Page 5 of 13

Re: 1193: Externalities

Posted: Mon Apr 01, 2013 8:47 pm UTC
by ThatOneRoadie
chrisjbillington wrote:hwæt version do eow have? ic have pyskein 0.6.2


ï obtained thè sourcè codë for vèrsion 0.7.1 of pySkein and compilëd ït mysèlf.

Vèrsions 0.7 ör bettèr havè Å nèw kèy cönstant, whïch cöuld èxplain thè differencès ïn hashès.

Re: 1193: Externalities

Posted: Mon Apr 01, 2013 8:57 pm UTC
by pitareio
correcthorsebatterystaple1097262 is 431 bits off.

Re: 1193: Externalities

Posted: Mon Apr 01, 2013 8:59 pm UTC
by kierajeng
Current page = Dell Comics perhaps? (Yes, I registered just for this.)

Re: 1193: Externalities

Posted: Mon Apr 01, 2013 9:08 pm UTC
by asmodai
Damn they're fast at protecting. I was locked out even though I tried 10 seconds after the last change. Doesn't even look like anybody got an edit in edgewise.

Re: 1193: Externalities

Posted: Mon Apr 01, 2013 9:10 pm UTC
by asmodai
Dammit, foiled again!

"The rise in popularity of the party is owed in part to the rise in popularity if pirates in general, subsequent to the release of the film [[Pirates of the Caribbean]] by [[The Walt Disney Company]]"

Re: 1193: Externalities

Posted: Mon Apr 01, 2013 9:11 pm UTC
by jjjdavidson
Regarding disparate values from various Skein-1024 implementations:

í suspect the problem to be encoding: ASCII/Unicode/Big-endian/et cetera. Åny Skein implementation should have been tested against several test values; unfortunately âll test values í've found hâve been untypeable hex strings starting with 0xFF.

Skein (well, âny hash) looks at byte values, not characters, so it mâkes â difference whether the input's ASCII, Unicode, or whatever. The PΘST code specifies enctype="application/octet-stream"; unfortunately í don`t kñow enough to say whât effect thât hâs.

Re: 1193: Externalities

Posted: Mon Apr 01, 2013 9:14 pm UTC
by Aaeriele
afschuld wrote:How do ic evaluate how close two hashes be? ic be capable of running the same hash algorithm as the website but ic dont ken how it determines it be ___ bits off thing.


A hash is actually a series of bits. Two hashes are two series of bits. Comparing the the Nth bit from one series to the Nth bit of the other, gives you a count of how many are different.

Re: 1193: Externalities

Posted: Mon Apr 01, 2013 9:20 pm UTC
by vasporig
why do i get "School not found" ? :(

Re: 1193: Externalities

Posted: Mon Apr 01, 2013 9:21 pm UTC
by pitareio
To put it differently : XOR hash1 and hash2, count how many "1" bits you get.

Re: 1193: Externalities

Posted: Mon Apr 01, 2013 9:25 pm UTC
by pitareio
pitareio wrote:correcthorsebatterystaple1097262 be 431 bits off.


I never expected it to work so well : there are now over 1000 unis with a score of 431 in best.csv :shock:

(sorry folks, this isn't getting you any nearer to entering the comic)

Re: 1193: Externalities

Posted: Mon Apr 01, 2013 9:41 pm UTC
by teelo
That alamater thing wont accept my university vuw.ac.nz

WHAT THE HELL RANDALL

Re: 1193: Externalities

Posted: Mon Apr 01, 2013 9:47 pm UTC
by pitareio
teelo wrote:That alamater thing wont accept my university vuw.ac.nz

hwæt THE HELL RANDALL


Judging from best.csv, only .edu, .uk and .au (this is where you should get mad) got in.

Re: 1193: Externalities

Posted: Mon Apr 01, 2013 9:48 pm UTC
by Mindor
pitareio wrote:
pitareio wrote:correcthorsebatterystaple1097262 be 431 bits off.


ic never expected it to work so well : there be now over 1000 unis with an score of 431 in goodest.csv :shock:

(sorry folks, this be not getting eow ænig nearer to entering the comic)

correcthorsebatterystaplecorrecthorsebatterystable is 461

Re: 1193: Externalities

Posted: Mon Apr 01, 2013 9:50 pm UTC
by ThatOneRoadie
öffline Vèrsion ïn Python3 för thosè whö wånt ït:

pastebin.c om/99pxf5kx

ï could'vè madë bettèr usè of whitèspace, but oh wèll

Re: 1193: Externalities

Posted: Mon Apr 01, 2013 9:54 pm UTC
by Flumble
Mindor wrote:
pitareio wrote:
pitareio wrote:correcthorsebatterystaple1097262 be 431 bits off.


ic never expected it to work so well : there be now over 1000 unis with an score of 431 in goodest.csv :shock:

(sorry folks, this be not getting eow ænig nearer to entering the comic)

correcthorsebatterystaplecorrecthorsebatterystable be 461

Makes me wonder how well the hash chain of correcthorsebatterystaple performs.

Re: 1193: Externalities

Posted: Mon Apr 01, 2013 10:01 pm UTC
by Eternal Density
When did Randall become Jimmy Whales?

Re: 1193: Externalities

Posted: Mon Apr 01, 2013 10:06 pm UTC
by pitareio
I hope to see a larger dog when/if we reach 25000 USD.

Re: 1193: Externalities

Posted: Mon Apr 01, 2013 10:07 pm UTC
by KrytenKoro
endolith wrote:
pitareio wrote:ic'd like to ken hwæt maketh so eow angry, and hwæt maketh eow reckon the donations (currently an healthy amount of cash) woll not largely compensate the slight disruptions caused by today's comic to Wikipedia.


Wikipedians have no sense of humor. They'woll just protect articles and permaban æferychon and whine about this, no matter how big the donations.

....right, 'cause deciding to do something constructive rather than whoring themselves out for the whims of some idiot vandals is so unhumored.

I mean, it's not like they already have such a bad reputation among academia because of all the -TOTALLY HILARIOUS JOKES- inserted by oxygen-wastes like the ones on this thread, is it?

Re: 1193: Externalities

Posted: Mon Apr 01, 2013 10:10 pm UTC
by edo
Flumble wrote:
Mindor wrote:
pitareio wrote:
pitareio wrote:correcthorsebatterystaple1097262 be 431 bits off.


ic never expected it to work so well : there be now over 1000 unis with an score of 431 in goodest.csv :shock:

(sorry folks, this be not getting eow ænig nearer to entering the comic)

correcthorsebatterystaplecorrecthorsebatterystable be 461

maketh me wonder how well the hash chain of correcthorsebatterystaple performs.


(not too well.)

Re: 1193: Externalities

Posted: Mon Apr 01, 2013 10:31 pm UTC
by airdrik
Hey, someone (from cmu) finally bested stanford's score. We'll see how long that lasts.
Nice to get a little change in scenery/wordage

Re: 1193: Externalities

Posted: Mon Apr 01, 2013 10:37 pm UTC
by rubseb
Donation campaign or no, I don't like this invitation for Wikipedia vandalism. Fun as it may be to solve these clues and see your company appear in the comic, it's one big pain in the ass for those of us who try to keep Wikipedia clean. It's like drawing penises in your school books except the whole world still wants to use them. And sure, it will all probably be cleaned up in time, but I hate to think that there are now some xkcd fans who think more lightly about vandalizing Wikipedia for fun. It just seems unnecessary and not very clever (he could have thought of a less destructive game).

Re: 1193: Externalities

Posted: Mon Apr 01, 2013 10:45 pm UTC
by wildgunman
rubseb wrote:Donation campaign or no, I don't like this invitation for Wikipedia vandalism.

Yeah, well the man in charge says you can lump it. https://twitter.com/jimmy_wales/status/318734598597115904

Re: 1193: Externalities

Posted: Mon Apr 01, 2013 11:02 pm UTC
by firinne
Ic eom a bit disappointed that "ic eom" isn't one of the filters :P What a pleasant surprise to see on the fora, nevertheless; something for everyone!

Re: 1193: Externalities

Posted: Mon Apr 01, 2013 11:02 pm UTC
by edo
you guys know what this is:
bdf4a1cd0e7a726c2108b41f4cbea8802001b226fcb5702e946c7270b3e0df0822c1e31e538cf9db696791ad17a8cffcdaf320e0d40db6dee1a0af5794b04c105c1578e9459f3ccb1c52d66ccacd01049e40d68173ab1fb6afa0919112bdaae7ab702d72c87894b548821da3fdc695e3c5e87d9bebe88efc3b08b7348fa7a823

Spoiler:
Hash browns

Re: 1193: Externalities

Posted: Mon Apr 01, 2013 11:07 pm UTC
by SWuh
Manchester were on top for a short while!! That made me superhappy :D --- I must be very easily pleased!!!

Re: 1193: Externalities

Posted: Mon Apr 01, 2013 11:21 pm UTC
by jaronflick
wildgunman wrote:
rubseb wrote:Donation campaign or no, ic don't like this invitation for Wikipedia vandalism.

Yeah, well the man in charge sayeth eow can lump it. https://twitter.com/jimmy_wales/status/318734598597115904


I'm impressed at how fast the editors were on top of the changes. Were they warned?

I know people in schools who hate Wikipedia. They would point to this and say "Look how easily wrong information was entered." I would point to it and say "Look how fast it was corrected!"

GO Wikipedia!

Re: 1193: Externalities

Posted: Mon Apr 01, 2013 11:27 pm UTC
by blowfishhootie
wildgunman wrote:
rubseb wrote:Donation campaign or no, I don't like this invitation for Wikipedia vandalism.

Yeah, well the man in charge says you can lump it. https://twitter.com/jimmy_wales/status/318734598597115904


I must be missing it. Where does that tweet say the vandalism is OK? Nothing about this comic makes vandalism necessary, though it certainly invites immature losers to do just that.

What else is Jimmy Wales going to say? Even if he were aware of the vandalism it has provoked, which I'm not sure of, he's in no position to demand Monroe do anything different. But his enjoyment of the comic in no way is an endorsement of the losers vandalizing Wikipedia, either through XKCD or anything else.

Re: 1193: Externalities

Posted: Mon Apr 01, 2013 11:34 pm UTC
by edo
BESURETODRINKYOUROVALTINE
Spoiler:
only 522, already checked it. variations only get me to about 470

Re: 1193: Externalities

Posted: Mon Apr 01, 2013 11:45 pm UTC
by gnutrino
tvynr wrote:
ThatOneRoadie wrote:It seems the hash engine be running off of python3 and the pySkein module.


Well, that's exciting. This thread has reports for four implementations of Skein-1024-1024 tested against the string "z". The J-ava, Haskell, and C implementations all agree on one hash and PySkein has another hash. Is PySkein broken?


I poked around and it seems PySkein uses v1.3 while I (and I guess the other people getting the wrong answers) was using 1.1 after upgrading it now works for me, for people that want a C implementation see http://www.skein-hash.info/downloads towards the bottom of the page where it says V1.3, don't use the 1.1 links at the top

Re: 1193: Externalities

Posted: Mon Apr 01, 2013 11:59 pm UTC
by AUS
So several things change based on several things. Anything new?

Re: 1193: Externalities

Posted: Tue Apr 02, 2013 12:05 am UTC
by pitareio
blowfishhootie wrote:
wildgunman wrote:
rubseb wrote:Donation campaign or no, I don't like this invitation for Wikipedia vandalism.

Yeah, well the man in charge says you can lump it. https://twitter.com/jimmy_wales/status/318734598597115904


I must be missing it. Where does that tweet say the vandalism is OK? Nothing about this comic makes vandalism necessary, though it certainly invites immature losers to do just that.


Come one. Get real. This comic IS a direct invitation to edit Wikipedia pages (and to abuse uni/lab computers with a silly hash break search, and to donate to wikimedia foundation). And Jimmy Wales seems pretty happy with it. Oh, I forgot, who is he to decide what's good or bad for wikipedia?

More to the point, I'm a little disappointed, the dog hasn't changed since we passed $12k or so.

Re: 1193: Externalities

Posted: Tue Apr 02, 2013 12:07 am UTC
by marathdamane91
Is the 3rd panel not updating correctly? It hasn't changed from the 4 hour drive from New York statement in like 2 hours that I've seen. Even when it was Baidu Inc....

Re: 1193: Externalities

Posted: Tue Apr 02, 2013 12:12 am UTC
by bugstomper
pitareio wrote:So, the key is sending as many requests as you can with different strings, and I suspect the top unis at http://almamater.xkcd.com/best.csv had scripts running for hours.


Not sending as many requests as you can -- A response from the server telling you how many bits wrong you got takes about a second. You have to compute the hash yourself as fast as possible to try some enormous number of hashes per second and only send in something to the server when you find one better than the current best.

My very rough and possibly wrong calculation of the odds of getting the current best, which is 394 from cmu, is about 1 out of 1.2e13 which means they have hashed over 10 trillion numbers so far to get there. Unless you are prepared to generate that many hashes you don't stand a significant chance of beating the top contenders.

Re: 1193: Externalities

Posted: Tue Apr 02, 2013 12:20 am UTC
by vollkorn
Alright, I want my schools domain to be added to the hash competition, but I cannot post it here because the spam filter keeps me from posting an url-like string.

Mayeb like this: uni-hamburg DOT de

Re: 1193: Externalities

Posted: Tue Apr 02, 2013 12:27 am UTC
by KarMann
bugstomper wrote:
pitareio wrote:So, the key is sending as many requests as you can with different strings, and I suspect the top unis at http://almamater.xkcd.com/best.csv had scripts running for hours.


Not sending as many requests as you can -- A response from the server telling you how many bits wrong you got takes about a second. You have to compute the hash yourself as fast as possible to try some enormous number of hashes per second and only send in something to the server when you find one better than the current best.

My very rough and possibly wrong calculation of the odds of getting the current best, which is 394 from cmu, is about 1 out of 1.2e13 which means they have hashed over 10 trillion numbers so far to get there. Unless you are prepared to generate that many hashes you don't stand a significant chance of beating the top contenders.

No, it doesn't mean that they have hashed that many. That's just a rough average of how many they ought to have hashed to get there. There's a small but finite chance they could have just hashed one, and gotten lucky; they might well have hashed 20 trillion, and been rather unlucky.

Re: 1193: Externalities

Posted: Tue Apr 02, 2013 12:28 am UTC
by chris857
The Dog, he CHANGED! His nose is pushing into the monitor now.

Re: 1193: Externalities

Posted: Tue Apr 02, 2013 12:28 am UTC
by SecondTalon
From the third post as I'm disapproving it

vollkorn wrote:Darn, do my posts need to be approved first or am I too stupid to use forums? I think I know why I don't really like forums...


The message you get when you post your first few posts literally tells you your messages are going in a queue.

So yes, your first two need approval. And if you'd read the rules, you'd also know why your URL links keep getting barked at as spam and rejected.

As for how to tell Randall ... he doesn't read the forums. Hasn't in years. And he doesn't take submissions.

So.. uh.. contact@xkcd.com? Or press@xkcd.com?

Re: 1193: Externalities

Posted: Tue Apr 02, 2013 12:31 am UTC
by pitareio
bugstomper wrote:
pitareio wrote:So, the key is sending as many requests as you can with different strings, and I suspect the top unis at http://almamater.xkcd.com/best.csv had scripts running for hours.


Not sending as many requests as you can -- A response from the server telling you how many bits wrong you got takes about a second. You have to compute the hash yourself as fast as possible to try some enormous number of hashes per second and only send in something to the server when you find one better than the current best.

My very rough and possibly wrong calculation of the odds of getting the current best, which is 394 from cmu, is about 1 out of 1.2e13 which means they have hashed over 10 trillion numbers so far to get there. Unless you are prepared to generate that many hashes you don't stand a significant chance of beating the top contenders.


Yep, I had already been corrected (and understood I could hash it myself). Took about 20 million tries to get down to 428, I'll leave it running overnight and see if I get lucky. Stanford and the others probably run highly optimized code on several powerful machines, I know the odds are not in my favour.

Re: 1193: Externalities

Posted: Tue Apr 02, 2013 12:47 am UTC
by gparker
KarMann wrote:
bugstomper wrote:
pitareio wrote:There's a small but finite chance they could have just hashed one, and gotten lucky; they might well have hashed 20 trillion, and been rather unlucky.


In fact I did get outrageously lucky when I hit 402 after about 30 minutes of searching. (Nobody matched that score for about eight hours.) Mathematically I had a good chance of reaching 402 only after running all day, and in fact I did run all day with only a 403 and a few 404s afterwards.

Re: 1193: Externalities

Posted: Tue Apr 02, 2013 12:48 am UTC
by gnutrino
I've gotta wonder how much cluster time in top universities Randall just tied up...