The Whiteboard

12:02 AM

panabox:~/xxx/crev> time java HandSome pokerlarge.txt
Player 1 wins 7520 times.
1.936u 0.148s 0:00.73 283.5%    0+0k 0+64io 0pf+0w
panabox:~/xxx/crev> time java HandSome p054_poker.txt
Player 1 wins 376 times.
0.696u 0.064s 0:00.32 234.3%    0+0k 0+64io 0pf+0w

durron597

@rolfl I find your program significantly harder to follow than mine. This is evidenced by your abundant (needed) comments

rolfl

Yeah, it is harder... I agree.

On the other hand, it is hopefully faster.... though I have not yet benchmarked yours.

durron597

~/workspace/problem54/target/classes
$ time java problem54.Problem
376

real    0m0.100s
user    0m0.015s
sys     0m0.000s

that's on cygwin, how do i make it look like yours

@rolfl is my 0.015s properly compared to your 0.064s?

rolfl

I'm using tcsh.

lte me make it bash

durron597

@rolfl good, because i don't think i have tcsh installed on this cygwin :-P

rolfl

12:08 AM

rolf@panabox:~/xxx/crev$ time java HandSome p054_poker.txt
Player 1 wins 376 times.

real    0m0.345s
user    0m0.760s
sys     0m0.048s
rolf@panabox:~/xxx/crev$

hmmm.... min'es on a linux machine, which currently looks like:

durron597

hard to compare because we don't know system stats

i know i'm doing my file i/o on an SSD

rolfl

I am in the middle of my monthly backup cycl, I am hammering away on multiple raid devices, and doing large rsync's with numerous hard links.

durron597

probably best would be to just compile and run mine on your machine

i don't have java 8 installed even

user55340

@durron597 returning to a previous topic... chrisbeaumont.org/holdem_odds/#8H+QS

durron597

@MichaelT Fascinating. I'd love to talk about that further but I have one foot out the door

user55340

12:16 AM

1

Q: Thought experiment: What would happen if we didn't have close votes?

After reading "My Love-Hate Relationship with Stack Overflow: Arthur S., Arthur T., and the Soup Nazi" one of our developers wondered what would happen if we didn't have close votes. Disclaimer: I don't know of any actual plans to actually get rid of close votes. Like legalizing marijuana, a chan...

rolfl

@durron597 - Hmmm. My solution is many times slower than yours.

sol@solarium:~/xxx$ time java -cp . problem54.Problem
376

real    0m0.101s
user    0m0.128s
sys     0m0.017s
sol@solarium:~/xxx$ time java -cp . HandSome
Player 1 wins 376 times.

real    0m0.755s
user    0m0.946s
sys     0m0.031s

durron597

@rolfl Time to bust out the profiler ;)

okay gotta run. have a good night everyone

rolfl

Hmmm... you cheat... ;-)

If I convert my code to read from a class loader's getResource... it runs in:

376

real    0m0.082s
user    0m0.101s
sys     0m0.026s
sol@solarium:~/xxx$ time java -cp . HandSome
Player 1 wins 376 times.

real    0m0.172s
user    0m0.371s
sys     0m0.016s

only twice as slow as yours now, which is 4 times better than before.

psr

@MichaelT I think there is a new program that is just a hair off of perfect play in 2 player hold 'em (academic paper with a proof). Seeing what it does vs. various hands ought to be interesting.

Robert Harvey

12:32 AM

@rolfl That's pretty nice work.

I'm trying not to look at the code too closely, because I'm going to steal your 64 bit idea, but write my own hand detectors.

user55340

Need to boot up the desktop and fix my perl implementation of PE54... its... interesting IIRC.

user55340

(I converted the hand into a string and sorted them)

Transcript for