Added a garbage collector to inca. And thanks to valgrind, it's not crashing so much anymore.

@undergroundmonorail - yes to TI BASIC

Can we define concensus to be some function of the number of active users on meta?

For example, "strongly agree" two years ago may have been 3 votes. "Strongly agree" now would probably be 12-15.

I'm just curious, because I don't see a way to get anything useful out of meta with all of these "discussion" question that lead to nowhere.

If we need to define consensus, it is probably better to look at the question views instead of the active users on meta.

I like that

Except that more controversial topics will have the same people returning again and again

That leads to more views and eventually makes the topic "un-concensusable" is that is a word

If Alice, Bob, and Charlie argued for weeks and then finally reached a concensus, the view count over the span of weeks might be too high

Hmm... then we would need something like unique views, but that feature doesn't exist.

That would be much better

Votes are like unique views I suppose

Then perhaps 'highest vote count' is enough.

How about something like % difference in vote count. If you assume that all views can be divided into roughly two categories for a given topic, then you can take the % difference to get some number.

@Rusher By "% difference in vote count", do you mean that an answer with +6/-3 has a 50% difference?

No, I mean that if it were possible to group the answers into two categories "those that think one way" and "those that think another way" then the difference would be the difference in the total number of votes between the two categories

So if I ask, "Should ProgramFOX be banned from CodeGolf?" on meta

I might get two answers that say No (for different reasons)

One answer that says yes

Total No votes is 10. Total Yes votes is -20.

The percent difference is abs((10 - (-20)) / -5) = 600% where -5 is the average of 10 and -20

That would put the community in an overwhelming 600% of not banning you from Code Golf

lunch time :)

@Rusher The problem with counting downvotes is that there's probably a lot of overlap. -10 of that -20 probably came from the same 10 people that voted the other one up. You'd probably get a more meaningful number if you count only upvotes, not total score.

That being said, I think trying to quantify "consensus" this way probably isn't going to work in the long run.

@Geobits If your assumption were true, then every two sided issue would have one answer with positive votes, and another answer with an equivalent number of negative votes. The total votes would always be zero. This is obviously not the case, and often there are multiple opposing answers with positive votes.

Can the down voters on this question revisit it to either retract or bring up a NEW issue? It's about to be randomly closed by the next bandwagon voter that comes along, even though the author thinks that he has addressed the issues brought up so far.

Right, but for every downvote, you can only guess whether it is balanced by an upvote on the other or not, that's my point.

I'm not saying it's true for every one, just that we have no way of knowing.

Just counting upvotes doesn't have this problem.

Well, your proposal only worked on the assumption that mine was broken. Your assumption was actually broken, and I agree that you can only "guess".

I argue that counting only upvotes has an equivalent problem. I may choose to upvote something only because I downvoted the opposing side.

My assumption was that we don't know, so it wasn't broken. That's why I used the word "probably" three times.

After you probably a hundred things together, you end up with a total score

Why would you use total upvotes and completely ignore the total downvotes?

Okay, but what's the benefit of counting downvotes for this?

What is the benefit of counting upvotes?

Every argument you make for upvotes, I will make for downvotes.

Since one is not better than the other, we consider them both equally meaningful

Right, so count only one of them. My point is that you shouldn't mix them up.

Which one?

I propose upvotes as the one to count because there are usually more of them.

Also, WHY should we not count both?

Because you can vote on both answers.

"I can vote on two answers, therefore the total score is misleading." is a nonsequiter

Try again

There's a reason you can't vote people down in elections. You get a positive vote only.

If you vote one up and one down, you essentially get two votes per person.

So, those that choose to vote the opposite view down have an outsized say in the matter.

And that is counterbalanced by the fact that everyone else can also do this. In a game where the same options are available to everyone, nobody has a disadvantage . There is a game theory law that says this, but I can't remember the name

Except that some people have some weird stigma against downvoting, and don't do it as often (even on meta where no rep is at stake). I'm not saying that stigma is logical or unfair, just that it happens.

So the net effect is that some people only vote up, while others get double the votes.

That doesn't help at all to prove that there's a consensus, which I thought was your goal. It only shows a winner in raw votes.

That's why I said quantifying consensus doesn't really work here. It only generally works if you have an idea of the number of potential voters, not the number of actual voters (which we don't even have, since it could be anywhere from totalVotes to totalVotes/2).

It's easy to say "a bunch more people voted this way" and take that as agreement. It's hard to say "we have a 600% consensus".

Are people that have that stigma more likely to be on one side or the other of a particular issue?

If not, then again I say that your argument is statistically not important.

We can argue the details if you like, but my overall point is that it doesn't matter. You can't put a solid number on this.

I'm all for arguing the details. I present to you issue X that has two sides, A and B. Which side, A or B, benefits from double-voters?

Whichever side the double-voters are on, obviously.

Which side is that?

Put this in a real world context. Two groups of people, one that gets two votes in an election, one that gets only one. Is that fair? No. You don't need to know ahead of time which side they will fall on to determine that.

You could say that in the long run, over the course of many questions, it might even out, but on any particular question, it's obviously unfair.

Are you saying that some people on Code Golf only have a single vote?

Your analogy was not true to Code Golf

Everyone here has the ability to vote as they choose, be it on one answer or all of them

Try this: There is a question with three answers. One positive, two negative (to whatever the question is). One person votes up the positive, and down on both the negative. How do the two negative votes add any information that the single upvote doesn't?

Allow me to present your own real-world context again. Everyone gets two votes in an election, but you don't have to use them both. Is that fair?

If they were both positive votes freely usable, then that could be considered fair. STV voting is an example of this. You don't see many actual voting systems with negative votes, though. There's a reason for that.

So any system with negative votes is, by default, unfair?

Yes, I suppose that's my point. In general, at least. There may be some esoteric system that uses them that I don't know about that would prove the absolute statement wrong.

We're clogging up this chatroom. Move here to continue this discussion chat.stackexchange.com/rooms/14143/room-for-rusher-and-geobits

Great... I close voted for the wrong reason and then retracted, and now I can't vote again

codegolf.stackexchange.com/q/26303/18487 <- CV please...

On the wrong forum and contains only a single sentence of spec.

The rest is a copy paste of all code-trolling questions

I just realized that the code trolling tag was actually encouraging users to copy a disclaimer into every post. I removed that bit, but it awaits peer review. The only time a tag should be mentioned in the question is when it needs additional clarification.

So I just finished that wall KOTH.

No idea why it has 4 CVs. There is a test program (even though I have no idea how to run it) and I still was able to solve it.

There wasn't one when most of those CVs were coming in.

Ah, makes sense.

^ I asked for people to revisit it and the votes remained

As for the code trolling post, kill it with fire. Please. Now.

oops, one of them was mine... retracted...

I already tried to cv the complex number challenge but I accidentally said it belonged on Stack Overflow

Then I thought "No wait, it belongs on Mathematics"

So I retracted and now I can't vote.

I really don't get the vibe about code-trolling questions here... like... I can see that there are some people not too fond of them in general (mainly the veteran users)... and I'm not sure I like them myself. But still most of them make it into the Hot Network Questions, and after a while they rake in loads of upvotes and answers. But then this one seems to be just outright hated... I don't see the difference.

No, it belongs in the firey pits of doom.

Well it's implied, because it's a duplicate on Mathematics, and would therefore be closed

Someone in the comments of this one said that CT questions were supposed to mimic bad questions on SO. That is so much the opposite of some of the better ones I have seen

I think some of the CT followers now believe that "clever" is the same thing as "bad"

Well my impression is exactly that though... (troll answers for give-me-teh-codez questions) I actually thought that's the purpose of CT as well. If it's not, what is?

(i.e. please show me an example of a CT question that is good/does the opposite)

btw, @Doorknob it would be great if you could look into some of the questions on the easter egg KotH ;) (also the tester.rb sometimes runs into an infinite loop right in the first round)