, they're there to answer the question
@Quill aka "Exactly what SO is about"
that's not what SO is about
Good answers are precious, good questions only exist to set up good answers.
SO is answer centric
2:22 PM
Not really... there's a reason pages don't load straight to the top answer
@Quill to achieve the ultimate goal here I don't think we can have something that is not really complicated...
What's up with all the stars?? lol
@Magisch Exactly. The pearls are the best answers.
Load a question -> read the question -> is this your question -> yes -> read the answers
                                                              -> no  -> try again with the next question
I'd love to see a change to the sort (I've spent a ton of time on it), but there is a lot of resistance to doing it. It's a massive change to how we've done things since inception.
@Quill you can see that everywhere from the "Optimising for pearls" blogpost to the +5/+10 discrepancy to the downvote cost
SO is focused on good answers, and good questions are only a facility to make them easier to provide
2:24 PM
then let's just abandon the question model. let's just set up one big page, and everyone can just answer every problem there
When I search for something, I already have the question. The only thing that matters about the question that I'm going to find is that its close enough to what I want. Outside of that the only thing thats of interest is the answer
@Quill That's not how I do it. Most of the time I look at the answers first and see if any have helpful info, then look at the question and either find search keywords or linked/related questions.
But we're talking about a drop in the bucket when looking at a the total number of questions on SO.
because obviously the question doesn't matter... a javascript question with great C# answers is fine because the answers are great
@bluefeet Maybe we could have a trial period
And an evaluation later on
2:25 PM
@bjb568isnotapebble there's a reason pages don't load straight to the top answer
The number of questions with an answer scored higher than the accepted answer is very small
But for these questions it matters
@Quill The questions provide a way to focus answers.
Take mine for an example @blue
An user could be mislead from that
I had no clue when I accepted that answer, and how coul I? I just asked about that.
@Quill It's great, except not easy to find. Questions function also as a signpost.
2:26 PM
The Qs where thats the case tend to be high-views high-traffic, too.
@Quill is there?
The answers are what you're looking for on SO, the questions are just signposts to them.
@bluefeet Then changing just those questions' answer sort is a small change.
and meaningful
good questions aren't singularly about people providing good info. I don't go to a mechanic to fix my car's tires, only to have him fix the engine, steering and exhaust excellently. Questions are for solutions to the problem
Not for solutions to "a" problem
the author has a problem, the answers fix the problem. the answers aren't there to fix everyone else's problem, they're there to fix the author's problem
Right, a set of focused answers is important. But once that's done, you don't need the question anymore.
2:28 PM
if everyone else has a problem, they're welcome to ask it, and have answers for their problem
Questions are for answerers, not googlers.
@Quill No. They're not welcome.
@bluefeet since Jeff left there is a feeling that things can radically change in SE, and that feeling is still there for me. All the changes so far were cosmetic or new additions, but I believe the current management and team can, and will, touch the very core and change it if convinced it is required. Not talking about removing downvotes or accepts, but something like letting votes expire, or change the sort algorithm is possible. Just need someone strong enough from the inside to push it. :)
@JonasCz The duplicate vote gets deleted, too. It's no longer there. Even if the target question gets undeleted, the vote is gone. (Source: experiment)
@ShadowWizard and a decent proposal....
@Quill actually most real changes come from the team itself.
2:30 PM
@bjb568isnotapebble individual, non-duplicate questions aren't welcome? huh....
@ShadowWizard I agree with that and we can do that. I'm still trying. :P
a bit hard to know what actually works
@bluefeet I have faith in you! ;)
@Quill If it's a good question, it'll be accepted. But most aren't. And answers are valued more.
@ShadowWizard proposals from the team aren't always a-ok
2:30 PM
@bjb568isnotapebble the preference would be to see a larger impact if making a big change
Hi all
in JavaScript: Do you even lift? on Stack Overflow Chat, 2 mins ago, by Neal
tried to make a canonical "why doesn't jquery's .data function work anymore" question/answer http://stackoverflow.com/q/37814855/561731
@ShadowWizard Wasn't the feeling there while Jeff was at SE? He did change a whole lot of very core things.
@Quill not talking about proposals, more like announcements, e.g. profile redesign.
Already getting downvoted by the JS room (just bc I think hrrrmf)
@bluefeet Well, make a bigger change, get a bigger impact. You can reform the answer/vote system, not just switch ordering in a small set of cases.
2:31 PM
@ShadowWizard oh, right
Those come after being decided already, they just seek input and feedback.
@ShadowWizard oof. A paragon of changing what's not really important.
In a way, may I add, that made things worse. The new profile page has only one feature for me: the link back to the old profile page.
@bluefeet Who is resisting changes to sort? People who wrote crappy answers 8 years ago?
@bjb568isnotapebble that was just an example. Like I said, there wasn't any change in the core of SE, it's all cosmetic.
Obviously, this burst of chat activity was meant to win the message #5000000 for the Tavern. Sadly, we lost to the employees... this was #4999999 and the three messages after that were in private rooms.
2:34 PM
@AaronHall When proposing big changes, we need buy-in internally.
> “Forty-two!" yelled Loonquawl. "Is that all you've got to show for seven and a half million years' work?"
"I checked it very thoroughly," said the computer, "and that quite definitely is the answer. I think the problem, to be quite honest with you, is that you've never actually known what the question is.”
haha, perfect fit
@amanaPlanaCAnalPAnaMA People don't like general reference questions. They don't think they're on-topic.
Sure, your shareholders might not want to annoy programmers with probably a decade of experience, but you currently destroy value for everyone else (wasting their time with inadequate answers) to do that.
the answer is important, but so is the question, knowing a solution isn't as important as the solution to the question
2:37 PM
@AaronHall The current system is also shedding off its experts as they get tired of wading thru the sea of crap that is created by the self-destructive cycle of solving problems for OP rather than generating a lasting Q&A resource.
@AaronHall yeah, but people build highways every day, and people generally don't complain for an invalid reason. the new highways are helpful and useful and help you get to work faster, unless they topple an issue somewhere else, only cynics will complain
@bjb568isnotapebble Haha they are most definitely on topic :-)
@ShadowWizard interesting. I did not know that tag existed (do we really need it?)
hey there @TIPS
2:38 PM
Q: How can I manipulate the DOM without a library like jQuery?

bjb568 is not a pebbleI am used to using jQuery for manipulating the DOM, for example: var mything = $("#mything"); mything.on("click", function() { mything.addClass("red"); mything.html("I have sinned."); }); But now I want to do the same things with Vanilla JavaScript. Is this possible? How can I do it? ...

I didn't notice the snippets reskin... awesome :D
@bjb568isnotapebble That is not the same thing. That one is asking how to do something without trying anything. That is not a Q for Stack Overflow (sorry :-( )
Many people forget what SO really is. Wiki, blog, forum, Digg/reddit.
2:40 PM
mailing list
@amanaPlanaCAnalPAnaMA Eh, maybe borderline, but it has Benjamin Gruenbaum's support. Questions don't have to have attempts, attempts are only helpful to give answerers more direction.
@amanaPlanaCAnalPAnaMA it wasn't closed as lacking an MVCE
@amanaPlanaCAnalPAnaMA of course we need it, the whole point of that post is to explain about a breaking change in the new jQuery version. Also, you better make the answer CW and flag the question asking to make it CW as well.
Seriously - the people who lose if we change sort are people who have 100+ upvotes on half-baked 8 year old answers.
@ShadowWizard Why should it be CW?
2:42 PM
> good answers would be too long for this format
I see no reason for CW
Problem is I already wrote the answer…
@amanaPlanaCAnalPAnaMA GR questions typically are CW. Makes it easy to edit by the community.
@amanaPlanaCAnalPAnaMA because it's like a faq - you answer something before it was actually asked. You didn't solve any actual problem someone was having.
@bjb568isnotapebble yeah, it was closed as too broad, it wasn't closed as lacking a Minimal, Verifiable and Complete example
@ShadowWizard Then Stack Overflow should not give the option to answer your own question in the same fell swoop. I think you are misunderstanding what community wiki is :-(
Jeff Atwood on July 1, 2011
The FAQ has contained one key bit of advice from the very beginning:
2:44 PM
@amanaPlanaCAnalPAnaMA point is not self answer. You didn't actually have that problem.
A: Tag Wiki vs Community Wiki vs Self-Answered Questions

Cody GrayThat close vote probably came from the giant wall of bold-italic text that takes up 1/2 of the question yet has nothing to do with the question. And the fact that the question isn't actually a question, it is a blog post. You need to ask the question in question format. You didn't do that here. ...

> You can make these community wiki if you want, but don't feel any pressure to do so. As long as the answer is mostly yours, you deserve the credit for creating it. Writing good canonical questions and answers is hard.
Ok, it's your choice.
@ShadowWizard And yes. I have had that problem thank you...
And I found the solution
But I think voters favor CW GR questions.
Therefore I wrote the question/answer.
2:45 PM
general reference
dup target
@amanaPlanaCAnalPAnaMA hmm... fine, won't fight you over this, just saying what I would do in such a case.
@ShadowWizard ok.
GR = canonical then?
2:47 PM
I considered making my top answer community wiki. But I don't want to.
thanks, I'll be here all... nope, I have work to do...
@bjb568isnotapebble 724 views in more than a year... doesn't look like this was a question that people actually ask.
@soup It was to be used as a dupe target. But that plan failed with the controversy over it.
At least I wrote a thing.
Maybe it was useful to somebody.
It surely was, just not to many.
2:57 PM
@bjb568isnotapebble what did you write?
"visited 1067 days, 784 consecutive"
I have a life, it's just always online.
[ SmokeDetector ] Bad keyword in title, phone number detected in body: !8!885O78O6 dell tech support by kalimata on civicrm.stackexchange.com
3:32 PM
@AaronHall Majority of the answer.
I'm on iOS 10.
@bjb568isnotapebble night
Hmm, so it looks like I missed a discussion on accept votes. Didn't bjb have this discussion years ago?
> SiriKit
should be interesting to see what new apps abuse this for a few days before being pulled from the App Store
@hichris123 and Brad posted his answer on meta just now, same topic
Oct 4 '15 at 21:40, by bjb568
What's really absurd here is how OP can accept an answer with 100 votes less than the top answer and make it rise above it.
Okay, not that much of a discussion...
3:36 PM
@hichris123 Hell, BJB with correct spelling and punctuation?!
Cats are coherent most of the time.
In cases where OP recognizes a new answer as best, it works out.
Real life example from an hour ago: I wanted to know how to select all in nano. So... superuser.com/questions/196425/…
But that's only happened like 3 times to me.
But my top 2 answers are such.
The top voted answer has way more votes than the accepted one. But there's a good reason: accepted answer addresses the question that was asked, top answer addresses the question that people search for when they land on that page.
3:39 PM
> Incrementalist
strange badge name
@Quill Huh? What's that from?
@hichris123 Docs beta, it's for getting a change that passes review
Oh, that thing...
That non-bloody, non-gutsy thing.
it has a meta thing
3:41 PM
That's why you need to take time into account
This answer has 1264 upvotes: stackoverflow.com/a/38990/541136
and it doesn't even work on Python 3.
@soup It actually seems to be a decent idea. But... the last time I checked, the Documentation wasn't too great...
@AaronHall It claims to...
ok, it added a piece at the end as a workaround.
Anyways, it's a case where it eventually worked out for OP's accept to pin at the top, when he changed it. But if we valued recency of votes, it would be moot. And Reddit users don't get all confusaled when posts aren't ranked by net votes.
They just accept it as an indication of post quality that when a post with more votes is ranked lower than another post that the post with more votes was simply more controversial as well.
Would someone please unpin consider unpinning "please a dude"?
@AaronHall :o You don't want to please dudes?
3:53 PM
Wait, the duck went back to @InfiniteRecursion.
Gah, why flag that people?
in Documentation Beta on The Stack Exchange Network Chat, Jun 7 at 20:45, by Adam
We sent out a bunch more invites earlier today. That's very likely the last wave of beta testers before the end we go public.
@Bart Flag what O' Bart sensei?
Your message to AaronHall
Mayhaps a misclick. I'll pass
@Quill Doc is on SE's chat? O_o
3:58 PM
Could be, though I never mis-flagged myself. Ever. There's no evidence of that. At all.
@TIPS it's nearly as dead as Language Learning chat ;-)
docs is really pretty though :)
@Bart Obviously not...
Good, that's settled then
4:04 PM
Feb 16 '14 at 1:13, by Bart
Wait what? I flagged that?
"next 6 months" That's incredibly optimistic. At the rate we're going we might as well permanently rebrand Stack Overflow with rainbows and unicorns. (Not that I'd be opposed to doing that apart from what is happening in the world...) — BoltClock ♦ 1 hour ago
5:11 PM
Oh, hey, @TimStone: Nick tweeted about live query plans in SQL Server 2016. Is that something we can get in SEDE?
5:23 PM
Hmm, maybe!
I'm not sure how optimistic I should be about that. :P
It was a nice solid "maybe". Look at that exclamation point.
How more noncommittal do you want?
but was it an excited ! or a sarcastic !?
Now we're getting to the important questions.
If it was sarcastic, he should have used the right punctuation.
Sarcastic Mark ^
Internet grammar, people. Come on.
insert tiny sarcastic mark here
5:33 PM
@hichris123 It depends on your definition of optimism
6:02 PM
Hmm, is this really too broad? stackoverflow.com/questions/394882/…
Eh, off-topic. Not really too broad though.
6:37 PM
[ SmokeDetector ] Blacklisted website in answer: Call me when you are back by lali on ell.stackexchange.com (@ShadowWizard)
1 hour later…
7:46 PM
[ SmokeDetector ] Bad keyword with email in answer, email in answer: How to protect my human from vampires by jackson on rpg.stackexchange.com
1 hour later…
8:52 PM
Q: Moderators Pro Tempore

Grace NoteThroughout the beta, we need members from the site whose focus is to engage the community, both in community-building issues and site management. That's why we select a few members from each community to act as temporary, provisional Moderators. You can read about the program here: Moderators Pro...

9:06 PM
@Quill Someone still hasn't said hi to room #4. :P
The fun part of the Reopen Votes queue is casting a reopen vote and waiting a couple hours to see what happens.
10:14 PM
8 hours ago, by Shadow Wizard
@bjb as bluefeet said, any of these would be too confusing for users. Seeing answer with score of 5 above answer with score of 30 is confusing and will make people think "this is bad", without bothering to check why it's like that.
This reads like an argument against pinning accepted answers.
@nicael :)
Anyways, calculate a hotness score (between 0 and 1? or 0 and 100?), post the score next to the net votes, and no one will be confusaled.
10:33 PM
Q: A&M EU Meetup Report!

Toshinou KyoukoJust last weekend Animecon was held in the World Forum - the same venue that has previously hosted such events as the EuroVision Song Contest. As we previously organised, a couple of our EU members (including myself) attended and met up. Unfortunately, a combination of an overly long leadup t...

11:15 PM
> Hi Quill,
Thanks for getting back to us!
We really appreciate your interest and support, and we'll get some goodies out in the mail for you shortly :)
If you like swag (and who doesn't?), you should know that entries are still open for the first HyperDev coding challenge. Details on this, and upcoming events, can be found on our blog:
- https://hyperdev.com/blog/developer-contest-get-back-nature/
The Arqade election report is remarkably simple. Two winners got 302 and 301 votes; so both got redirected to mod team.
in The Bridge on The Stack Exchange Network Chat, 3 hours ago, by Grace Note
@GodEmperorDune I was just asking internally about it. We've had multiple instances with at least one candidate well past threshold on the first round, but none where I can think that every seat was filled with more than one seat to check.
00:00 - 14:0014:00 - 00:00

« first day (2141 days earlier)      last day (1790 days later) »