The Nineteenth Byte

m.buettner

00:28

and here we go again

ProgramFOX

10:44

@TheDoctor No, beta sites don't have Ads.

TheDoctor

13:19

2 more CVs: codegolf.stackexchange.com/questions/26303/…

TheDoctor

13:50

Same here: codegolf.stackexchange.com/questions/26434/…

Rusher

14:06

@TheDoctor I can't find a reason to close vote the latter one. I downvoted it though.

Geobits

15:49

-1

Q: Why are waffles blue?

I want to know why waffles are blue. It is very weird and I want to find out more. Waffles are pancakes with built-in syrup reservoirs. They're very waffly, and better than pancakes, but they're blue, which is a real problem for me. print(Waffle.color == Colors.BLUE); // true

This isn't code trolling, it's just trolling.

Rusher

7 Minutes. That's a record huh?

Peter Taylor

@Geobits Two more delete votes needed.

Rusher

Do deleted questions count towards the minimum questions per six months needed to keep a tag alive?

Geobits

Hmm. I thought it had to be closed for two days before deletion? I don't have the option, just share/edit/reopen/flag.

Maybe I need 4k for that.

Peter Taylor

I think that once it gets to -3 it's eligible for instant deletion.

Geobits

15:58

Ah, found it on the privileges page:
> You must wait for a question to be closed for 2 days before you can vote for deletion. This restriction is removed for trusted users.

Will flag anyway.

Maybe I should sprint to 4k ...

@Rusher I don't think they count, but I couldn't be sure. I think when the script runs, it queries for questions within 'date-range', but I don't know if that's a mod-query which can see deleted or a regular search.

Good question for MSE if it hasn't been asked already.

Rusher

@Geobits Meh. Why not.

Einacio

16:15

i have been thinking of creating a simple site to showcase the king-of-the-hills challenges on "manual mode", i.e. human vs human instead of program vs program. what do you think about it?

Rusher

@Einacio You mean like... I tell my Wolf how to move?

Or Geobits tells his pack how to chase the flock?

Geobits

I'm not sure how feasible that is. If you're asking for human input on each turn.... mine normally runs to 10k turns each round. Others vary, but it's normally a lot.

It would work fine for some, though.

Rusher

I would hate to be waiting for 1000 other people to submit their move

Einacio

16:31

my idea was to try first with the simpler ones, like Battlebots or Easter Egg Hunt

Rusher

Uh oh, we might have a problem.

Apparently, deleted questions count towards tag survival. At this rate, the tag will never disappear.

In other words, the only way to get rid of the tag is to untag everything that is code-trolling.

I guess that would be fine, because they are by default a pop contest as well

Volunteers?

Geobits

16:59

Just to be clear, you're asking for volunteers to untag every code-trolling question so the tag will be deleted by script?

An easier option (if it were decided that it was not welcome) would be to historical-lock the current ones and deprecate the tag by setting the wiki to "Don't use this tag, it sucks".

If you just edit out the tag and it gets deleted, there's nothing to stop it from being used again anyway.

If there is a concrete meta post saying "don't use this tag", then future uses could easily be closed as off-topic with a link to that post.

Peter Taylor

It's bad enough that people get the idea that code trolling is acceptable. If they got the idea that that's what the pop con tag is about, it would be even worse.

Before deleting the tag, the questions should be deleted.

And that requires mod intervention.

Geobits

I'm not sure that deletion is necessary in all cases. I see it as similar to old list/pop posts on SO. Most of those (or at least the popular ones) got a historical lock. That's the purpose of the lock, I believe, and it states clearly "dont use this post as an example of what's acceptable". The downvoted ones could clearly be deleted.

3

It would also serve as an easy signpost. Point to a locked one and say, "These aren't welcome here. Notice the lock?"

Sawarnik

17:17

The Twentieth Byte.

Peter Taylor

@Geobits I could accept that as a compromise.

m.buettner

It's funny, because this is exactly what I suggested yesterday but nobody seemed to be getting that. :D

Geobits

The problem with doing that as-is is that the only two meta posts I see really discussing it are these: meta.codegolf.stackexchange.com/q/760/14215 and meta.codegolf.stackexchange.com/q/812/14215 . The most upvoted answer for each says "no, don't kill it with fire". For a historical lock/ban to work, you have to have something that actually says we don't want it.

Otherwise it will just be seen as whim-of-the-high-rep-users, which is not a great perception for users to get.

Peter Taylor

17:37

> For a historical lock/ban to work, you have to have something that actually *says* we don't want it.

There's http://codegolf.stackexchange.com/help/on-topic

Geobits

I assume you mean the "objective winning criteria" bit, but the next obvious argument is how to differentiate that with [pop-contest]. I'd argue that trying to rid the site of those would be much harder, but the premise is similar in regards to objectivity. If you meant a different part in particular, feel free to correct.

Einacio

how about codegolf.stackexchange.com/help/dont-ask?

m.buettner

17:53

But that's literally exactly what I was getting at yesterday. If we want to close those as "not a good fit for this site" we need an official decision on meta for that. Maybe Doorknobs future post will help with that (see the 3-star post in the list of starred posts on the right). But if the official decision remains to be "don't kill if with fire", then said high-rep users should accept that decision as well and instead try to figure out what makes a good code-trolling question.

but one step at a time...

Geobits

The thing is, I really don't know. This is one of those hard problems where community doesn't (seem to) agree with the opinions of higher-ranking users. The more you flex the rep-power, the less you're seen a "community-driven" site. Then again, sometimes the masses are wrong. I'd imagine that's what mods are for (to help with this arbitration), but our mod presence is... not always available?

Rusher

18:09

Isn't code-trolling just a variant of pop contest?

Geobits

As used now, basically. Code-trolling doesn't have to be pop contest (as stated in the wiki), but I fail to see a useful scoring measure other than that.

I don't believe any other criteria has been used for one yet.

Oh, except for this gem:

> Note : the most surprising, imaginative and trolling answer will be marked as accepted.

Peter Taylor

18:48

@Geobits That, and also the bit about having a "clear specification of what constitutes a correct submission".

They also fail the "Too broad" test, which may be what @Einacio is pointing at.

Rusher

I don't mean to open fifty cans of worms at once, but many problems could be solved at once by only allowing one type of tag for this site.

"A tag should be a winning criteria that can stand alone without any other tag."

Geobits

I get that many could be called too broad. That's not an argument for killing the whole tag, though, unless you're saying that it's not possible to have a clearly specified code troll. It's a good argument for individual questions, though.

@Rusher Then [image-processing] shouldn't be used? Or [math]?

Rusher

That would mean code-trolling can no longer piggy back on pop contest, all meta posts would have to actually be moved to meta (yea, I mean you you code tips questions), and what is left would be a nice collection of relevant tags.

Geobits

Limiting each question to one tag and limiting them to a set list defeats the entire purpose of the tag system.

Rusher

Nah don't limit them. Introduce all the new ones that you want. But they have to be able to stand alone as the only tag on a challenge.

This is a site wide stack exchange policy, not something I just made up

If a tag can't stand alone and requires another tag (or set of tags) to live, then it shouldn't exist.

Geobits

18:56

I think that's a bad interpretation of the policy. Clearly if you are going to require certain tags (criteria), then a literal interpretation says that all other tags are invalid by default, as they cannot stand alone on a question.

Rusher

A quote from Jeff Atwood since you think I misinterpreted it: From this point on, meta-tagging is explicitly discouraged.
How can you tell you’re using a meta-tag? It’s easier than you might think.

If the tag can’t work as the only tag on a question, it’s probably a meta-tag. Every tag you use should be able to work, more or less, as the only tag on a question. Meta-tags, like [beginner], [subjective], and [best-practices], are useless by themselves — they tell you nothing at all about the content of the question.

Geobits

@PeterTaylor I kinda assumed that, but the consensus that I see on meta is that it is possible to have a good code troll.

Peter Taylor

@Geobits I am saying that.

Rusher

That's what I tried to say, but it was either Doorknob or PT that said I took Jeff Atwood too literally

I mean, the part about standing alone is in bold. I don't see how I misinterpreted that

Peter Taylor

If you take it literally then you would ban all but 4 tags in every single meta stack

graph-theory is not like beginner

Rusher

18:58

@PeterTaylor Name the four and Ill name a fifth that can stand alone

Peter Taylor

bug feature-request discussion support

The software enforces that you include one of them

Rusher

I guess meta is another story

Geobits

@Rusher I remember that conversation before, and I said the same thing I'm saying now. That comment by Jeff is not applicable to sites that have mandated-use tags (like metas and ppcg).

Or at least not to be enforced literally.

Rusher

@Geobits But doesn't the reasoning behind it still apply, even if to a lesser degree?

Geobits

No, because I like to be able to search for [ascii-art] if that's what kind of challenge I want to do. Your proposition would ban that.

Rusher

19:02

Actually, you could search for "ASCII art" and accomplish the same thing

Geobits

Or [graphical-output], [math], [encryption], or whatever.

I can't add it to my favorite (or ignored) tags, though.

Rusher

You also can't add beginner to your ignored tags on SO

Geobits

So are you honestly proposing that we eliminate every non-criteria tag? If so, I'll just stop.

Rusher

I'm not going to use "Can Geobits add it to his list" as a criteria

Geobits

If not, I'll probably also just stop.

Rusher

19:05

The reason you are reacting so extremely to my proposition is because nobody stopped the garbage from piling up before it got out of hand. Now it's commonplace, and we have crap like tips tags and tags like code-trolling that are extensions of other tags.

Geobits

No, I'm reacting negatively because you might as well not have tags at all. There's a reason the software supports up to 5 tags on a question.

2

Rusher

So, if I create a challenge involving bicycles, should I create a tag called bicycles?

If fifty people write challenges about zombies, should we create a tag called zombies?

Geobits

I can weigh the merits of each individual tag without making a blanket statement to destroy them all. Can you?

You added a [java] tag to your KotH question. Why? It can't stand alone.

Rusher

I obviously should not have done that.

Geobits

The point to tags (IMO) is that users can see roughly what a question is about before clicking on it (sometimes the excerpt isn't that helpful). For instance, if I don't like kolmogorov-complexity questions, I can avoid wasting my time. If it were just tagged [code-golf], that wouldn't be possible.

Rusher

19:18

If you don't like beginner questions on Stack Overflow, well you are just SOL.

Geobits

I fail to see how that's relevant....? [beginner] doesn't tell you what a question is about, whereas [kolmogorov-complexity] does.

Rusher

I disagree. Both share an "is a" relationship with the question.

Geobits

Yet beginner is subjective. What one person think of as a beginner question may be advanced to another. My example is more objective. The question is either about [kolmogorov-complexity] or it isn't.

Rusher

So only subjective meta tags should be removed? I could get on the side of that idea.

That's at least better than saying "You read it wrong and your idea is terrible.", which is all I got last time I brought this up.

Geobits

> If the tag commonly means different things to different people, it’s probably a meta-tag.

That's from blog.stackoverflow.com/2010/08/the-death-of-meta-tags . If you read it, you can clearly see what kinds of tags they're talking about.

Particularly:

> The reason meta-tags are a problem is that they do not describe the content of the question.

Rusher

19:26

Alright. I agree with you. Jeff is clearly describing subjective tags in your first quote.

Geobits

Call me impatient, but waiting for results is not fun. I've about lost interest in the challenge already, and judging hasn't even begun yet.

Peter Taylor

@Geobits Where do you see that?

Geobits

One example is meta.codegolf.stackexchange.com/a/761/14215 , with 30+ votes

Peter Taylor

19:44

One problem is that most of the answers in the threads about trolling contain several ideas. It's not clear how many of those 32 upvoters agree with all of them.

I don't think there's really a consensus about trolling at all. Compare the +13/-5 for this opinion that all of them should be deleted with the +8/-14 for this one.

Geobits

That's a good point, to be honest. The question then remains, what do we do when there is no consensus?

I'd argue that doing nothing (and having constant close/open cycles on each new post) is more destructive to the site than picking one of a) allowing them, or b) banning them.

It's absolutely not fair to users as it is, who have no idea what they need for a successful post. Even if they look around at other CT questions, they have no way to know if it will be well-received or voted into oblivion.

Rusher

20:34

@Geobits Simply doing nothing and also not talking about it so much would alleviate some of the constant close reopen cycles. For example, every time I vote to close a CT question, @Quincunx or another user gets vocal and says "Why was this closed?" and then asserts that it is a perfectly valid question. Then, the question gets reopen votes in response, and then someone brings it here and declares war against it.

I admit that I have done the whole "CV this plz" thing, but I won't be doing it again after a discussion I had yesterday

Quincunx

22:03

@Rusher I only do that when I feel that those questions should not be closed. And I've only done that before the question was closed. I actually wasn't complaining about the question being closed, only about the incorrect close vote.

Rusher

@Quincunx Well I'll make a deal with you. If I feel that one of the listed close vote reasons does not fit or is not specific enough, I will use a custom reason or I will follow up with clarification in a comment. If it doesn't sit well with you, you can cast a reopen vote and clarify in the comments if needed. This can all take place without verbally challenging each other in the comments.

And slowly, this behavior can spread to the people who close and reopen out of spite.

This will avoid the "pity" votes that come from people who are like "I agree with Rusher, so I'll close vote it too!" or "I agree with Quincunx, so I'll reopen it too!"

Less publicity = less rollercoaster voting.

grovesNL

23:19

At what point do the "how slow is this python code" questions become duplicates?

3

They are all just reposts of each other with slightly altered details. I may be biased because I find the challenge pretty uninteresting as the entire function of the code is pretty meaningless.

Transcript for