The Nineteenth Byte

@Peter, I'm sorry you're right... deleted the answer and CVed... although I'm not sure why you didn't CV. I'd figured I might just delete the answer if the OP clarifies that this was not his intention, but that's not right.

this is beautiful:

Thank you for teaching us how big our world is by sharing so many of your own.

Martin Büttner

10:19 AM

huh, MF went public beta ... I'm quite surprised, seeing how low their activity is

@Optimizer have you given the string rectangle a go in CJam?

Optimizer

i tried

could not seem to get it anywhere near 41

Martin Büttner

well the 41 didn't work

Optimizer

well, mine too

Martin Büttner

just gave it a try with the naive approach

31

now go ahead an beat Pyth big time ;)

Peter Taylor

10:35 AM

@MartinBüttner I didn't CV because I don't think the question is unclear.

Martin Büttner

well at the very least it's missing very important test cases

it also doesn't specify how much of the moon is guaranteed to be visible afaict

Peter Taylor

I don't think you need much, although obviously if you have more then you can get lower errors.

Take the bounding box of the white pixels; pick three widely spaced white pixels on the bounding box; find the circumcentre.

Martin Büttner

sure but if, due to aliasing, the sickle only contains a handful of pixels without much curvature it'll probably be impossible

Peter Taylor

Does Mathematica do interval arithmetic natively?

Martin Büttner

what's that? like [1,2]+[2,3] --> [1,3]?

Peter Taylor

10:40 AM

No, [1,2]+[2,3] -> [3,5]

Martin Büttner

oh, yes it threads most arithmetic over lists

I'd actually used that in my answer

Optimizer

you have an answer in 31 ?

Martin Büttner

yes

Optimizer

all of a sudden I forgot all CJam

what is the naive approach ?

Peter Taylor

Or [1,2] - [2,3] -> [-2, 0]

randomra

10:41 AM

@PeterTaylor that fails on convex non-full moon

Peter Taylor

Ah, true.

Martin Büttner

@Optimizer the one that most of the current answers use. print the top line, loop over the sides, print the final line

randomra

you need two pixel with the biggest distance imo

Martin Büttner

@PeterTaylor yes

wait no

that would be [-1, -1]

(in mathematica)

I'm not sure how you arrive at [-2, 0]

Peter Taylor

Interval arithmetic != zipWith.

Martin Büttner

10:42 AM

okay, then my question still stands: what's interval arithmetic

Peter Taylor

Given that x is in the interval [1, 2] and y is in the interval [2,3], x - y is in the interval [-2, 0]

Martin Büttner

oh I see

I don't think so, but I can check

it does

Optimizer

go ahead and post it then @MartinBüttner

you have my up vote

Martin Büttner

nah, I'm not a fan of winning my own code golfs

Peter Taylor

@randomra That's potentially iffy with narrow crescents.

Optimizer

10:45 AM

@MartinBüttner meh, just wait for APL or something

or Joe

@Peter 2^

fail

let me test something

nope

Peter Taylor

10:46 AM

Circumradius(Interval, Interval, Interval)?

randomra

@PeterTaylor I still select a 3rd point (maybe one with maximal distance from both for best results) and then compute the circumcenter the same way you wanted

Peter Taylor

Yes, that should cover convex and concave.

Martin Büttner

Optimizer

@MartinBüttner no really, post it man. I wanna see

what sorcery it is.

Martin Büttner

it's really pretty straightforward

give me a few minutes to write up a post

@Optimizer done

Optimizer

11:09 AM

ah, slicing form both sides for the sides...

neat. I was getting stuck at that part only

@MartinBüttner L* -> s

Martin Büttner

ah, nice

Optimizer

how come pyth is not using the same algo for a sub 25 score ?

oh, the use of stack here is nice and helpful

Martin Büttner

I think it is

the 37 is using the same approach I think, and Jakube also tried it and didn't get less than that

Martin Büttner

11:59 AM

meh I wanna start a discussion on meta about our tagging practices

mainly that we're not tagging enough, and that some people occasionally seem to be putting effort into getting rid of tags where we should add them... and what we're considering meta tags and what not

but I don't really know what the question is :/

also, what's with all our math tags... which ones are chosen for any particular challenge is completely random... math number arithmetic plus all the more specific ones.

Peter Taylor

You complain that people are putting effort into getting rid of tags and then you give examples of tags which are almost completely useless because they don't tell you anything about the question...

Martin Büttner

sounds reasonable

my personal pet peeve is printable-ascii

Martin Büttner

12:39 PM

@Doorknob I was already on it ;)

New Meta Posts

1

Q: Tagging: are we doing it wrong?

I wanted to post this for a while... First off, I think that we're not tagging enough. Currently, we have an average of 2.1 tags per question. I've checked some (thematically) comparable sites, and Stack Overflow, Code Review and Programmers all have between 2.6 and 3 tags on average. Tags are a...

Martin Büttner

another test (sorry Peter):

interesting

PhiNotPi

1:06 PM

Is it possible to sort a list of all tags by followers?

Martin Büttner

@felix you are Lembik's sock puppet? :P

@PhiNotPi does data.SE have that as a field?

doesn't look like it

PhiNotPi

:(

Martin Büttner

@felix are you eleanora as well? ;)

PhiNotPi

I would say that number of followers is a measure of tag usefulness.

Martin Büttner

that doesn't really help you in deciding how useful a new tag might be, though

PhiNotPi

1:13 PM

If it hasn't already been done, we can define several different categories of tags.

Martin Büttner

there is this

9

Q: The Tag Categorization Project

This is not a question; rather, it is a collection of categories of tags. The purpose of this post is to post information regarding specific categories, such as templates for tag wikis, and also to consolidate all of the tag categories in one place. Please edit the categories below if you find ...

discussion tags

PhiNotPi

Source layout has 0 followers, apparently.

Martin Büttner

seems like an argument that followers are not a good indication :P

PhiNotPi

Apparently.

Martin Büttner

for the record I don't follow any tags on ppcg at all

Geobits

1:16 PM

People follow tags? :)

Martin Büttner

if I'm interested in questions with a tag I just search for them

Sp3000

I'm trying to categorise all the tags and I'm a bit stuck as to where to draw the line between "general themes which don't help at all (math)" and "slightly more specific themes (arithmetic, combinatorics)"

PhiNotPi

Restricted source has 4 followers and 94 questions, Python has 63 followers and 32 questions.

Martin Büttner

I don't even think math isn't helpful... I've actually filtered by it a few times, and I can definitely see people being interested (or not interested) in math questions in general.

Peter Taylor

@PhiNotPi I was thinking along similar lines. pastebin.com/raw.php?i=60f2Pp0q

@MartinBüttner But at a philosophical level, every computer program is a mathematical object, so every properly specified question could be tagged math.

Martin Büttner

1:25 PM

true

Geobits

At a philosophical level, you could argue that they're all [code-challenge], also.

randomra

What I would like instead of math is an advanced-math tag which indicates that solutions need some advanced thinking

PhiNotPi

You could probably do some intense statistical analysis to determine which tags are often posted with other tags. If a set of tags are mutually conflicting (never appear together on the same question), then there's a good chance that those tags are meaningful.

Martin Büttner

@randomra that seems very subjective

randomra

And this is just an idea: it would be nice to search questions based on difficulty generally e.g. by having easy/medium/hard tags (of course this couldn't work like this)

Geobits

1:27 PM

eww :P

randomra

@Geobits for which one? :)

Martin Büttner

Btw, honestly guys, are Lembik's sock puppets something to worry about? It also seems like the felix account has been sort of suspended on math.SE, if you look at the rep.

@randomra this has been discussed

Geobits

@randomra Difficulty thing. You can almost search difficulty by searching the numbers of answers it gets anyway.

PhiNotPi

I think difficulty can be determined by a large number of question upvotes but no answers.

Martin Büttner

7

Q: Tags regarding difficulty

I think that we should tag puzzles (or whatever) with an estimated difficulty tag. That way people who want a real challenge can find it and those who are just starting out can find ones they might be more able to solve. For example consider: Distance between hands on clock vs: Shortest path i...

discussion tags

Geobits

1:28 PM

Well, difficult or boring.

Sp3000

Related: link (by Doorknob on Puzzling)

Martin Büttner

the only real solution I see is to have some sort of star rating on the question for users (maybe even just for answerers). but that would require quite a change in the software, so it's not going to happen.

randomra

@Geobits That's what I do now, high rep users questions with very few answers.

Geobits

I have another related post on Puzzling meta, since people started putting it in the title rather than tags: meta.puzzling.stackexchange.com/q/1437/1563

@MartinBüttner Link to accounts? I haven't noticed him causing trouble with them yet.

Martin Büttner

well I wouldn't know how to tell

randomra

1:32 PM

Maybe something like votes/answers shows quite well the difficulty.

Martin Büttner

math.stackexchange.com/users/66846/felix and math.stackexchange.com/users/134430/eleanora (the latter is a bit of a guess, with the former I'm quite sure)

Geobits

Interesting, but I guess I haven't seen any interaction between them, so I couldn't say.

The mods could see voting times/patterns better, I'd assume, but if he's using them to prop up his other account, he's not doing it very well.

Martin Büttner

well as I said, it looks like some mod already noticed on math.SE

Geobits

Why do you say that?

Martin Büttner

101 rep with loads of upvoted questions. it would be quite a skill to offer bounties to exactly cancel all that rep

(it's actually funny, that I joked about this when felix was online yesterday and asked about binary arrays)

Geobits

1:41 PM

It looks about right to me, based on the rep history. It looks like he throws up a bounty every time he goes over 150, 350 total so far. Use a little downvote for flair, and it shouldn't be too hard to line up rep.

You've got screenshots for a few "nice" rep totals yourself, I think ;)

It's weird, no doubt. Hard to say it's wrong, though, since it doesn't look like he's throwing those bounties on himself.

Martin Büttner

true, but I think Math.SE doesn't appear on the account's SE profile

PhiNotPi

I;m about to plot a questions vs. followers graph.

Martin Büttner

oh no, I'm not saying he's actually using them to boost the other accounts

because we have no way of figuring that out anyway

Geobits

Oh. Then what's there to do or care about?

Martin Büttner

well that was my question :D

Geobits

1:44 PM

I figured you wouldn't be asking unless you thought there was cause for concern ;)

Martin Büttner

well no... I don't know what the general sock puppet policies are... that's why I asked

Geobits

AFAIK, nobody cares unless you start voting yourself up. I'm fairly sure many high-rep users have one.

I made one once to test some stuff with low/no privileges on SO, but I forgot the name to it years ago :D

randomra

I think I even seen a meta on math.se about what to do with people not using their real names so their policy might be quite stricter.

Martin Büttner

haha okay

randomra

hmm, might not be math.se, I can't find now

Martin Büttner

1:49 PM

I wish PhiNotPi's MathJax post had 20 stars... maybe that could convince the devs :D

PhiNotPi

I wish I could star it too.

Martin Büttner

lol

Geobits

I'm not sure that stars convince anybody of anything except that people like clicking stuff :)

Martin Büttner

in excessive amounts they do show community agreement ;)

PhiNotPi

Martin Büttner

1:51 PM

@Geobits that being said, upvotes on feature requests aren't any different in that respect

Geobits

Oh I know. I think stars are the lesser of them, though, since they can't show disagreement, where you can downvote to counteract an upvote.

Martin Büttner

that's true

Sp3000

2:04 PM

I'm going through all the tags and boy do we have some really confused tags

Geobits

"Movie 43 is the Citizen Kane of awful" < I have to see this movie now.

"An execrable waste cooked up by a hell's kitchen of directors and writers. It's death-of-laughter by committee. Its title? Because it's like one of those many asteroids out there—a dismal chunk of rock hurtling through an empty void, without purpose."

Optimizer

any good challenges coming in today ?

anything promoting from sandbox ?

New Sandboxed Posts

2:23 PM

0

A: Sandbox for Proposed Challenges

Navigate My Time Machine challenge-type-undecided This is a pretty broad idea, but time travel is a lot cooler than space travel. The basic idea is that the program will have to sort out the path of various objects through time, given certain constraints as to what must be where, when. There ...

Geobits

That's a neat time chart for Primer, but this one seems more accurate to the movie: xkcd.com/657

PhiNotPi

I think it's a good idea for a challenge. I've never actually seen the move, though.

Geobits

It's good, but better the second time through.

Rainbolt

@Geobits That xkcd chart has Frodo sailing off to the west on a ship

I hate that they did that...

They could have made four three hour long movies and left in everything.

But looking back, Tom Bombadil really was a completely useless character

Peter Taylor

@randomra MathOverflow has a real name policy, IIRC.

Geobits

2:31 PM

Did he not sail in the movies? I haven't seen them since they came out (but have read the books more recently), so I don't remember all the diffs.

Rainbolt

@Geobits He did in the movies unfortunately

No bandits for the movie

They merged entire armies in the movie too

Geobits

There were surely a lot of differences, but you could say that about most epic book/movie pairings. Of course, nerd rage knows no bounds :)

Rainbolt

Well, I think they left out more than usual

Lembik

@MartinBüttner no! We just used the same computer... which is quite annoying

@MartinBüttner felix that is.. what is a sock puppet?

Geobits

@Rainbolt I'd say it's better than the alternative. Look at the way they've stretched the damn Hobbit into three movies.

@Lembik It's when you put your hand in a sock and make it move like it's talking and stuff.

Lembik

2:35 PM

@Geobits sounds painful :)

Geobits

For the sock maybe.

Lembik

exactly :)

@PeterTaylor mathoverflow has no real name policy that I have ever seen enforced. There are plenty of silly aliases there

and plenty of geniuses too

@MartinBüttner sometimes people know each other without actually being the same person :)

@Geobits what was this about?

Geobits

felix's rep on mathematics

TheBestOne

@Geobits The Hobbit movies are a completely different work than the book. The Lord of The Rings movies are actually somewhat based on the story line of the books.

Lembik

@Geobits are you trying to line up my rep with felix's for some reason?

Geobits

2:40 PM

I don't see how you can say it's a totally different work than the book The Hobbit...

@Lembik No. I was saying it looked legit to me....

Lembik

@Geobits ok but I am still confused. What could be non-legit?

Geobits

Have you read the conversation above it?

Lembik

I am trying to understand it.. is the idea that we are involved in some sort of plan to increase our reps?

Martin Büttner

@Rainbolt I thought he also sailed off in the books?

Geobits

@MartinBüttner I'm fairly certain he did.

Lembik

2:43 PM

by awarding bounties to each other? How would that work?

isn't that a zero sum game?

Geobits

@Lembik I don't see how it's not clear that I'm saying the accounts look fine to me.

Lembik

@Geobits no that part is fine :) What I am trying to understand is what it would mean for them not to be fine :)

what is the crime I am being acquitted of?

just nice to know

Geobits

Devious sock-puppetry, I guess. Nobody accused you of anything that I know of, except having socks.

Lembik

I actually have no socks on right now...

(the shame)

Geobits

I thought originally that Martin did, but it doesn't seem that way now.

Lembik

2:45 PM

but I don't understand what a sock-puppeteer has to gain from it?

maybe my mind isn't devious enough :)

Geobits

From bounties? Not much. You can increase rep by upvoting, though.

Lembik

oh ok... but you can vote up as much as you like once you have minimal rep, right?

Martin Büttner

exactly that's the problem

Lembik

in that plan you don't seem to need to do much other than open 100 accounts and get whatever minimal rep you need to upvote in each

Geobits

Well.... yes and no. Search meta.se for "serial voting" to see that while technically you can, there are triggers that help detect it.

Lembik

2:47 PM

@Geobits ah ok

I have to say, you can't actually eat rep :)

Geobits

Can't you? I've been living off mine for some time now.

Lembik

:)

Geobits

Tasty, tasty rep.

Lembik

@MartinBüttner you do remember how I told you ages ago that my computer was shared ?

hence why my name didn't show up properly in chat for a long time

in any case.. let's focus on the tragedy that is LCS :)

Martin Büttner

Just to join in on the LotR discussion, having read the books (*after* the films, admittedly), I have to say that I find almost all of the changes very much justified. Not all things that are great in a book work in a film. And also some of Tolkien's style (in particular, dialogues) is actually occasionally a bit disappointing, and Jackson really improved on that.
What Tolkien was really amazing at is crafting this world with this incredible amount of detail - and Jackson put a *lot* of effort into being true to that vision.

GoT on the other hand is very close to the books, but the few changes they did make seem a lot more random and therefore I found those a lot more disappointing at times.

@Lembik ah yeah, now that you mention it

Geobits

2:51 PM

GoT seems to wander in and out of "true to book" to me. Season one was very close. Two, not as tight, but they veered back a bit midway through season three or so.

It'll be interesting once they catch up and pass the books, though.

Martin Büttner

yeah

I actually watched each season after reading the corresponding book and managed to get spoiled twice :D

Geobits

@Lembik No tragedy, I just don't find it interesting to golf. The O(n) requirement means a tedious algorithm. The comments are spot on IMO.

Lembik

@Geobits ok so let me disagree with parts of that. First the O(n) algorithms are clever, interesting and excting! How can you call them tedious?!

Geobits

Tedious to golf.

Martin Büttner

plus, at least two of "clever, interesting and exciting" are very subjective ;)

Lembik

2:56 PM

@Geobits second.... I think there are two ways to golf. One is syntactical tricks and the other is at a higher level. I have no interest in the syntax tricks either. But is it more boring to golf something clever and sophisticated than someone small and easy?

@MartinBüttner yes.. my views :)

Martin Büttner

@Lembik the answer to your last question is probably yes, especially if I didn't come up with the clever and sophisticated solution myself

2

Geobits

But by requiring certain algorithms (which you've severely limited by the requirement), you take away half of that "high level" golfing, leaving mostly syntactical golf.

Lembik

people could win by just implementing something in their favourite language then running it through an automated golfer of course

@Geobits oh that's an interesting view... you felt there were two few algorithmic options to golf. Well there are about 12 linear time suffix array construction algorithms and at least 3 linear time suffix tree construction algorithms

Martin Büttner

@Geobits I think what I'm probably most disappointed in with the series is Stannis :D

Lembik

@Geobits but the lesson I learn is not to pose code golf questions.. I am clearly very bad at it :(

I just don't see it as a very interesting to write a very short piece of code that is completely useless

it is slightly interesting.. :),.. just not very interesting

Geobits

3:00 PM

@Lembik What can I say? I looked at a couple of the links provided in your post and decided it didn't look fun for me. No big deal, it'll either get answers over time or it won't. It's not like it's sitting at -5 or anything. It's not a bad thing to have some open questions on the site.

Lembik

@Geobits thanks. I find it interesting to find the limits of the site as well. I have had some very very clever answers to previous questions. But it seems that this particular sort of sophistication doesn't work well

I am not sure I could summarise what works and what doesn't yet :)

Geobits

I know what you mean. I still haven't gotten a "good" (totally subjective :) answer for this one, probably for much the same reason.

Lembik

@Geobits that does look nice

I did wonder if the fortnightly challenges might be more focussed on collaboratively solving harder challenges

or if people might think was an interesting model

Geobits

@MartinBüttner Stannis is okay... but he does lack a certain something :)

Peter Taylor

@Geobits The only way to really golf that would be a brute-force approach which generated candidates cheaply and then tested whether they met the labelling criterion. Would you consider that "good"?

Martin Büttner

3:06 PM

in the book (at least initially) it seemed to me that he never really wanted to be king, but just felt this really strong sense of obligation because of how law-abiding he is. whereas in the series he just seems to be some power-hungry dude who wants his candy.

Geobits

@PeterTaylor By good I mostly meant "actually golfed". Maybe it would have been better off as something other than a golf, but when the one solution I've gotten is ~5k of Java including a bunch of debug printing, it's hard to call it good.

It also doesn't compile, now that I look closer :(

New Meta Posts

3:29 PM

0

Q: Can I make a Sandbox meta post instead of using the normal Sandbox?

My next challenge will rely heavily on Stack Snippets and http://api.stackexchange.com/ to gather data from the answers (similar to this or this). I want to Sandbox it to iron out any kinks, but none of the data gathering will work if I answer in the normal Sandbox because answers can't have ans...

discussion sandbox

Rainbolt

@Martin @Geobits This is why I don't recall Frodo leaving at all. The very last chapter has Frodo and Sam going back to the Shire and finding it to be corrupted and taken over by bandits. en.wikipedia.org/wiki/The_Scouring_of_the_Shire

Except that it isn't the last chapter. It's the second to last...

PhiNotPi

The new meta post reminds me of the sandbox subdomain idea: meta.codegolf.stackexchange.com/questions/1846/…

Rainbolt

And yea, he finally sails away in the chapter after that. I wonder if I just skipped it entirely or if I read it and didn't care. lotr.wikia.com/wiki/The_Grey_Havens

Calvin's Hobbies

@Rainbolt Frodo did leave! So does Sam, not to mention Gimli and Legolas.

Just at different times.

Geobits

@Rainbolt Yea, they definitely left that part out, which sucks. It feels somewhat justified to me, since even in the books it felt like a break in the narrative flow.

Rainbolt

3:39 PM

Sam leaves outside of the trilogy though

You have to wonder why three books made three movies and cut out a ton of canon material, and then one book makes three movies with a ton of non canon material.

TheBestOne

@PhiNotPi mbomb's bomber is working well now.

Geobits

@Rainbolt $$$

They might not have known the original trilogy would make as much as it did.

Once they saw that, they decided to milk the crap out of the next one.

Sp3000

@MartinBüttner I posted my train wreck of a tagging answer. Wasn't quite sure what to say, but at least I'm pretty confident what one of our major problems is.

Rainbolt

Oh right. They filmed it all at once. I forgot about that

Martin Büttner

@Sp3000 already reading it ;)

@Sp3000 I like how you sorted [conversion] and [base-conversion] into different categories ;)

Sp3000

3:44 PM

As I note, the last two are a bit of a mess

My biggest problem was where to put counting, sorting and hashing

Martin Büttner

also I feel like [image-processing] is an input type

I usually think of it as the counter-part to [graphical-output]

Sp3000

That makes sense, moved

Geobits

I think most image processing will have both graphical input and output, no?

Martin Büttner

not necessarily

Sp3000

Typically input, not always output I think

Martin Büttner

3:48 PM

e.g. counting grains of rice

Geobits

ah, right

Martin Büttner

or the moon challenge from last night

@Sp3000 I like your answer, but I'm not sure how much I agree with the first bullet point... there's only so much the tag name can tell you without looking at the tag wiki excerpt.

Sp3000

Ideally image-input and image-output or similar would be the most ambiguous, but I doubt they were made at the same time

Geobits

I might have done the moon thing if it were actual photos. That seems more fun, if harder to judge.

Sp3000

Well the first bullet point is basically "don't be like one of the tags I pointed out above", if there's a better wording for that :P

Geobits

3:52 PM

I like that we have the [division] tag in addition to the broader [math] and [arithmetic] tags >.>

Now we just need addition, subtraction, multiplication, modulus, exponentiation, tetration....

Martin Büttner

wow that tag is used 8 times

Sp3000

Yeah I almost had a "way too specific tags" category

But I gave up on trying to work out where "binary tree" and "bitwise" should go

Calvin's Hobbies

Stuff left out off the top of my head: Frodo gets the ring when 23 but leaves for Bree when 50. Farmer maggot. The shire elves. Crickhollow. Fatty Bolger. The old forest. Old Man Willow. Top Bombadil. Barrow Downs. Bill Ferny. Stone trolls (in extended). Glorfindel. Radagast. Wolves before Moria. Road to Lorien (rope bridges). Boat carrying. That's just Fellowship...

Martin Büttner

what we don't have is a tag for rationals...

Calvin's Hobbies

We've moved on :(

Geobits

3:55 PM

Just like Frodo

TheBestOne

@Calvin'sHobbies When 33, not 23 ;)

Geobits

I thought Radagast was in the movie? Maybe in a different time/place.

Martin Büttner