The Nineteenth Byte

12:10 AM

@Geobits Okay, I'll try it starting with both sets empty and [] as the first answer. I'll give it a try tomorrow morning, playing as perfectly as I can (only looking one step ahead). Would you mind giving it a go as well, if you find some time?

Geobits

Sure, but I can tell you already that I typically don't enter regex challenges for a reason, so me saying "this is f*ing hard" may not mean so much ;)

If I say it's easy, though....

Martin Büttner

lol yeah, but a lower bound would still be useful :P

if you want you could even flip a coin for deciding which set each regex goes into ^^

Martin Büttner

12:30 AM

@Rainbolt how is this not still "unclear what you're asking"?

2

COTO

@xnor The two things I don't understand are the heuristics by which the art terminators judge a challenge to have "too much" art in it, and why programming has to be the unchallenged principle skill in winning the contest.

The consensus in the meta thread came across loud and clear that programming needed to be a principal component of any challenge, but not the principal component of any challenge. I happily upvoted the answer since this cutoff is perfectly logical. I can't think of any good reason why we should further restrict the rules and throw out an enormous class of potentially wonderful, creative challenges.

Nor can I understand why, given the conspicuous absence of rules in the help center, we can't let the appropriate mechanism of community content voting dictate whether such challenges live or die. We literally have to wipe these challenges out of existence.

The answer "the public can't be trusted" is closed-minded and elitist. It leads to dead sites with only a handful of active users. I'm absolutely not the "come as you are; it's all good" guy, but there needs to be balance. PPCG significantly lags the SE minimum questions-per-day quota, and 50% of the precious few questions that are asked get closed in a heartbeat.

According to the helpful stats posted by @xnor, 80% of closures are imposed by 15 users (quite possibly the exact 15 reading this thread). 80% of re-openings, which are significantly rarer, are implemented by a (nearly) completely different set of users.

Has it always been this way? Has the PPCG question rate always been what it is now, with the same high closure rate, and just kept right on truckin'? Have there always been 30-40 users responsible for 90% of the activity on the site, and just 15 responsible for liquidating off-topic content? If so, please point me to some statistics proving it. I'll shut up and take my medicine. If not, then I formally submit this rant for your consideration.

End COTO Rant v3.0.1

Rant terminated successfully after 3.551 seconds.

Geobits

12:52 AM

@COTO What those stats also say is that nobody else votes. I can't help but notice that despite your apparently strong interest in what gets closed or not, you don't seem to be voting much (at least according to the stats I've seen), despite having the privilege to do so for around two months. Somebody has to do it, and I don't see how railing against the only groundskeepers this site has while not pitching in yourself accomplishes anything.

Plenty of people can cast those votes. If they don't, that's a vote, too.

COTO

1:33 AM

@Geobits: I'm not sure why I don't appear on either list. I've voted 4 or (I believe) 5 times to close. I've never voted to reopen, except on my own art challenge just now. And indeed I'm aware that somebody has to do it. I've joined others in the past expressing appreciation for the work it takes.

Geobits

Yea, those stats only show successful votes, which makes them somewhat less useful.

I noticed that after posting them, go figure.

COTO

@Geobits: It's got to be more than that, because I'm quite certain that everything I've voted to close has ultimately been closed. The page indicates that results are cached, hence maybe we're seeing old data that somehow isn't being reflected in the numbers.

Geobits

Could be that, too. Normally the cache is weekly (IIRC).

hosch250

@Geobits Re your comment on my post: There was never an answer 0. The answers started at 1.

Oh, never mind. I get it.

We are using mod to calculate it.

Bother :(

COTO

@Geobits: I do appreciate users keeping the site neat and tidy, and keeping users on the straight and narrow. I'm not trying to be a malcontent. The art thing just bothers me. You can note in my conversation with @xnor last night that I've seen some absolutely brilliant open-ended art code golf challenges in the past.

It just seems to me that if we were to loosen up restrictions a little, allowing a bit more artistry in would be a good place to compromise.

And I'm a bloody engineer. :D

Geobits

1:40 AM

@COTO Yea, I get that. It just gets me when people complain about voting without (seemingly) putting in the votes to reverse it.

COTO

I shall be more diligent in future to cast such votes.

Geobits

Also, I may be a bit biased on this, but.... you don't downvote :(

You ask for us to let the community voting determine the life/death of these kinds of things, but downvoting is critical to that.

There are a lot of people here that don't DV much (or at all), so many times the absolute vote count is very lopsided.

COTO

I've been told many times that voting is subjective. I upvote what I like, I vote to close what the rules say shouldn't be there, and I leave what I don't like be. Apply Kant's imperative to this M.O.: if everyone upvoted and nobody downvoted, the answers would still ultimately be sorted in order of community preference. There would just be a DC offset. :)

hosch250

@Geobits What does the delimiter mean in the first 99 Fibonacci numbers?

Also, how should I print number 99? A long only holds up to number 46.

Using C++ this time.

Geobits

@COTO Oh I agree they're subjective. I'm not saying you have to vote, but if you want voting to be the main action/judgement (rather than closing), participation should be both ways (up and down).

But I'm not so sure that your hypothetical works. If a question has +21/-21, I don't see it the same as one that has +21 only (simply because you can't DV).

It's clear in one case that it's a controversial issue, because the signal is there. Without down votes, you lose that.

COTO

1:50 AM

@Geobits The controversy is opinion-based. And there's a reason you can't cast ballots against candidates in public elections. People dislike all kinds of things for all kinds of reasons.

Geobits

Of course the controversy is opinion-based, isn't that what you want when you say "the appropriate mechanism of community content voting"?

COTO

I dislike the gibberish code golf languages, for example. To me they're too close to compiled code. But I'm not going to downvote them because of that. And if they're particularly clever, I'll upvote them.

Geobits

What if an answer knowingly breaks a rule?

COTO

@Geobits I mean that unpopular questions quickly drop into obscurity, while unpopular ones don't. And the popular ones invariably accrue more votes.

@Geobits I like good satire, hence I won't delete those. If others want to, more power to them. But anything else, yes.

Geobits

Hmm. Let me put it this way: When I see a negative score on something, I look at it. It's a signal that something might be wrong with it. I may not agree that it deserved it, but by drawing my attention to it, it caused me to make some sort of judgement on it.

Whether good/bad/neutral, it doesn't matter. Another person looked at it.

When I see a 0/1/2 score, that doesn't necessarily happen.

COTO

1:55 AM

@Geobits What are some examples of things you'd downvote but not vote to close, delete, etc.? Can I see that in your profile?

Geobits

Questions or answers? I downvote a lot.

COTO

Answers.

Geobits

Glad you chose that, I wrote this on meta already for those :D

Almost all of my down votes on answers are in these categories: doesn't work, breaks rules, doesn't meet spec (in some other way)

COTO

Hence you use downvoting as a substitution for delete votes? I have no problem with that.

Geobits

I think deletions are for a different purpose. I don't see it as a substitution at all.

COTO

2:00 AM

In that case, I have "downvoted" 5 times, I just voted in the slash n' burn "delete/close" way.

I delete anything I feel breaks the rules. If an answer is junk, or shows no effort, or doesn't answer the question, etc.

Maybe I'll switch to just downvoting instead.

Geobits

Well, the thing about deletion is that I believe you can only vote to delete negative-scoring answers, so using deletion only instead of downvoting seems backwards to me :P

Calvin's Hobbies

@Geobits Yes! I was just about to say the same thing.

COTO

Fair enough, I'll try to be more diligent in downvoting rule-breaking answers and questions.

OK. Well, I'm off to dream up a new challenge. Thank you gentlemen for the conversation. Cheers and goodnight. :)

Calvin's Hobbies

Off topic: Can you always sort a list of 2 or more distinct integers with only these 2 operations: (A) swapping the first and last element, (B) moving the first element to the end of the list? I want to say yes but I'm not certain (no it's not homework).

Geobits

I swear I saw that on SO at some point where the answer was yes...

Wouldn't you basically do a bubble sort? If you move the first to the end and swap, it's the same as swapping the first two. You can do that repeatedly until sorted, right?

Wait.. something doesn't seem right about that.

xnor

2:22 AM

@Calvin'sHobbies Yes. What you're asking is whether the symmetric group is generated by a cycle of all the elements and a transposition of two adjacent elements.

by cycling, swapping the first and last elt, and cycling the rest of the way, you can swap any two adjacent elements

by a sequence of swaps of adjacent elements, you can move any single element any position

and from there, you can move each the lowest element into the lowest position, then the second-lowest element into the second position, and so on until the list is sorted

Calvin's Hobbies

2:36 AM

@xnor Thanks! Makes perfect sense :)

xnor

are you looking to make a challenge where you sort like this?

Calvin's Hobbies

Yes in fact. Presumably doing it optimally (with fewest A B moves) is tricky, so that might be the scoring criteria.

xnor

interesting

you need a large enough number of elements though that it can't just be brute forced finding the shortest path on n! nodes

Calvin's Hobbies

Right, I'd have time limits.

xnor

i'm curious how many operations you need asymptotically

it's at least n lg n because each sequence of A's and B's encodes a permutation

and it's at most n^3 because you just can naively use the proof i gave

you can do O(n^2) if you're careful

Rainbolt

3:08 AM

@MartinBüttner I clearly understood what he said. He wants a program that toggles another program on and off. You guys are clearly making that harder than it sounds.

xnor

3:47 AM

is there anything to stop a question from repeatedly flipping between getting closed and reopened?

codegolf.stackexchange.com/questions/40713/… might get closed again

Rainbolt

I want to ask everyone to reconsider their close votes in light of the complete rework I just did of Code to toggle a program

xnor

looks ok; I retracted my close vote and downvote

i'm not actually familiar enough with processes to know if everything is specified that needs to be, but i trust others will

Rainbolt

@xnor I am very familiar with starting and stopping processes in C#, and I'm very confident that everything is there. If you want reassurance, you can barely read the fist two sentences of msdn.microsoft.com/en-us/library/z3w4xdc9(v=vs.110).aspx

It takes "The friendly name of the process."

xnor

thanks, but i trust your and others' judgment on this

Rainbolt

Well, it somehow got another close vote. I'll campaign for it in the morning, or I'll just sandbox it myself and make it happen.

Night all :)

xnor

4:01 AM

good night!

so, codegolf.stackexchange.com/questions/40713/… got closed a second time

is there really nothing to stop a question from flipping back and forth if at least five users favor closing and five favor reopening?

the same five people can keep voting to close or reopen, right?

this is really silly

grc

it was closed by 5 different people this time

xnor

is that incidental, or can previous close voters not close vote again?

grc

you can only vote to close once

xnor

ah, and same with reopen i presume?

grc

yes

xnor

4:13 AM

that's a funny way to determine the majority

eliminate groups of five from each side until one no longer has five

though realistically, people are likely to just stop paying attention to the question

grc

it's already got a reopen vote again

xnor

that's me, sorry :-P

we really do need a better system for this though

i wonder if it's been discussed on meta.se

something where people can vote to keep open or close immediately, and whenever it reaches 5 in the given direction, it closes/reopens

closest i could find: meta.stackexchange.com/questions/210757/…

of which the top-rated and only response basically just says a moderator should deal with it

grc

4:30 AM

I think these kinds of posts are rare enough for mods to deal with

New Sandboxed Posts

6:49 AM

0

A: Sandbox for Proposed Challenges

Elevator Controller Sandbox Note I am posting this in the sandbox because the specification is far from complete and I need a place to keep working on it. So please don't be surprised that things are still missing. However, this also allows anyone to give early feedback or ideas, which of co...

Quincunx

7:08 AM

Ugh, I'm taking a beginning C++ programming class, although I already know C++. It's saddening all the bad practices it teaches. For example, it teaches using namespace std; as something "every program needs" (or else it can't run). It also teaches using the new operator to instantiate objects when a pointer is desired, but as far as I know, using new without a corresponding delete somewhere results in a memory leak. I don't think vector<object*> will auto-delete pointers added to it.

Calvin's Hobbies

7:45 AM

@xnor codegolf.stackexchange.com/questions/40744/…

Martin Büttner

10:34 AM

@Rainbolt Well, judging by Peter's 5 unanswered question, the spec was still very much ambiguous. I'll have a look at your revised version.

ProgramFOX

11:04 AM

Does anyone actually check the "delete / undelete" tab on Meta?

Sometimes Community♦ auto-deletes an unanswered question that's downvoted while I think that question might still get an answer so I vote to undelete, but it never gets undeleted unless I flag it.

Martin Büttner

@ProgramFOX where is that?

ProgramFOX

@MartinBüttner Here: meta.codegolf.stackexchange.com/…

Martin Büttner

oh that

I almost never check the mod tools

ProgramFOX

I think this question is a perfectly valid and answerable question but it was auto-deleted because it had a downvote, could anyone cast an undelete vote please?

Martin Büttner

done

ProgramFOX

11:11 AM

thanks

Martin Büttner

@Rainbolt okay looks good now

New Meta Posts

11:25 AM

0

Q: Code in Spoiler broken (again?)

This one looks pretty funny actually: Found in this answer.

Sp3000

@MartinBüttner Hm? Works in my Chrome though... :/

Remember last time how I tried to put a list in a spoiler? Well I played around with <pre> and it kinda worked

Martin Büttner

Ah I wanted to put OS and Chrome in there

New Meta Posts

1

Q: We Have a Messy Sandbox - Follow Up

In We Have a Messy Sandbox the consensus was that the best way to clean our Sandbox is to delete posted challenges but keep abandoned ones. The consensus also seemed to be that: These deleted posts should be edited down to a simple link to the real challenge (presumably with the title of the r...

discussion sandbox

Martin Büttner

12:06 PM

@Doorknob great work! :)

Doorknob

12:22 PM

Thanks ;) Keyboard shortcuts made this a lot less painful than it would seem.

Martin Büttner

heh

@Geobits An alternative to scaling the denominator would be to raise each answers score to a slowly increasing power. The advantage would be that there is still a hard limit on the total regex size, so that I can be sure that this gets really hard at some point.

Of course, I'd need to figure out a reasonable scaling for said exponent, but getting that slightly wrong might not be as bad as getting the denominator wrong.

Geobits

1:04 PM

Interesting. That seems like it would work okay.

Also: I played around with the regexes last night. I wish I could say otherwise, but I ended up getting frustrated/bored after 15-20 of them and hatched some Eevee eggs instead. Still no shiny :(

Martin Büttner

heh

do you still have the 20 regexes around?

Geobits

Well, not here, but I think I still have it open at home.

Martin Büttner

do you remember how long they grew?

Geobits

20 chars or so. I'm bad at this :D

Martin Büttner

hm that's alright... I'm at 10 patterns now, and the longest has 12 bytes

Geobits

1:09 PM

I'd guess you're better at putting them in the right sets to make it harder on yourself, though.

Martin Büttner

possibly... although I just always decided depending on whether the regex matched itself or not... if it did, I put it in the fail set, if it didn't I put it in the match set

btw, this also means, that people could churn out a bunch of answers by cooperating (and putting it in the easy set on purpose), but at least in this case, others could easily thwart that by chiming in (and I'm playing, too :D)

hm, I'll try with a different initial string now

Rainbolt

1:29 PM

Success!

Martin Büttner

?

Geobits

Eureka!

What are we talking about?

Martin Büttner

don't leave us hanging!

Rainbolt

A challenge that I wanted to happen is happening.

It will be good for the future of KotH controllers everywhere :)

And when I say good I mean short, not good.

Beta Decay

Which challenge is that?

Rainbolt

1:40 PM

4

Q: Code to toggle a program

Challenge Write a program that toggles another. Your program should accept exactly two command arguments. Specifically, your two tasks are: If the names of any running processes are identical to argument one, close them and exit If the names of zero running processes are named identical to arg...

code-golf

I pretty much rewrote the entire thing after four close votes had already been cast.

Martin Büttner

they're still not revoked though

Rainbolt

Well, one was revoked, and a new one has been cast.

Beta Decay

What's the meaning of toggle in the case? Is it just start and stop?

Rainbolt

@BetaDecay See "Specifically, your two tasks are:"

Geobits

start/stop, on/off

Beta Decay

1:43 PM

Ahh. This sounds good then.

Geobits

It definitely looks cleaner than it did when reopened ;)

Rainbolt

Can someone help me understand Dennis's comments on that challenge? What is the difference between a "real problem" and a "challenge based on a real problem"?

And if it isn't a "real problem", does that make it a "fake problem"? Are all of the problems we solve here fake?

I don't want to get into with him in the comments on the challenge. I was probably overstepping my bounds with the one comment I already posted.

New Sandboxed Posts

0

A: Sandbox for Proposed Challenges

Let's talk about this... I am planning on hosting a King of the Hill challenge in which bots will have to coordinate each other in order to be successful. The idea is to play a Diplomacy-like game between bots. The engine (still in development) will start the bots and communicate with them via s...

Geobits

I thought it was best summed up by his "...if you're actually facing a problem, asking for help on PPCG is a bad move...". If the OP was looking for production code, then advising him ihe wouldn't necessarily get that here is a nice thing to do IMO.

It was hard to tell from the original revision.

Rainbolt

How does "must run on BASH or Linux" equate to "production code"? He even had it tagged Code Golf

Geobits

1:52 PM

The original didn't have that either. The entire text was:

> Create a program that toggles another. It should take 1 argument and, if it is running, close it. Otherwise it should open it. It should work with options or arguments to the program.

It looks almost like he pasted his homework.

Rainbolt

It's ironic that everyone thinks that it's both unclear and a homework question (which would typically be well specified)

Whatever suits your close voting needs at the time I guess.

Geobits

Meh, I didn't vote to close it, just looking back at it now.

Martin Büttner

@Rainbolt Most homework questions I remember from programming classes were certainly less precisely specified than we require for PPCG challenges.

Geobits

However, I've seen some shockingly ambiguous homework pasted on SO, so it's not like it can't be both.

Rainbolt

I'm just having a hard time figuring out where people got the idea that this specific problem was a homework question. The reason I care is that it almost certainly got at least one person to close vote it for that reason.

And after Geobits said that the Bash/Linux restriction was added after the fact for clarity, I'm just flat out convinced that there is no evidence that this was homework, except for the fact that it has some real world usefulness.

Since there is no inherent problem with real world usefulness, I think the challenge received some unwarranted negative attention.

Thus, the attitude changed from "Let's make the challenge better" to "Let it die"

phew now that that's out of my system I can go back to work.

Geobits

2:08 PM

It was closed even before the bit about Bash was added, so that part definitely wasn't the issue with the original closing. It could be that Peter asked a bunch of questions and none were answered.

The second closing, maybe. I dunno.

Martin Büttner

Rewrite:

10

A: Sandbox for Proposed Challenges

A Different Kind of Meta Regex Golf domino-coding code-golf regular-expression I just wanted to record this idea here before I forget it. However, I will hold off posting it until I'm convinced that domino coding is actually a good idea for PPCG (which I'm currently not). Feedback under the pre...

Geobits

Then again, I don't see closing as just "let it die". It's supposed to be used to support "let's make it better", but stop answers from coming in until it's made better.

3

It's also used as "let it die" in some cases, to be sure, but I don't see why people seem to think it's the kiss of death. Reopenings really do happen :P

Rainbolt

@Geobits You know very well that the 9 close voters have made zero effort to improve the question. I can actually prove it using the revision history.

The attitude of "Ask questions, close as unclear" really needs to stop when the answers are close enough to obvious that you could just edit it yourself to make it clear.

And seriously? Peter asked what it means for a program to be "running"? He was obviously gunning for it to be closed by asking stupid questions.

Peter Taylor

@Rainbolt, in my book spending a few minutes to produce a list of missing details is making an effort to improve the question.

Geobits

Sure, and I almost rewrote it myself yesterday when I saw it was headed that way. I didn't have the time to do it right, though. But I didn't say that if you cast a close vote that you have to be the one to make it better.

A close vote literally means "This question is not fit for the site in its current form" to me.

Rainbolt

2:17 PM

@PeterTaylor You went overboard with your "six questions". Half of them weren't relevant.

Martin Büttner

@Rainbolt I'd say that's up to the OP. We can't know his intentions, so asking for clarification in specific points is making as much effort as I think you can expect.

Geobits

The main problem is that people take this shit too personally. So a question got closed. Fix it, ask for help fixing it, or move on.

A close vote isn't "you suck and we don't want your puzzle here", but people seem to take it that way.

Rainbolt

Ok, so my original point here was not that you shouldn't close vote. It's that I think the question was treated unfairly.

Geobits

Is the question better now? (I'd say so)

Rainbolt

And I actually had an agenda other than reopening this specific question. I for one don't want to see "real problems" auto closed.

But if everyone here is going to pretend that it didn't play a factor, then I have nobody to talk to.

There's always a secondary close reason in the back of your heads that you are ready to throw out to avoid admitting why you really closed it.

Martin Büttner

2:20 PM

Or you could accept that the people in this room did close it for other reasons.

Rainbolt

I accept that some people closed it for other reasons.

You need to accept the same thing.

There's a comment on the question itself campaigning against real problems.

Martin Büttner

Oh yeah sure, some people might have closed it for looking like homework, but there is no point arguing with us here, if we closed it for being unclear.

Geobits

That wasn't brought up (at all) in comments until after it was closed.

Rainbolt

Plus, it received a close vote after I fixed it, despite the fact that I answered all of Peter's questions expliticly.

Peter Taylor

@Rainbolt Every single one of my questions was relevant to whether or not the specific ideas I had for how to implement it were permitted or not.

Martin Büttner

2:23 PM

But I don't think you're arguing with the person who cast that close vote.

Rainbolt

2:36 PM

@PeterTaylor Saying that all six of your questions were relevant doesn't suddenly make it true.

In fact, the question as currently defined doesn't say anything about what it means for a program to be "running", because "running" is a well defined state. Every programmer knows what "running" means.

So that's at least one question that you asked that wasn't relevant.

"Does it have to have been started by the toggler?" is also an irrelevant question. The answer was "No." before you even asked it.

Rainbolt

3:03 PM

How to get the transcript moving quickly: Have an unpopular opinion.

Peter Taylor

@Rainbolt, you clearly have more confidence in your mind-reading ability than I have in mine.

Rainbolt

@PeterTaylor I think you completely missed the point of what I said. It's not that I am able to read minds. It's that if the spec doesn't say something should be there, then it shouldn't be there.

@PeterTaylor You could have asked "How do I know if this challenge is limited to GolfScript only?" The OP is not obligated to even acknowledge that question.

Do you understand what I'm getting at?

Peter Taylor

No.

Rainbolt

@PeterTaylor If the challenge was flawess and then I asked "Is this challenge is limited to GolfScript only?" and the OP didn't answer my question, would you vote to close as unclear?

In other words, does asking a question automatically make it a relevant question?

Peter Taylor

I think you're deliberately being silly.

Rainbolt

3:13 PM

@PeterTaylor I had to exaggerate because you didn't understand my more realistic example.

Please consider it only a hypothetical example that I used to shine light on the real situation.

Nowhere did the original spec say that you only had to worry about processes that you started. Therefore, you automatically do have to worry about them.

What you were asking for was basically an anti requirement. You can't possibly list all of the things that aren't required.

I don't know why I even bother responding to you. You resorted to calling me a silly mind-reader rather than addressing what I said. Why do you always have to be so arrogant and dismissive when you talk to me?

Peter Taylor

I don't think the original "spec" was a spec at all: it was a requirement which needed more detail to become a spec. There are various existing models of daemon control, and no-one other than the OP can know a priori which ones he knows exist and therefore are what he's expecting.

Rainbolt

@PeterTaylor I understand that you had some valid concerns about the original spec. I wish you had enumerated only those.

Here, let me help you turn your irrelevant question into a useful suggestion. Try "I think that we should only be held responsible for closing processes that we started."

It's still irrelevant, but now it doesn't look like a concern that must be addressed in order for the question to be valid.

Great. Now the question is closed again.

Apparently, some users have concerns that haven't been addressed, but because they aren't being vocal about it, I can't possibly know what they are.

Martin Büttner

3:35 PM

well I hadn't voted to reopen yet

I think they might just have seen it in the review queue and thought "sure why not"

Rainbolt

Gee I wonder why they thought that?

It couldn't possibly be all of the misleading comments on the question could it?

If the question fails to reopen a second time, and no visible effort has been made by the OP to get it reopened or sandboxed, I intend to sandbox it myself in about a week.

Gotta do things the way the community wants you to do them, unless you are popular like Calvin

Peter Taylor

3:51 PM

@Rainbolt, if you're looking for arrogant and dismissive comments in this chat thread, you could start by looking at chat.stackexchange.com/transcript/message/18417214#18417214 .

I will dismiss strawmen, because I certainly don't want to waste time and muddy waters by arguing about them.

Rainbolt

@PeterTaylor How did I misrepresent your question? I literally quoted it verbatim and then explained why it was an irrelevant question.

I also acknowledged that some of your concerns were valid.

If you're looking to avoid muddying the waters, why do you ask irrelevant questions?

I'd say that's three times that you have dismissed my concern without addressing it now. Once to call me a mind-reader, once to call me silly, and once to call my entire argument a strawman.

How about this: Does anyone else in this room think that it's unclear to say that a program is "running"?

COTO

4:12 PM

@Rainbolt I don't doubt there are some UNIX/Linux kernels out there that have esoteric application states other than "running" or "not running", such as memory-resident code triggered by interrupts, etc. But for your standard top-level application, which I assume the challenge is targeting, a process is either running with its PID in the list or it isn't.

The question is valid, it just needs to be more specific. Such as: "Is an xyz process considered to be 'running' or not?", so long as xyz is something reasonable.

Rainbolt

I just don't know how we've gotten this far if "running" is ambiguous. I've run multiple KotH challenges where I stated that I was start and stop your submission at certain points of the game. I wonder why I wasn't challenged on this point.

COTO

;)

Rainbolt

I think the commenters on this particular question overstepped what they really meant to say, making it difficult to address all of their concerns. In general, I got a feeling that some users just wanted it closed no matter what. I imagine the OP felt the same way.

Martin Büttner

looks like doorknob nuked the comments

COTO

I already ticked enough people off yesterday, hence I'm going to stand aside for a while. ;)

Geobits

4:18 PM

I flagged for a general cleanup of comments there, since most of them didn't really apply to the edit.

COTO

For all you SE gurus: The ScriptWarz Challenge looks interesting, but it suffers from the same problem as many of the KoTH challenges that all source code is immediately public to everybody, hence it's next to impossible to program a "crafty" bot.

My question is this: is there any reasonable mechanism by which somebody could convey the full code to the OP while placing temporary, placeholder code as an answer until the the challenge was run?

The code would eventually be revealed, but it would at least get to go a round or two without being stripped apart and undermined.

Geobits

I don't see that as a major problem, but a feature. I've seen several non-PPCG tournaments held for tank-war like programs where they were all open.

COTO

@Geobits All of them are open. That's the problem. We couldn't have a closed one if we wanted to, as far as I can tell.

Geobits

Right, I'm saying that's not a big issue (to me, at least).

What's the difference between "the code will eventually be revealed" and "post your code" in the long run? These things shouldn't really be temporary IMO.

COTO

@Geobits You would get at least a few rounds where your crafty bot could do something crafty. The way I envision it, the code would be revealed after the first round it was used in.

Geobits

4:25 PM

Strategy through obscurity? Meh :P

COTO

In the previous KOTH challenge I participated it, it was a trivial matter for users to make slight changes to existing designs that capitalized on highly specific weaknesses.

Geobits

Sure, but that would still happen. You're only delaying it by one run.

COTO

And obscurity is fine, only the OP has disallowed it, presumably because he's going to be running these as executables on his machine and wants to know exactly what they'll be doing.

Anyway, the whole point of "should we do it" is moot if we can't do it. Hence regardless of whether it's a good idea or not, for future reference, can anyone think of some system that could make it happen? We can't send private messages here, as far as I know.

Martin Büttner

You could ask for hashes of the code. The codes (matching the hashes) are revealed after two weeks. Then the tournament is run.

search for "iterated prisoners dilemma"... that one had a system like this

COTO

@MartinBüttner But how to get the code safely to the OP?

Martin Büttner

4:33 PM

by revealing it after 2 weeks... you could only run the tournament like that once

Rainbolt

If the challenge were extremely popular, how many submissions do you think would come back a week later and post the actual code? I'd guess less than half.

COTO

Ah. Now I get it.

Rainbolt

So, you could make privacy optional, and give them the mechanism you described, so that only those who are worried about being undermined can just post a hash and then return later.

And those that don't can just post and be done with it.

COTO

The OP does seem to be "collecting" posts for a one-time run in this case. That's an excellent idea. I wonder if he'd consent to it.

Martin Büttner

sounds good

@COTO Ask him.

COTO

4:35 PM

I shall indeed.

Geobits

Hashing led to overall disinterest in the Prisoner's game I think. The anemic voting on answers would be a disincentive for people to post hashes, and voters aren't nearly as likely to return to them later.

That one in particular has over 1k views, but the highest voted answer is a +2.

Peter Taylor

@Rainbolt, you misrepresented my question by quoting it as though it stood alone, although it was explicitly linked to the following three. I've already addressed the reason that it needed answering in chat.stackexchange.com/transcript/message/18418731#18418731

COTO

@Geobits Yeah, that did occur to me. People aren't likely to vote for a black box, and by the time the code is revealed, few people besides the participants themselves would be interested in the outcome.

Geobits

It really only matters for rep. If your main goal is to win the game, it's not a big deal. Rep is a pretty big motivator for most, though :D

Then again, I have fun looking at other entries specifically to see how I can do better than them (not necessarily targeting them, but better approaches, etc). That's half the fun of a KotH to me, so the black box approach doesn't appeal. To some I'm sure it would.

COTO

4:53 PM

@Geobits Exactly the reason I just deleted my comment in the SW challenge. A bunch of black boxes is too tournament-esque, and cheats any observers. Open source is the way to go.

Well, that misadventure was a fantastic waste of my lunch hour. Cheers all. :D

Geobits

:D

Rainbolt

@PeterTaylor You had a total of six questions, and at least two of them were either unimportant or already answered. The concerns brought up in chat.stackexchange.com/transcript/message/18418731#18418731 are not even remotely related to the two specific questions I mentioned.

In other words, "The spec was poor, therefore all six of my concerns were valid." is a non sequitur.

Dennis

@Rainbolt How can a single comment on a single question be campaigning against real problems?

hosch250

5:25 PM

Someone please post answer number 7 here: codegolf.stackexchange.com/questions/40727/leave-your-mark

I have answer number 8 all ready.

No time to create number 7 now, posted on the run.

Peter Taylor

@Rainbolt, but it's not what I'm saying. What I'm saying is that each of my questions had more than one reasonable answer, and that the answers depended on which model of daemon control the OP had in mind.

Beta Decay

6:03 PM

Pumpkins are so much easier to carve than swedes...

Geobits

@BetaDecay By 'swedes' you mean the food, not people from Sweden, right?

Beta Decay

@Geobits Haha yeah. I doubt I'd be announcing my carving up of a human so readily ;)

Geobits

It sounded like you announced your attempt at carving up humans. I figured you switched to pumpkins because humans were too fast for you :P

3

Beta Decay

Ahh I see. Maybe next year...

Geobits

Off topic from Academia, great comment:

To be honest, I'm more concerned about your chances of succeeding in the real world if you feel guilty for something like drinking a cup of coffee. It's a world of cut-throats out there and you're going to get eaten alive if you don't toughen up. — CaptainCodeman 3 hours ago

hosch250

6:19 PM

What is the general consensus on Evolution of Hello World?

I have an idea, where we take a Hello World program and change it into a calculator that supports +, -, *, and / with just two digits and no person can change/add/delete more than 10 chars at a time.

Don't know if the community will approve though.

Beta Decay

What do you do when you've met the goal of a calculator?

Geobits

What would the intermediate steps do?

(not to mention that basically anything beside "hello world" would be a better start)

hosch250

@Geobits Just be a compilable program in the language.

@BetaDecay Start from there and make a more advanced calculator?

Beta Decay

@hosch250 I think that you should have a starting point, set an objective, once people have reached that objective, change the objective and so on

Constantly changing challenge

Geobits

So it's a single-language challenge, or can my intermediate be something like Whitespace, where anything beside a very limited syntax is ignored and almost anything is "compilable"?

hosch250

6:23 PM

@BetaDecay I considered that, but I thought the community might not like it being changed that much.

@Geobits No, just like the Evolution of Hello World challenge, any language can participate, and we must work from that.

Geobits

Ok. So what's the incentive? How is it scored?

hosch250

We just have a limit on the number of chars to add/remove/change each submission.

How is the Evolution of Hello World scored?

Rainbolt

@PeterTaylor That's exactly my point though. Some of your questions did not have multiple reasonable answers. I explained this thoroughly already.

Geobits

> Once things settle down, the user who submits the most (valid) answers wins. Ties go to the user with the most cumulative up-votes.

(which I think is probably not the best scoring method, but....)

hosch250

OK, that is how this could be too. Or maybe most votes on their submissions (pop-contest?).

Geobits

6:25 PM

What I mean is, why would I care whether or not my edit gets closer to a calculator if it doesn't end up being one?

I could just edit any ten characters to make it compilable in "something".

Rainbolt

@Dennis You actually commented twice, both times outright stating that real problems are not welcome here.

hosch250

Yeah, it needs a lot of work if it will succeed.

Beta Decay

@hosch250 A pop-con wouldn't work a second time. You should come up with a unique scoring system.

hosch250

I see. Yeah, it needs a lot of work.

Supposing I got the problems ironed out, do you think it would be popular?

Geobits

I think that mostly depends on those details ;)

hosch250

6:27 PM

Of course.

Geobits

I just mean there's no way to know without knowing the specifics. It could be a hit or a flop at this point.

hosch250

Yeah. What is the general consensus of Evolution of Hello World? It is awful similar, so...

Beta Decay

I think we need a gap before another challenge in that vein comes out. Too many have come out in such a short time period

hosch250

OK.

Like Code-Troll?

Beta Decay

Well I wasn't here for the code trolling period, but I suppose so

hosch250

6:30 PM

I lost a lot of rep when those were removed.

Geobits

I don't think it's as similar as you seem to. Every entry in that one has a single, specific objective (print hello world). Changing one program into another is quite different, and without knowing the intermediate steps, I don't know how to compare it.

Dennis

@Rainbolt My second comment was an attempt to clarify, since you misinterpreted the first. You also seem to have misinterpreted the second, so I'll try again:

I don't care where a good challenge came from. It can be a real-life problem, it can be something made up. But if you actually need help and intend to use the code submitted in the answers, a code-golf challenge is the worst possible way to solve your problems.

What are you going to do in real life with code that invokes undefined behavior, has O(n!) runtime or is written in GolfScript?

Peter Taylor

I'm not sure that it's been discussed much other than in meta.codegolf.stackexchange.com/q/2415/194 . I think that Martin's answer could be summarised as "I want it to work, but it doesn't fit the software".

Dennis

The shortest answer is rarely the best...

Rainbolt

So the usability of the answers was your main concern? Your comment did not reflect that.

Peter Taylor

6:32 PM

I still think that chat is better for that style of development than Q&A.

3

hosch250

Ah, a flash! I can define a list of problems your program has solve for an intermediate state.

Answers below n: output "Hello World!" or solve any of the problems for the higher answers.

Answers below o and above n: input and output two numbers or harder.

Answers below p and above o: input two numbers and output their sum or harder.

...

Get it?

@Geobits and @BetaDecay

Dennis

@Rainbolt Yes, exactly. I don't remember exactly what my second comment said, but it was something about code-golf answers rarely containing production code.

Rainbolt

@Dennis If I misinterpreted your comment, it is because it looked like a hard and fast rule against real world problems. I didn't want other voters to think that having real world usefulness was a valid reason to close a question, so I stepped in to clarify.

Geobits

That makes sense. I probably wouldn't play (I didn't play evolution, either), but at least each one has an objective.

Beta Decay

Yeah, that'll probably work @hosch250

hosch250

6:37 PM

Yeah. Some people would want a challenge, others just want to program with others for the sheer fun of it.

OK, so I will work the problems out a little better, experiment some to get the ideas, then come back after I have a solution.

Geobits

Sounds like something that may need to spend a few days in the sandbox, too. I'm sure others can poke more holes in it than me ;)

hosch250

Yeah. Hope the sandbox doesn't spoil it by ruining the surprise factor.