01:25:17: <bvowk_> omfg.
01:26:01: <bvowk_> I just facepalmed so hard that I'm pretty sure my brain has a hand print on the frontal lobe now.
02:12:39: <Doxin> bvowk_: do tell! what didya do?
02:58:51: <bvowk_> I accidentally flipped two values in a protobuf, so everything for months has been slightly broken.
03:14:02: <Doxin> bvowk_: oops
03:54:30: <bvowk_> you have no idea how long I looked at that and didn't notice.
04:14:21: <Doxin> I know the feeling
04:14:39: <Doxin> bvowk_: thing that tends to get me: misplaced bracket in javascript.
05:16:40: -!- Doxin has quit (Remote host closed the connection).
06:47:15: -!- impomatic has joined #corewars.
08:29:10: -!- myndzi has quit (Ping timeout: 260 seconds).
08:30:25: -!- myndzi has joined #corewars.
09:28:43: * impomatic casts a spell of summoning :-)
09:28:52: -!- fluffy has joined #corewars.
09:28:58: <impomatic> Hi Fluffy! :-)
09:29:45: <fluffy> So it is still alive
09:29:54: <fluffy> or rather not dead
09:32:24: <impomatic> Yes, still a bit of activity in IRC. The newsgroup has been spammed almost to death by Skybuck :-(
09:32:51: <impomatic> There's also a bit of activity on twitter and G+.
09:33:01: <fluffy> that name rings a bell
09:33:25: <impomatic> 94nop has been busy and we've had four tournaments over the last couple of years.
09:36:55: <fluffy> No time for tournaments :-(
09:38:35: <impomatic> I think most players are the same. Life takes over :-D
09:40:23: <impomatic> It's difficult to find time here. Fortunately I work for myself so I make time for a bit of Core War at work.
09:52:37: <impomatic> fluffy: Did you see The Art of CoreWar passed age 1000 before it left the hill? :-) http://corewar.co.uk/artofcorewar.htm
09:54:02: <fluffy> yes, but it is dead now
09:54:29: <fluffy> ... and I might have a reason to create a new warrior
09:54:39: <fluffy> ... in a couple of year's time
09:54:56: <fluffy> or decades, depending on the available time
09:55:13: <impomatic> lol I know how you feel.
09:56:35: <impomatic> It can take a while to set up the tools and an up-to-date benchmark before you can even get started.
09:57:19: <fluffy> It takes me longer to write a new MARS
09:57:35: <fluffy> which I would have to do, because otherwise it feels like cheating
09:57:37: <fluffy> ;-)
09:59:55: <impomatic> The only thing that feels like cheating to me is using someone else's quickscan. That's why I had to write my own :-)
10:01:13: <fluffy> Agreed.
10:01:56: <fluffy> Since I'm not capable of writing decent warrior I have to write programs which write decent warriors
10:02:10: <fluffy> So I have to write a MARS
10:08:50: <impomatic> The Core War twitter feed has a fairly complete rundown of Core War news from the last couple of years. (There's even a few photos from tournaments) https://twitter.com/xcorewar
10:17:21: -!- verpeteren has joined #corewars.
10:18:19: * verpeteren hay
10:18:33: <verpeteren> nice to see some familiar names here
10:18:33: <impomatic> Hi verpeteren :-)
10:18:34: <fluffy> hola
10:18:55: * verpeteren was janssenmetdubbels longlongago
10:19:18: * impomatic remembers :-)
10:22:12: <impomatic> Here's a sneak preview of Silkworm - a binary launched paper/imp with a wimp and silk-imps http://corewar.co.uk/silkworm.htm
10:23:05: <impomatic> Nothing on 94nop is currently beating it :-D
10:24:36: -!- gb_master has joined #corewars.
10:25:42: <fluffy> Hrm! It's been years since I've last looked into cw and now I feel the urge again to play
10:26:02: <fluffy> damn you, impomatic and you mail!
10:26:09: <impomatic> lol :-)
10:26:16: <fluffy> I don't have time.
10:26:54: <verpeteren> for me it's been awile to. but i don't have the urge to play, I am just procrastinging some work that i need to do.
10:27:43: <fluffy> have cw ever been something else?
10:28:11: <impomatic> Something else?
10:28:30: <impomatic> Ah, okay?
10:28:30: <verpeteren> procrastination
10:28:35: <fluffy> than procastination
10:28:50: <verpeteren> abstract thinking to the max.
10:28:56: <impomatic> I misunderstood and was about to mention Darwin, etc :-D
10:29:47: <fluffy> Is there still an active tiny hill?
10:30:15: <impomatic> http://sal.discontinuity.info/hill.php?key=tiny
10:30:45: <fluffy> Doesn't seem very active
10:31:19: <fluffy> It still has one of my warriors
10:32:07: -!- gb_master has quit (Quit: Leaving).
10:32:36: <impomatic> Only 13 successful challenges over the last year :-(
10:34:07: <impomatic> I'd be happy to write a few new tiny warriors. I'm just waiting for someone to know a few of my warriors off first.
10:34:31: <impomatic> bvowk_ is currently working on something for tiny :-)
10:35:40: <fluffy> last time I've checked it was too easy to generate some warrior for tiny
10:35:57: <fluffy> ... not tiny, nano
10:36:07: <fluffy> but too difficult to do that for tiny
10:36:27: <fluffy> now, tons of years later tiny might be within reach
10:36:50: <fluffy> now that I've thrown away my last 486
10:46:24: <impomatic> I ought to try nano again, just to see how well hand-coded warriors score these days.
12:00:13: -!- Payhn has quit (Quit: ZNC - http://znc.in).
12:02:22: -!- Payhn has joined #corewars.
13:10:05: -!- yoR has joined #corewars.
13:10:20: <impomatic> Hi yoR :-)
13:11:30: <yoR> A reunion!
13:12:26: <impomatic> Anyone reading the logs will think they've opened a log from 2005!
13:13:43: <yoR> Oh well, nothing has changed really...
13:14:14: <yoR> Monster_Human_Grunt and Hullabaloo are still on the hill
13:14:44: <impomatic> Did you see Sputnik (simple paper with Satellite clear) http://corewar.co.uk/sputnik.htm and Silkworm (binary paper/imp with silk-imps) http://corewar.co.uk/silkworm.htm
13:15:22: <yoR> hill94xm hasn't seen action in over a year and a half...
13:15:37: <yoR> Yes, I've seen them :)
13:17:12: <impomatic> Okay, sent something to 94xm :-P
13:18:31: <yoR> What surprises me is that Sputnik makes so little copies, it is more a clear than a paper
13:19:09: <impomatic> About 20 copies I think. By that time it has about 7200 processes.
13:19:57: <impomatic> Hand picked constants :-P
13:26:37: <yoR> I'm using OS/X and CoreWin (using wine) is always crashing/hanging :(
13:27:15: <impomatic> :-(
13:27:38: <impomatic> How about the OS/X pMARS binary?
13:29:42: <yoR> That works okay, but I'm so used to having everything CoreWin has now, I'll try WineBottling (yes, that is how it is called) again (repackaging)
13:43:07: <yoR> 84% ties hehe... hmmm
13:46:42: <impomatic> Oh no... are we going for the minimum losses / maximum ties challenge again?
13:47:24: <yoR> Nope, it is just part of a new warrior
13:51:19: <impomatic> Satellite imps?
13:52:25: <yoR> Yup, 87% now :)
13:52:38: <yoR> What was our record? Do you remember?
13:53:05: <yoR> I know what happened last time, For John was the result :P
13:53:39: <impomatic> No, I'd have to check the logs but first I need to update them for the latest 375 challenges.
13:55:13: <impomatic> Hmmm... 89% ties but didn't make it onto the hill :-(
13:55:47: <yoR> Well score doesn't matter, this is just an 'empty spot'
13:56:50: <yoR> Add a wimp to the one you just send!
13:59:29: <verpeteren> hai yoR.
13:59:36: * verpeteren loves your blogposts
13:59:47: <yoR> Thanks!
14:04:45: <impomatic> Wimp added to e1test :-)
14:11:03: <yoR> Good, 90% :)
14:29:06: <yoR> I should really get/make a new benchmark and some automated testing software... firing blindly against the hill is exiting but not really efficient :P
14:30:11: <impomatic> I'm only using Randy, MTS and Exmars
14:36:07: <yoR> Oh wait, I wrote my own randy/exmars thing in Java, it is running exmars from Java code (native) so I can use smart constant-picking and still have the speed
14:36:09: <impomatic> yoR: you should try adding a qscan. I predict you'll enter the hill in 9th place, score 126 - 18% wins, 9% losses, 73% ties. :-)
14:36:47: <yoR> Nah, the constants are still random, that is not 'how I roll' :-)
14:45:38: <yoR> And I'm still playing around with the structure of the warrior, basically the whole idea has changed
15:42:00: -!- myndzi has quit (Ping timeout: 256 seconds).
15:45:31: -!- myndzi has joined #corewars.
16:13:19: <yoR> Not really impomatic... 9th? tsss... 20th!
16:23:32: <fluffy> Oh, now it is really 2005 again
16:29:00: -!- yoR has quit (Ping timeout: 252 seconds).
17:13:44: -!- yoR has joined #corewars.
17:22:34: <impomatic> yoR: use the qscan constants I sent you instead of the default ones. 369 / 3730 do really badly on the hill for some reason.
17:24:04: <impomatic> 2631 / 1272 will do the trick.
17:30:13: <bvowk_> whoa.
17:30:33: <bvowk_> We're packing a whole year worth of activity into this last little bit :)
17:30:51: <bvowk_> how are you doing fluffy?
17:37:27: <fluffy> Still alive
17:39:59: <bvowk_> working on anything interesting?
17:40:40: <fluffy> No, just playing with my xmas presents
17:41:22: <yoR> Musical instruments?
17:41:43: <yoR> Oh bvowk: Thanks again for the delicious pancakes :-D
17:43:03: <bvowk_> That place does a pretty great brunch eh? :)
17:43:16: <bvowk_> Next time you're out we'll get some BBQ perhaps?
17:43:20: <yoR> Absolutely, hungry now thinking about it
17:43:57: <yoR> Someone showed me a picture yesterday of the 'best burger in Germany'.... dying to go there now :-)
17:44:59: <yoR> https://pbs.twimg.com/media/CXa_AuqWsAQJNmk.jpg:large
17:45:58: <bvowk_> damn. Maybe I should be coming to visit your side of the pond :)
17:47:56: <yoR> You absolutely should, technically speaking Germany is fluffy's territory, but it is probably closer to my home (Düsseldorf)
17:54:24: <bvowk_> heh, I like the idea of there being territory for the CW folks.
17:54:51: <bvowk_> it does seem like for most areas there can be only one.
17:55:21: <bvowk_> Fizmo had to go to Asia, PK and I have to split the Americas in half..
17:55:25: <yoR> One per country...
17:56:20: <yoR> verpeteren: You'd better not be Dutch :D
17:56:27: <bvowk_> heh
17:57:14: <bvowk_> yoR: this is the place we should go when you come back http://www.magnoliasmokestack.com/
17:58:00: <bvowk_> Although they haven't posted the menu for today yet :(
17:58:11: <yoR> If you say so :-)
18:03:26: <bvowk_> What are you working on right now? Anything interesting?
18:03:57: <yoR> Satellite papers, an optimized version of Golden Eye, with a bit faster boot and smaller footprint
18:05:11: <bvowk_> I stole some ideas from the uGP folks yesterday
18:05:31: <bvowk_> and figured out why my evolver liked clears so much.
18:05:46: <yoR> And why is that?
18:06:15: <bvowk_> because I flipped ties and losses in one of the datastructures I passed to the evolving part.
18:07:01: <yoR> Ouch... my old evolver has a bug as well, but I don't know what it is, never found out and ditched it
18:07:56: <yoR> It went well for a couple of iterations, improving slightly every generation, and suddenly a very bad warrior would populate the entire group and ... ? Failure
18:08:08: <bvowk_> it was really stupid, it's all handled automagically everywhere else, except for this one bit where I was massaging some old code to work with the new data format. I shouldn't have done it.
18:08:31: <bvowk_> a very bad? like one that scored well but didn't have anywhere better to go?
18:08:39: <bvowk_> Or one that scored badly and just took over anyway?
18:08:52: <yoR> No, one that was suicidal or something, pretty crazy, clear bug
18:09:11: <yoR> But coulnd't find it :')
18:09:18: <yoR> Oh well, that was in 2005 or something
18:09:26: <bvowk_> heh
18:10:06: <yoR> Maybe I'll write another evolver tomorrow, being the 31th, nobody at work
18:10:21: <yoR> Nothing better to do, just found my old code, should be easy enough
18:10:23: <bvowk_> turns out you can actually generate some pretty interesting warriors even if you screw up losses and ties ;)
18:10:28: <yoR> (a simple evolver that is)
18:10:41: <yoR> What features does yours have now?
18:11:20: <bvowk_> it's pretty simple, it generates pretty much everything from random mutations.
18:12:10: <yoR> Shit, gotta run, please elaborate, I'll read it tomorrow, I've got a volleyball tournament starting in 20 minutes, 15 cycling from here, and I still have to pack my bag
18:12:20: <yoR> 15 min
18:12:26: <bvowk_> All my sauce is in the distribution parts that make it run all over the place, keeping diversity in the islands it runs, and automatically adjusting the mutation rates.
18:12:32: -!- impomatic has quit (Read error: Connection reset by peer).
18:12:37: <bvowk_> Cya :)
18:14:19: <verpeteren> yoR: i am dutch.
18:14:29: <bvowk_> uh oh.
18:17:16: <verpeteren> so what is it today that everybody is 'back'. or was i the only one that was gone?
18:17:45: <bvowk_> I think everyone is likely on holidays and was checking in?
18:17:55: <verpeteren> cool.
18:17:58: <bvowk_> But it's definitely been a serious reunion :)
18:18:06: <bvowk_> I don't think I've seen fluffy in 10 years.
18:18:30: <verpeteren> We should plan something, (with beer or wine)
18:18:53: <verpeteren> any body going to fosdem "NEXT YEAR"?
18:21:42: -!- john_metcalf has joined #corewars.
18:29:18: <bvowk_> Sadly, not me. Although I'll definitely be in Ireland a few times this year.
18:31:50: <bvowk_> I don't recognize your name tho verpeteren, do I know you from corewars past?
18:57:12: <verpeteren> +- 5-7 years ago, janssenmetdubbels
18:57:36: <bvowk_> Yeah, that sounds familiar :)
18:57:37: <verpeteren> never realy played cw, but tried to make a redcode parser
18:57:39: <bvowk_> Nice to have you back.
18:57:44: <verpeteren> :-)
18:58:31: <bvowk_> Redcode isn't so easy to parse.
19:01:13: <verpeteren> the icws94 describes two parts, the forrof/equ is tricky. i got stuck in "unary operators inside macros" that i could not expand correctly.
19:01:38: <verpeteren> i'll giveit another try once i'm a milonaire.
19:02:14: <fluffy> I wouldn't do that ... except may be because it sounds like fun.
19:02:45: <fluffy> Almost no published warrior uses those parts of the "spec"
19:03:59: <verpeteren> hmm. interesting.
19:04:39: <bvowk_> yeah, I pull out most of the for/rof and some of the tricky equ stuff in the warriors I use for benchmarks and stuff.
19:04:39: <fluffy> I've ended "fixing" the one or two warriors that do that
19:04:50: <bvowk_> otherwise they work in some mars and not others :(
19:05:31: <verpeteren> so you run them throug pmars to get the expanded codeL
19:06:09: <bvowk_> or I hand expand them.
19:07:15: <verpeteren> ah, i see. Does that mean that pmars isn't the preferred engine for the hills any more?
19:08:46: <bvowk_> sal uses the exmars that has the pmars front end.
19:08:47: <verpeteren> bvowk_: if i remember correctly you were deep in the warrior-evolving-marses? what is the preferred mars nowadays?
19:08:51: <bvowk_> I use pycorewars
19:09:02: <bvowk_> or exmars for some of them.
19:09:08: <verpeteren> ah.
19:10:31: * verpeteren knows that he has *soon* 3 weeks off. these weeks are planned for a *sideproject*.
19:11:25: * verpeteren feels his dissicpline leaking away
19:11:28: <bvowk_> heh
19:11:59: -!- impomatic has joined #corewars.
19:12:50: <verpeteren> damn, I installed leechblock. that banned me from HN and reddit. then I installed irc to fill this gap. Now I want to dive into cw again.
19:13:57: <verpeteren> i think i can deinstall leechblock and irc. reading HN/reddit and slashdot is a harmless waste of time. CW is a endless pit of hours and fun.
19:18:09: <fluffy> no, weeks, months and years!
19:18:37: <bvowk_> I always keep corewar stuff idling in the back of my head.
19:18:53: <bvowk_> I use it as a test bed for new things I play with.
19:19:00: <bvowk_> new hardware? Put corewars on it.
19:19:12: <bvowk_> new machine learning idea? Try it on corewars
19:20:09: -!- impomatic has quit (Ping timeout: 245 seconds).
19:27:35: <fluffy> I've managed to stay away from cw for quite a long time. And now.
19:28:23: <fluffy> It took only one mail from John and I'm thinking about running my old MARS
19:28:28: <fluffy> ... to generate new warriors
19:29:38: <bvowk_> John wanted you to add debugging and output to pycorewars :)
19:30:30: <bvowk_> if you made an open cl version so I could have it run on the phi's without running python natively, that'd b make me happy ;)
19:51:19: <fluffy> Are you serious about the pycorewars thingy?
19:53:33: <fluffy> What very are you currently using?
19:54:06: -!- impomatic has joined #corewars.
19:54:27: * fluffy pokes impomatic
19:54:30: <fluffy> Which version?
19:58:42: <impomatic> Wifi is playing up due to the weather.
19:58:56: -!- Fizmomaster has joined #corewars.
19:59:10: <impomatic> I only wondered if there's already anything in PyCorewar I could take a screenshot of
19:59:11: <Fizmomaster> hi hi
19:59:13: <impomatic> Hi Fizmo :-)
19:59:26: <Fizmomaster> hi impomatic
20:01:16: <fluffy> impomatic: You could take a screenshot of how quickly it returns to the command line ;-)
20:02:04: * Fizmomaster thinks fluffy things returns today ;-)
20:06:59: <impomatic> I've made a start on the mirror :-) http://corewar.co.uk/gutzeit/quickscanners/
20:09:48: <fluffy> That was a long time ago
20:18:49: * impomatic wonders when Fluffy will write this section "Part VI will probably deal with quickscanners for nonstandard-cores, especially for tiny and nano cores."
20:19:30: -!- Doxin has joined #corewars.
20:19:48: <impomatic> Hi Doxin
20:19:54: <Doxin> hi impomatic!
20:20:00: <Fizmomaster> Hi Dioxin
20:20:02: <Doxin> whatcha up to?
20:20:15: <Fizmomaster> pretty busy the channel today :-P
20:21:03: <impomatic> Tweaking some satellite code, mirroring a website and a few other bits and pieces :-)
20:21:34: <impomatic> Fizmomaster: I've just received a scan of the Japanese Core War Newsletter :-D
20:21:55: <Fizmomaster> Yay :-D
20:22:14: * fluffy kicks impomatic - Which version of pycorewar are you using?
20:22:23: <bvowk_> whoa! hey fizmo!
20:22:56: <Fizmomaster> hey bvowk :-D
20:22:57: <bvowk_> PyCorewar-0.2.0.28
20:23:28: <fluffy> 28 ... most the most recent version that I can find on my disk is 24
20:24:40: <bvowk_> I think you made me this one for a bug fix.
20:25:14: <bvowk_> http://www.math.ualberta.ca/~bvowk/files/PyCorewar-0.2.0.28.tgz
20:25:39: * Fizmomaster accepts bvowks suggestion to take over CW Asia :-D
20:27:02: <impomatic> fluffy: I not using PyCorewar, I was just trying to get it up and running to take a screenshot :-)
20:28:17: <fluffy> There is nothing for a screenshot
20:28:52: <bvowk_> sure there is..
20:29:19: <bvowk_> you can show it parsing files into core and running.
20:29:20: <fluffy> k, maybe you use a different definition of 'screenshot'
20:29:28: <impomatic> How do I check which version of PyCorewar I have?
20:29:40: <fluffy> bvowk_: One does not simply take a screenshot of a console
20:29:47: <bvowk_> heh
20:29:52: <Doxin> impomatic: try print PyCorewar.__version__ or w/e the module is called
20:30:11: <fluffy> I feel pretty ashamed for the code
20:30:22: <fluffy> but that's the way it is
20:30:31: <bvowk_> [bvowk@bvowk-macbookpro wade]$ python
20:30:32: <bvowk_> Python 2.7.10 (default, Oct 23 2015, 18:05:06)
20:30:32: <bvowk_> [GCC 4.2.1 Compatible Apple LLVM 7.0.0 (clang-700.0.59.5)] on darwin
20:30:32: <bvowk_> Type "help", "copyright", "credits" or "license" for more information.
20:30:32: <bvowk_> >>> import Corewar
20:30:34: <bvowk_> >>> Corewar.__version__
20:30:36: <bvowk_> '0.2.99'
20:30:38: <fluffy> I'm also ashamed for my last month's code
20:31:01: <Doxin> fluffy: if you're not ashamed of the code you've written you're doing it wrong
20:32:02: <bvowk_> Doxin: likley you're just not looking at it again.
20:32:08: <Doxin> :P
20:32:15: <bvowk_> I find I like my code until I go back and look at it again ;)
20:32:16: <fluffy> ... and I don't seem to be great on the basics
20:32:19: <fluffy> ... like counting
20:32:37: <fluffy> the packge is 0.2.0.28 and __version__ is 0.2.99
20:32:49: <bvowk_> So it seem.s
20:32:53: <fluffy> for big 0.2.0.28 that is probably ok
20:33:23: <bvowk_> I've got a few hojillion hours of testing of 0.2.0.28/0.2.99
20:33:30: <bvowk_> I'm pretty sure it rights correctly :)
20:33:39: <fluffy> wow .. it actually compiles
20:33:42: <bvowk_> er.. runs correctly.
20:33:55: <bvowk_> it's even fast with modern compilers too.
20:34:14: <bvowk_> I didn't expect it to cross compile well with the intel compilers.
20:34:28: <fluffy> for phi?
20:34:32: <bvowk_> yeah
20:34:48: <fluffy> ok, damn you John, now I'm hooked
20:34:49: <bvowk_> I cross compiled all of python and numpy to get that working.
20:35:10: * fluffy compiles pmars
20:36:44: <fluffy> bvowk_: Since you have a "hojillion hours" of testing ... what changes (bugs) would you like to have?
20:41:05: <impomatic> I can't compile PyCorewar :-(
20:41:25: <fluffy> be sure to use python2
20:41:36: <impomatic> C:\Python27>python setup.py install
20:41:37: <impomatic> running install
20:41:39: <impomatic> running build
20:41:40: <impomatic> running build_py
20:41:42: <impomatic> running build_ext
20:41:43: <impomatic> building 'Corewar.Redcode' extension
20:41:45: <impomatic> error: Unable to find vcvarsall.bat
20:42:22: <fluffy> hrm ... windows
20:43:15: <fluffy> I can't remember that I've ever compiled it for windows
20:45:24: <impomatic> Fizmomaster: I've sent you a copy of the Japanese newsletter :-)
21:03:18: <impomatic> These are the two versions of PyCorewar I have http://corewar.co.uk/pycorewar.htm
21:04:01: <impomatic> Not sure if that's 0.2.0 or 0.2.0.?? because I can't get it to run on Windows :-(
21:05:37: <fluffy> impomatic: I'm currently looking for a vm with Windows
21:06:00: <fluffy> those operating systems for fringe groups are hard to find ;-)
21:06:37: <fluffy> if all else fail I can compile it at work
21:07:01: <fluffy> but that will the second week of next year
21:07:10: <fluffy> *fails
21:09:21: <bvowk_> fluffy: if you could make it use OpenCL to run multicore or on the Phi's that would be amazing :)
21:10:59: <fluffy> hm ... opencl ..
21:13:08: <fluffy> I don't think that I have time for opencl
21:13:23: <fluffy> does the phi have someting like pthreads?
21:13:34: <bvowk_> or openMP?
21:14:07: <bvowk_> yeah, the idea with the opencl/openmp stuff is that you use some language/compiler magic to spawn running threads on the phi cores and do work there.
21:15:12: <fluffy> how do you actually use it?
21:15:19: <fluffy> do you test all possible positions?
21:15:28: <bvowk_> I do both
21:15:33: <fluffy> I see
21:16:25: <fluffy> it is probably the easiest thing to run several different positions in parallel
21:16:27: <bvowk_> most of the evolving is with a fairly noisy number of rounds until towards the end.
21:16:47: <fluffy> but then you'd have to setup several (mars)cores
21:17:17: <bvowk_> there's 35MB or so per thread on the cards
21:17:24: <fluffy> more than enough
21:17:27: <bvowk_> plenty for cores.
21:17:42: <fluffy> Is there any problem with sticking to python2
21:17:51: <bvowk_> not for me..
21:18:12: <bvowk_> I wouldn't have a problem with using something newer either.
21:18:38: <fluffy> My main problem is that I don't know much about the phi
21:19:02: <bvowk_> Well, I can give you access to a box with all that.
21:19:15: <fluffy> The best I can do here is some 24 core
21:19:16: <bvowk_> But it should be retargetable between the various backends.
21:19:27: <bvowk_> so multicore should work with the phi too.
21:19:37: <bvowk_> and maybe some gpu
21:20:54: <bvowk_> if you wanted to go full crazy, you could try to SIMD your mars to run multiple cores in the wide instructions.
21:21:29: <bvowk_> the phi's and the newest cores have AVX-512 https://en.wikipedia.org/wiki/AVX-512
21:21:53: <impomatic> Sascha Zapf sends greetings to all in #corewars :-)
21:23:07: <fluffy> bvowk_: Intel says that pthreads will works
21:23:30: <fluffy> bvowk_: Does the 0.2.0.28 already compile on the phi?
21:25:42: <bvowk_> if I cross compile..
21:25:50: <bvowk_> I was hoping to avoid that in the future.
21:26:07: <bvowk_> just have the host offload the number crunching parts to the phi
22:31:33: -!- gb_master has joined #corewars.
22:47:22: -!- Fizmomaster has quit (Quit: ChatZilla 0.9.92 [Firefox 43.0.3/20151223140742]).
22:54:13: -!- fluffy has quit (Quit: nighty night).
23:02:24: -!- gb_master has quit (Quit: Leaving).