00:12:48  * mikealquit (Quit: Leaving.)
00:17:56  * c4miloquit (Remote host closed the connection)
00:21:28  * bentkusquit (Quit: Lost terminal)
00:23:42  <piscisaureus>alright, I'm going home
00:23:52  <piscisaureus>bnoordhuis: tomorrow no-ben wednesday?
00:29:21  <piscisaureus_>tootallnate: that's easy to fix
00:29:37  <TooTallNate>piscisaureus_: glad to hear :)
00:29:54  <piscisaureus_>apparently your arm gcc doesn't support uint128_t
00:29:57  <piscisaureus_>which is not surprising :-)
00:31:09  * joshthecoderquit (Quit: Linkinus - http://linkinus.com)
00:31:24  <TooTallNate>uint128_t
00:31:29  <TooTallNate>hah, never used that before
00:31:43  * c4milojoined
00:31:48  <piscisaureus_>nobody did :-)
00:32:36  <CIA-129>node: Bert Belder openssl-1.0.1c-wip * ra07316c / deps/openssl/openssl.gyp : fix arm ? - http://git.io/x-da5g
00:32:37  <piscisaureus_>c'mon CIA
00:32:39  <piscisaureus_>TooTallNate: ^-- can you try that?
00:32:59  * TooTallNatetrying
00:33:25  <TooTallNate>piscisaureus_: OS X builds btw; still need to run tests
00:33:52  <mjr_>Allow me to also be surprised that uint128_t exists.
00:34:08  <piscisaureus_>I wonder what the preferred method is for setting defines
00:34:22  <piscisaureus_>should I edit the opensslconf.h or the gyp file ?
00:35:12  <TooTallNate>piscisaureus_: good Q. I think we define some gyp defines in some places, but I edited the android conf.h directly to get arm working
00:35:26  <piscisaureus_>yeah, I always just edit the gyp file now
00:35:45  <TooTallNate>ok, sounds good to me
00:35:46  <piscisaureus_>I suppose editing the .h file is better for this flag
00:38:17  <piscisaureus_>also, 3 new perlasm files were added for aes
00:38:24  <piscisaureus_>I have no clue which one to pick now?
00:38:44  <piscisaureus_>4 even ,,,
00:42:43  <TooTallNate>piscisaureus_: arm build works
00:44:15  * mikealjoined
00:45:48  <piscisaureus_>TooTallNate: cool, thanks
00:45:59  <piscisaureus_>TooTallNate: now I have to figure out which patches to back out
00:46:02  <piscisaureus_>after that I will land
00:47:32  * mikealquit (Client Quit)
00:48:06  <CIA-129>node: Ben Noordhuis v0.8 * r4870a4e / doc/api/http.markdown : doc: http: expand request.headers documentation - http://git.io/p7_ubg
00:48:30  <CIA-129>node: Bert Belder openssl-1.0.1c-wip * r1f5a7dd / (5 files in 3 dirs): openssl: update the GYP build to work with openssl 1.0.1c - http://git.io/B2PRFw
00:48:30  <CIA-129>node: Bert Belder openssl-1.0.1c-wip * rf70a4b0 / (11 files in 2 dirs): openssl: remove obsolete patch files - http://git.io/Y3posg
00:54:15  <isaacs>TooTallNate: https://github.com/TooTallNate/node-lru-cache/commit/cb30000e9cadbbf8d02d5093deb953c1f2dd3b10
00:54:18  <isaacs>TooTallNate: that'll break my stuff.
00:54:31  <isaacs>TooTallNate: i'm using it to cache file descriptors, and when they fall off the cache, they get fs.close()'d
00:55:01  <isaacs>but i guess i could just be more careful about not closing them until they're done being used.
00:55:21  <TooTallNate>isaacs: well they'd never fall off immediately in that case then right?
00:55:38  <TooTallNate>cause a fd has constant size
00:55:45  <isaacs>hm...
00:55:53  <isaacs>right.
00:55:56  <TooTallNate>my case was a big Buffer, too big for the cache, immediately falls off
00:56:01  <isaacs>right
00:56:04  <isaacs>what did dispose() do in that case?
00:56:12  <TooTallNate>calls git_free()
00:56:15  <TooTallNate>on the Buffer
00:56:22  <isaacs>what's git_free?
00:56:31  <TooTallNate>part of libgit2 :)
00:56:34  <isaacs>ok
00:56:48  <isaacs>so, the tricky thing there is that you need to be careful yo udon't call dispose() until you're done using it
00:56:53  <isaacs>ie, don't put it in cache until you're through with it
00:57:16  <isaacs>in that case, maybe it's better to just hook your git_free() to the Buffer's gc
00:57:30  <TooTallNate>isaacs: well what's happening is the buffer is too big, never gets put in the cache (also dispose() is never called), so there's a mem leak
00:57:35  <isaacs>right
00:57:39  <TooTallNate>isaacs: the GC hook *is* an option
00:57:46  <TooTallNate>probably a good one at that
00:57:46  <isaacs>it's probably the better option
00:57:50  <TooTallNate>ya
00:57:59  <isaacs>dispose() is kind of a weird dangerous race condition
00:58:16  <isaacs>since otherwise you won't leak memory, you'll offset memory that isn't yours.
00:58:28  <isaacs>memory that's been free()d
00:58:50  * loladirojoined
00:59:07  <isaacs>and in this case, actually, if st has a lot of files (more than 1000 or so) it'll potentially close() them before they are done
00:59:12  <isaacs>that just hasn't happened yet.
00:59:27  <TooTallNate>isaacs: ok i think you've convinced me
00:59:31  <TooTallNate>patch withdrawn
00:59:36  <isaacs>heh
00:59:50  <TooTallNate>i just need to add the "callback" functionality to ref
00:59:53  <TooTallNate>or just use node-weak
00:59:55  <isaacs>hm. yeah.
01:00:26  <isaacs>and actually, i'm not convinced now that even turning down your patch makes my use case safe.
01:00:37  <isaacs>i still need to handle this
01:03:24  <TooTallNate>i mean i don't see how it would hurt… but i see the better solution
01:03:31  <isaacs>TooTallNate: yeah, your patch is actually good.
01:03:43  <isaacs>but you cna't rely on it
01:04:34  <TooTallNate>isaacs: so the inherent problem is that you fetch an item from a cache, and usually do async things with it, so we never really know when it's safe to call dispose()?
01:05:14  <isaacs>right
01:05:27  <isaacs>so... it's more like a weakref callback in a way
01:05:40  <isaacs>like, "This is no longer in the cache, so... you can delete it. it's ok with me if it's ok with you"
01:12:02  * dapquit (Quit: Leaving.)
01:16:48  * pooyaquit (Quit: pooya)
01:29:32  * abraxasjoined
01:41:49  * wavded_quit
01:42:26  * loladiroquit (Quit: loladiro)
01:47:46  * ArmyOfBrucequit (Excess Flood)
01:48:15  * ArmyOfBrucejoined
01:50:37  * ArmyOfBrucequit (Excess Flood)
01:51:06  * ArmyOfBrucejoined
01:55:31  <bnoordhuis>the ninja build is pretty noisy with the openssl upgrade
01:55:43  <bnoordhuis>lots of -Wmissing-field-initializers warnings
01:56:00  <piscisaureus_>huh
01:56:07  <piscisaureus_>die heb ik nergens gezien
01:56:58  <bnoordhuis>piscisaureus: what version of gcc?
01:57:18  * piscisaureusboots linux
01:57:44  <piscisaureus>bnoordhuis: 4.6.1
01:58:02  <bnoordhuis>ah, strange - i'm compiling with 4.6.3
01:59:29  <piscisaureus>bnoordhuis: maybe just add -Wno-missing-field-initializers ? :-)
01:59:33  <bnoordhuis>yeah, perhaps
01:59:42  <bnoordhuis>it's not really an issue, the warnings are harmless, just annoying
02:00:04  * mikealjoined
02:01:41  <bnoordhuis>piscisaureus: is sha1_armv4_large.patch only relevant for android?
02:01:58  <piscisaureus>bnoordhuis: as far as I can tell, indeed
02:02:26  <bnoordhuis>i'm trying to figure out what it does but...
02:02:53  <piscisaureus>bnoordhuis: you should probably find the patch in the android OS repos
02:02:58  <piscisaureus>bnoordhuis: I just reapplied it
02:04:15  <bnoordhuis>"This patch eliminates memory stores to addresses below SP." <- i don't know what to make of that
02:05:05  <bnoordhuis>i mean, i see that that is what it does. but why?
02:05:54  <bnoordhuis>http://comments.gmane.org/gmane.comp.encryption.openssl.devel/21365 <- interesting
02:07:06  * TooTallNatequit (Quit: Computer has gone to sleep.)
02:07:16  <bnoordhuis>piscisaureus: it got fixed in upstream openssl recently
02:07:23  <bnoordhuis>where 'recently' means august 17
02:07:27  <piscisaureus>bnoordhuis: aha
02:07:41  <piscisaureus>bnoordhuis: so we just float an upstream, that's fine i suppose
02:07:51  <piscisaureus>bnoordhuis: where is the source control for openssl. I couldn't find it
02:08:01  <bnoordhuis>piscisaureus: it's in cvs but there are a couple of git mirrors
02:09:08  * mikealquit (Quit: Leaving.)
02:09:56  * avalanche123joined
02:11:56  <CIA-129>node: Ben Noordhuis openssl-1.0.1c-wip * r1e36fe2 / deps/openssl/openssl/crypto/sha/asm/sha1-armv4-large.pl : openssl: apply upstream sha1-armv4-large.pl patch - http://git.io/ESYKOQ
02:13:08  * avalanche123quit (Read error: Connection reset by peer)
02:16:29  <piscisaureus_>bnoordhuis: weird. It would surprise me somewhat if the android maintainers *also* got the ABI wrong when they did the old patch.
02:21:38  <bnoordhuis>piscisaureus: oh sorry, i didn't revert the old patch
02:23:35  <CIA-129>node: Ben Noordhuis openssl-1.0.1c-wip * ra3fbfcc / deps/openssl/openssl/crypto/sha/asm/sha1-armv4-large.pl : openssl: apply upstream sha1-armv4-large.pl patch (+9 more commits...) - http://git.io/qOv3KA
02:25:14  <bnoordhuis>piscisaureus: ^ hope that's okay
02:25:28  <piscisaureus>bnoordhuis: what, exactly?
02:25:31  <piscisaureus>bnoordhuis: rebasing?
02:25:33  <bnoordhuis>piscisaureus: yes
02:25:36  <piscisaureus>bnoordhuis: I guess it's too late now
02:25:38  <piscisaureus>:-)
02:25:41  <piscisaureus>bnoordhuis: go aheada
02:26:00  <bnoordhuis>i can undo it, still have the old tip in my reflog
02:26:08  <piscisaureus>isaacs: you need to push tags for v0.8.9
02:26:16  <piscisaureus>bnoordhuis: nah, I know my rebases
02:28:20  * mikealjoined
02:28:27  * mjr_quit (Quit: mjr_)
02:28:30  * mikealquit (Client Quit)
02:31:12  * mikealjoined
02:34:02  <isaacs>piscisaureus: i did already?
02:34:15  <piscisaureus>hmm, v0.8.9 doesn't show up here somehow
02:35:00  <isaacs>piscisaureus: https://github.com/joyent/node/commit/v0.8.9
02:35:04  <isaacs>piscisaureus: git fetch ry --all ?
02:35:18  <piscisaureus>isaacs: ry? hehe
02:35:44  <piscisaureus>oh that works
02:35:47  <isaacs>piscisaureus: or "joyent" if you prefer
02:35:53  <isaacs>or "upstrea" or "origin" or whatever.
02:36:03  * isaacsmisses the good old days when "people" worked on "programs"
02:38:30  <bnoordhuis>piscisaureus: so what are you gonna vote today?
02:38:47  <piscisaureus_>bnoordhuis: sgp
02:38:52  <bnoordhuis>very good
02:38:57  <piscisaureus_>bnoordhuis: or partij van de toekomst
02:39:11  <bnoordhuis>mens
02:39:24  <bnoordhuis>^ that's a party
02:39:44  <piscisaureus>rly?
02:43:05  * joshthecoderjoined
02:50:34  * brsonquit (Quit: leaving)
03:00:27  * TooTallNatejoined
03:03:42  * mjr_joined
03:05:19  * c4miloquit (Remote host closed the connection)
03:06:02  * TooTallNatequit (Quit: ["Textual IRC Client: www.textualapp.com"])
03:15:12  <piscisaureus_>bnoordhuis: unfortunately there is no trace of the review/discussion of npn.patch.
03:25:06  <bnoordhuis>okay, off to bed
03:25:13  <bnoordhuis>sleep tight, bertje
03:25:18  <piscisaureus>yeah
03:25:20  <piscisaureus>late :-(
03:25:21  <piscisaureus>very late
03:25:47  <piscisaureus>bnoordhuis: when you wake up all the "drop" patches are gone
03:26:36  * CIA-129quit (Ping timeout: 240 seconds)
03:29:39  * bnoordhuisquit (Ping timeout: 252 seconds)
03:31:18  * mikealquit (Quit: Leaving.)
03:37:35  * piscisaureusquit (Quit: Lost terminal)
03:38:55  * piscisaureus_quit (Quit: ~ Trillian Astra - www.trillian.im ~)
04:01:23  * piscisaureus_joined
04:12:42  * FUUFREENODESTOPBjoined
04:13:41  * mjr_quit (Quit: mjr_)
04:16:05  * dshaw_quit (Quit: Leaving.)
04:23:52  * FUUFREENODESTOPBquit (Quit: Leaving...)
04:33:51  * piscisaureus_quit (Ping timeout: 248 seconds)
04:36:35  * loladirojoined
04:46:20  * mjr_joined
05:05:02  * FUUFREENODESTOPBjoined
05:16:10  * FUUFREENODESTOPBquit (Quit: Leaving...)
05:18:13  * mjr_quit (Quit: mjr_)
05:29:57  * loladiroquit (Quit: loladiro)
05:36:34  * paddybyersjoined
05:56:02  * philips_quit (Excess Flood)
05:59:20  * philips_joined
06:05:24  * theColequit (Read error: No route to host)
06:18:06  * mikealjoined
06:18:06  * paddybyersquit (Quit: paddybyers)
06:22:12  * joshthecoderquit (Quit: Linkinus - http://linkinus.com)
06:27:30  * paddybyersjoined
06:38:04  * travis-cijoined
06:38:04  <travis-ci>[travis-ci] joyent/libuv#670 (v0.8 - 1988f5e : Ben Noordhuis): The build is still failing.
06:38:04  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/12c25e1027e2...1988f5e581e4
06:38:04  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/2414830
06:38:04  * travis-cipart
06:56:23  * stagasjoined
06:57:21  * rendarjoined
07:02:35  * paddybyersquit (Quit: paddybyers)
07:03:43  * paddybyersjoined
07:12:07  * dshaw_joined
07:19:43  * FUUFREENODESTOPBjoined
07:20:09  * FUUFREENODESTOPBchanged nick to `3rdEden
08:13:06  <indutny>hoya
08:34:21  * stagasquit (Quit: ChatZilla 0.9.88-rdmsoft [XULRunner 1.9.0.17/2009122204])
08:52:21  * stagasjoined
09:52:20  * hzjoined
10:30:15  * AvianFluquit (Quit: AvianFlu)
10:55:53  * bentkusjoined
11:14:31  * stagasquit (Ping timeout: 246 seconds)
11:18:47  * mmaleckijoined
11:40:00  * stagasjoined
12:14:13  * bnoordhuisjoined
12:46:07  * bentkusquit (Ping timeout: 248 seconds)
12:46:33  * bentkusjoined
12:46:53  * rendar_joined
12:47:44  * c4milojoined
12:47:48  * ircretaryquit (Ping timeout: 240 seconds)
12:48:01  * dshaw_1joined
12:48:26  * dshaw_quit (Ping timeout: 240 seconds)
12:48:26  * rendarquit (Ping timeout: 240 seconds)
12:48:45  * philips_quit (Ping timeout: 240 seconds)
12:55:49  * philips_joined
12:58:23  * abraxasquit (Remote host closed the connection)
12:58:49  * loladirojoined
13:16:35  * loladiroquit (Quit: loladiro)
13:22:48  <indutny>bnoordhuis: hey
13:23:08  <indutny>bnoordhuis: I've fixed almost every issue
13:24:34  <indutny>bnoordhuis: what can I do about consts
13:24:57  <indutny>oh
13:25:02  <indutny>I see your replies now
13:35:53  <indutny>bnoordhuis: I'm going to change loop in fsevents.c too
13:36:05  <indutny>bnoordhuis: it'll split list in mutex, and operate further without locks
13:36:27  <indutny>running tests now
13:37:01  * loladirojoined
13:54:01  * loladiroquit (Quit: loladiro)
13:59:12  * piscisaureus_joined
13:59:47  * bentkusquit (Ping timeout: 246 seconds)
14:00:07  <indutny>piscisaureus_: hey man
14:00:46  <indutny>piscisaureus_: can you please test it on mac once again - https://github.com/joyent/libuv/pull/525 ?
14:05:36  * piscisaureus_quit (Ping timeout: 240 seconds)
14:11:34  * bentkusjoined
14:12:38  <bentkus>There is an async c torrent library which is designed to be hooked into async socket apis
14:12:42  <bentkus>does anybody know the name of it?
14:15:05  <tjfontaine>one of the libtorrents uses asio
14:22:37  * AvianFlujoined
14:22:57  * dapjoined
14:25:23  * dapquit (Client Quit)
14:25:48  * dapjoined
14:29:42  * `3rdEdenquit (Quit: rebooting)
14:32:45  * FUUFREENODESTOPBjoined
14:35:49  <tjfontaine>indutny: only process_title seems to fail for me on 10.8
14:35:56  * FUUFREENODESTOPBchanged nick to `3rdEden
14:40:28  * loladirojoined
14:49:55  * loladiroquit (Quit: loladiro)
14:55:48  * mjr_joined
15:05:16  * loladirojoined
15:12:47  * loladiroquit (Quit: loladiro)
15:17:01  * TheJHjoined
15:29:52  * dshaw_1quit (Quit: Leaving.)
15:33:09  <indutny>tjfontaine: cool, thank you
15:33:21  * kristatejoined
15:36:37  <indutny>bnoordhuis: indeed it does leak memory
15:36:40  <indutny>bentkus: but github is down
15:36:46  <indutny>at least for me
15:36:55  <indutny>or
15:37:44  <indutny>oh, I just had dns problems
15:39:48  <bentkus>github had the last 2 days problems
15:40:07  <indutny>yeah
15:41:37  * stagas_joined
15:42:53  * stagas_quit (Read error: Connection reset by peer)
15:43:28  <indutny>bnoordhuis: sent patch in email
15:43:38  <bnoordhuis>ah, so you did
15:43:50  <bnoordhuis>coding like it's 1999
15:43:51  * stagas_joined
15:43:56  * stagasquit (Ping timeout: 245 seconds)
15:43:56  <indutny>bnoordhuis: yeah
15:44:02  <indutny>bnoordhuis: very comfortable and secure way
15:44:07  * stagas_changed nick to stagas
15:44:08  <indutny>bnoordhuis: does github works for you?
15:44:14  <bnoordhuis>yes
15:44:18  <indutny>odd
15:44:45  <bnoordhuis>try 8.8.8.8 as your dns server
15:45:07  <indutny>already did
15:45:18  <indutny>ping github.com
15:45:18  <indutny>PING github.com (207.97.227.239): 56 data bytes
15:45:18  <indutny>Request timeout for icmp_seq 0
15:45:18  <indutny>Request timeout for icmp_seq 1
15:45:18  <indutny>Request timeout for icmp_seq 2
15:45:32  <indutny>ping google.com
15:45:32  <indutny>PING google.com (173.194.35.192): 56 data bytes
15:45:32  <indutny>64 bytes from 173.194.35.192: icmp_seq=0 ttl=56 time=10.459 ms
15:45:32  <indutny>64 bytes from 173.194.35.192: icmp_seq=1 ttl=56 time=16.695 ms
15:45:33  <bnoordhuis>hah, patches in non-monospace fonts are so terrible
15:45:37  <indutny>hahaha
15:45:46  <indutny>I'll try rebooting my wifi router
15:45:47  <indutny>brb
15:45:49  * bnoordhuisstarts pine
15:47:57  <tjfontaine>I think there's a gmail labs option to enable monospace
15:48:26  <tjfontaine>hm or there was once
15:49:18  <indutny>ok
15:49:48  <indutny>crap
15:49:57  <indutny>it doen't work from russia
15:51:26  <indutny>going to reboot
15:52:35  <indutny>oh
15:52:37  <indutny>that's impossible
15:52:42  <indutny>it just doesn't work
15:54:12  <indutny>tjfontaine: what's a github ip for you?
15:54:57  <tjfontaine>207.97.227.239
15:55:01  <tjfontaine>but it's probably anycast
15:55:19  <indutny>same for me
15:57:20  * stagasquit (Ping timeout: 244 seconds)
15:57:59  * stagasjoined
15:59:20  * stagas_joined
16:00:50  <isaacs>good morning
16:01:06  <indutny>morning
16:01:48  <isaacs>bnoordhuis, piscisaureus: maxogden is in your country today, i think. you all should hang out :)
16:02:18  * stagasquit (Ping timeout: 252 seconds)
16:02:21  <bnoordhuis>isaacs: where in the netherlands though?
16:02:31  <bnoordhuis>i mean, we're a small country but not *that* small
16:03:32  <bentkus>how big is a socket on windows?
16:03:53  <isaacs>oh, i thought netherdlands was like a neighborhood in the city of europe.
16:03:54  <isaacs>no?
16:03:56  <indutny>bnoordhuis: don't tell me about that
16:04:22  * stagas_quit (Ping timeout: 246 seconds)
16:04:31  <indutny>bnoordhuis: siberia is like all netherlands
16:04:59  <isaacs>siberia's way more nether, that's for sure.
16:05:39  <indutny>Total 139,700 km2
16:05:42  <isaacs>though, nether originally meant low, then in english at least, gradually took on the connotation of being far away or distant.
16:05:46  <indutny>it's only omskaya district
16:05:58  <indutny>41,543 km2
16:06:02  <indutny>that's whole netherlands
16:06:08  * stagas_joined
16:06:29  <indutny>and yeah
16:06:32  <indutny>siberia is nether
16:07:35  <indutny>http://en.wikipedia.org/wiki/Omsk_Oblast
16:09:22  <mmalecki>indutny: hey
16:09:26  <indutny>hoi
16:09:31  <mmalecki>indutny: do you want a server you can proxy through?
16:09:40  <indutny>mmalecki: dunno
16:09:43  <indutny>mmalecki: give me it's IP
16:09:45  <mmalecki>we like, work for a cloud company or something, I think
16:09:52  <indutny>mmalecki: my one in joyent cloud isn't reachable for me
16:10:03  <indutny>and jit.su doesn't work for me too
16:10:12  * loladirojoined
16:10:34  <mmalecki>that's very interesting, does it resolve at all?
16:10:45  <indutny>yes
16:10:46  <indutny>it does
16:10:56  <indutny>Apparently, my ISP has routing problems
16:11:36  * stagas_quit (Ping timeout: 268 seconds)
16:13:34  * stagas_joined
16:14:02  <indutny>bnoordhuis: so does this patch work for you?
16:14:41  <indutny>I can't even post about my ISP on twitter
16:14:52  <indutny>because it's down for me too
16:14:57  * joshthecoderjoined
16:15:21  <bentkus>long live russia
16:19:46  * stagas_quit (Ping timeout: 246 seconds)
16:21:59  <indutny>long live beeline ISP
16:24:23  * stagasjoined
16:27:10  <indutny>isaacs: going to bring crypto in modern age
16:27:23  <indutny>isaacs: Buffers
16:27:24  <isaacs>indutny: love it
16:27:38  <isaacs>indutny: shim whatever you have to to keep backwards compatibility, though
16:27:47  <indutny>isaacs: ok
16:27:51  <indutny>isaacs: I'll try :)
16:28:08  <isaacs>if that's not possible, then we can just have it be some kludgey trick that upgraders have to do
16:28:36  <isaacs>but start with the correct API, and then shim for compatibility
16:32:18  <indutny>ok
16:37:47  <indutny>bnoordhuis: switched to reserve 3g connection
16:37:53  <indutny>bnoordhuis: and force pushed Pull Request
16:38:07  <indutny>it's similar to what I've sent you by email
16:41:13  * joshthecoderquit (Quit: Leaving...)
16:48:10  * mikealquit (Quit: Leaving.)
16:56:24  * loladiroquit (Quit: loladiro)
16:59:41  * hzquit
17:10:28  * bentkusquit (Ping timeout: 260 seconds)
17:12:01  * joshthecoderjoined
17:14:40  * mikealjoined
17:17:12  * bentkusjoined
17:21:53  * dshaw_joined
17:26:15  * TooTallNatejoined
17:30:56  <bentkus>freaking fd_set
17:33:35  * brsonjoined
17:35:48  * loladirojoined
17:39:20  <TooTallNate>bnoordhuis: i'm running `make test` on my pi with the openssl 1.0.1c stuff now
17:40:41  <bentkus>do we have to applaud?
17:53:27  <TooTallNate>bnoordhuis: here's one relevant failure https://gist.github.com/f2eb3e9ec54689427e2d
17:56:06  * hzjoined
18:04:57  <TooTallNate>bnoordhuis: the rest look like they're timing-related since this thing isn't so fast
18:05:16  <TooTallNate>bnoordhuis: you think we should loosen some of the timeout values in the tests in the repo?
18:14:10  * AutowiredConstrujoined
18:16:19  * dapquit (Quit: Leaving.)
18:19:04  * hij1nx_joined
18:19:17  * hij1nx_quit (Excess Flood)
18:30:31  * loladiroquit (Quit: loladiro)
18:32:35  * loladirojoined
18:32:35  * loladiroquit (Client Quit)
18:32:41  * paddybyersquit (Quit: paddybyers)
18:37:09  * paddybyersjoined
18:43:30  * `3rdEdenquit (Quit: stop, hammer time)
18:59:31  * dapjoined
19:03:56  * piscisaureus_joined
19:04:07  <piscisaureus_>indutny: bnoordhuis: https://chromiumcodereview.appspot.com/10825183
19:04:16  <indutny>das v8 hacker
19:04:17  <piscisaureus_>yes, I know you read hacker news too
19:04:23  <indutny>yeah
19:04:34  <indutny>I don't think we're using compression at all
19:04:48  <indutny>but I wonder how is it exploitable
19:04:59  <indutny>and why is it simplier to turn it off, instead of fixing openssl
19:05:03  <indutny>probably, a protocol error
19:05:56  <piscisaureus_>indutny: I don't know, we don't turn it off either
19:06:03  <indutny>really?
19:06:23  <indutny>compression. Saves memory - do it in userland.
19:06:24  <indutny>#if !defined(OPENSSL_NO_COMP)
19:06:24  <indutny> STACK_OF(SSL_COMP)* comp_methods =
19:06:29  <indutny>err
19:06:36  <indutny> // Turn off compression. Saves memory - do it in userland.
19:06:41  <indutny>that's from node_crypto.cc
19:09:54  <indutny>piscisaureus_: comments?
19:10:04  <piscisaureus_>indutny: about what?
19:10:12  <indutny>piscisaureus_: about this
19:10:19  <piscisaureus_>indutny: OPENSSL_NO_COMP
19:10:27  <piscisaureus_>weird that the flag got renamed :-)
19:10:30  <indutny>and we're doing pretty the same thing
19:10:34  <piscisaureus_>anyway
19:10:48  <piscisaureus_>indutny: ok, well that's good
19:10:56  <indutny>ok
19:10:57  <piscisaureus_>indutny: no worries then
19:11:01  <indutny>but that's pretty interesting
19:11:10  <indutny>I'm quite sure that everyone has copy-pasted this option
19:11:15  <indutny>in their project
19:11:19  <indutny>and this is fun
19:11:33  <indutny>like, google can make reverse commit after two days
19:11:44  <indutny>with a link to rickroll video
19:11:52  <indutny>that would be awesome
19:12:05  <piscisaureus_>hahaha
19:12:48  <indutny>piscisaureus_: https://chromiumcodereview.appspot.com/10825183/
19:12:52  <indutny>look at bottom comment
19:13:11  <piscisaureus_>:-)
19:13:15  <piscisaureus_>they don't know you
19:13:26  <piscisaureus_>they will now think indutny is the biggest troll of all time
19:13:40  <indutny>piscisaureus_: that's unfair
19:13:48  <piscisaureus_>why
19:13:51  <indutny>bnoordhuis is the biggest troll
19:13:59  <piscisaureus_>yes but he didn't comment
19:14:08  <TooTallNate>piscisaureus_: only 1 ssl-related error on arm with v1.0.1c
19:14:16  <piscisaureus_>TooTallNate: yeah, cool
19:14:20  <piscisaureus_>TooTallNate: I saw that
19:14:26  <TooTallNate>cool
19:14:42  <piscisaureus_>TooTallNate: thanks for testing btw. And I have no clue whether this is a genuine failure, or how to fix it ...
19:14:50  <TooTallNate>haha
19:14:53  <TooTallNate>ya me either :p
19:14:56  <TooTallNate>maybe indutny
19:15:27  <piscisaureus_>TooTallNate: what happens if you run that specific test 10 times in a row?
19:15:33  <indutny>sure, I am here to fix any openssl shit that happend to you
19:15:36  <piscisaureus_>TooTallNate: is the failure consistent
19:15:46  <piscisaureus_>\o
19:24:11  <indutny>btw
19:24:16  <indutny>I don't like design of our website
19:24:18  <indutny>isaacs: ^
19:24:22  <indutny>we should look more mature
19:24:23  <indutny>like
19:24:24  <indutny>http://www.php.net/
19:24:52  <indutny>you see, people are visiting that site and immediately getting understanding of how old the platform is
19:24:55  <indutny>node.js is old too
19:25:06  <indutny>though, not as old as hypertext preprocessor is
19:25:52  * dshaw_1joined
19:26:53  * dshaw_quit (Ping timeout: 264 seconds)
19:27:04  <indutny>isaacs: yt?
19:27:22  <indutny>isaacs: hash.update(...).digest('buffer') or just .digest()
19:28:27  <indutny>or
19:28:30  <indutny>remove 'binary'
19:28:39  <indutny>and make it buffer instead
19:28:52  <indutny>I think I'm going to try this path
19:33:29  <isaacs>indutny: hey
19:33:46  <isaacs>i think bufer should just be the default. that'll break anyone depending on 'binary' output, but whatever.
19:33:46  <indutny>ho
19:34:01  <indutny>so, leave 'binary' there
19:34:01  <isaacs>mostly people use hex or convert to a buffer or pass into another crypto function anyway
19:34:04  <indutny>but output buffer by default
19:34:10  <isaacs>yeah
19:34:12  <indutny>ok
19:34:14  <isaacs>like every other part of our api
19:34:25  <isaacs>that's going to break some people, but let's try it and see how bad it is
19:35:49  <AutowiredConstru>like every other part of our api => horrible
19:36:10  <indutny>рфрф
19:36:11  <indutny>haha
19:36:13  <indutny>indeed
19:36:19  <mmalecki>AutowiredConstru: huh?
19:36:30  <AutowiredConstru>sorry
19:36:46  <mmalecki>i find node.js api to be one of the most pleasant APIs to work in
19:59:31  <mjr_>crypto has just been ignored for too long, and it's time to modernize it.
20:01:18  * pfewpepwepwpepwjoined
20:01:27  * pfewpepwepwpepwchanged nick to `3rdEden
20:02:48  <indutny>mjr_: ++
20:02:51  <indutny>mjr_: doing this right now
20:02:53  <indutny>isaacs: wanna review?
20:03:00  <indutny>I'm running tests now
20:03:11  <indutny>PR should be sent in ~5 minutes
20:06:22  * loladirojoined
20:06:34  * dshaw_joined
20:06:49  * hzquit
20:07:35  <bnoordhuis>TooTallNate: does simple/test-tls-server-verify fail consistently or randomly?
20:07:43  * dshaw_1quit (Ping timeout: 252 seconds)
20:16:05  <bnoordhuis>oh, i guess bert already asked that
20:21:36  * rendar_quit
20:21:54  * rendarjoined
20:25:04  * kristatequit (Ping timeout: 246 seconds)
20:29:35  * `3rdEdenquit (Quit: Linkinus - http://linkinus.com)
20:30:58  <TooTallNate>bnoordhuis: piscisaureus_ it's consistent it looks like
20:31:07  <bnoordhuis>well, bummer
20:31:19  <bnoordhuis>i'll dig up an arm device somewhere and try to reproduce
20:35:00  <isaacs>AutowiredConstru: if you think the node api is horrible, then this is the right room to be in. please be prepared to elaborate, though, and perhaps to be disagreed with.
20:35:19  <AutowiredConstru>i meant libuv api
20:35:28  <isaacs>AutowiredConstru: but i do actually like criticism. it's ok to be pretty agressive with it even, as long as it's constructive.
20:35:42  <isaacs>AutowiredConstru: same rules apply, then :)
20:36:12  <isaacs>node is libuv in javascript
20:36:24  <isaacs>of course, javascript != C
20:36:27  <isaacs>so there's big differences
20:36:47  <bnoordhuis>piscisaureus_: seen the exit polls?
20:36:53  <piscisaureus_>bnoordhuis: yep
20:37:03  <piscisaureus_>bnoordhuis: I am looking at the tv in the background
20:37:14  <bnoordhuis>piscisaureus_: so what did you vote?
20:37:16  <piscisaureus_>(I was surprised it still works :-))
20:37:28  <piscisaureus_>bnoordhuis: what did you vote?
20:37:32  <bnoordhuis>i developed a conscience at the very last moment and voted d66 :/
20:38:53  <bentkus>i'm done integrating cares into my bindings
20:39:04  <bentkus>does anyone know what the ns, srv, txt and ns replies are?
20:39:23  <bnoordhuis>bentkus: you mean protocol wise or the c-ares functions?
20:39:28  <indutny>isaacs: bnoordhuis: piscisaureus_: https://github.com/joyent/node/pull/4007
20:39:30  <indutny>TooTallNate: ^
20:39:41  <bentkus>protocolwise
20:40:06  <bentkus>i tried them out, but the server just returned empty buffers
20:40:08  <indutny>in next commit I'll fix style in node_crypto.cc
20:40:16  <bnoordhuis>bentkus: ns=name servers, txt=text, srv=varies
20:40:27  <bnoordhuis>bentkus: not all domains have them
20:41:32  <bentkus>Yeah, it would be nice to get some domains which support these 4 so I could test my code
20:42:07  <bnoordhuis>bentkus: you can use for _jabber._tcp.google.com for srv
20:42:28  <piscisaureus_>TooTallNate: you have an actual arm device to test on?
20:42:28  <indutny>bentkus: probably I can help you
20:42:44  <piscisaureus_>TooTallNate: or do you use some form of virtualization?
20:42:49  <TooTallNate>piscisaureus_: yes, i can give you or ben access if you like
20:42:58  <TooTallNate>it's a raspberry pi sitting at home
20:43:12  <TooTallNate>it's running http://n8.io though so don't crash the entire machine please :)
20:43:13  <piscisaureus_>so you log in over ssh?
20:43:18  <TooTallNate>piscisaureus_: yup
20:43:18  <piscisaureus_>that'd be cool
20:43:25  <TooTallNate>piscisaureus_: send me your key
20:44:36  <bnoordhuis>bentkus: oh, and gmail.com for txt
20:45:11  <piscisaureus_>TooTallNate: https://gist.github.com/bc1e11efd69acb93e069
20:46:00  <TooTallNate>piscisaureus_: ok you can log in with `ssh pi@n8.io`
20:46:30  <piscisaureus_>TooTallNate: it asks for my pwd ...
20:46:55  <TooTallNate>piscisaureus_: one more time
20:47:05  <piscisaureus_>TooTallNate: nope
20:47:07  <TooTallNate>maybe authorized_keys doesn't like blank lines in it
20:47:09  <TooTallNate>wtfff!!!
20:47:13  <TooTallNate>is that the right key?
20:47:25  <piscisaureus_>ehm, yes, I use it everywhere...
20:49:00  <piscisaureus_>TooTallNate: I can use it to log in to node@nodejs.org
20:49:41  <TooTallNate>well that's very strange
20:49:52  * paddybyersquit (Read error: Network is unreachable)
20:49:52  <TooTallNate>i'd like to think i know how to add ssh keys...
20:50:10  * paddybyersjoined
20:50:15  <piscisaureus_>TooTallNate: add it to ~/.ssh/authorized_keys :-)
20:51:57  <TooTallNate>piscisaureus_: i mean this is the whole damn file https://gist.github.com/f25cf5545139b2ef6af8
20:53:48  * mbalhojoined
20:53:53  <mbalho>WHATS UP NETHERLANDIANS
20:54:15  <isaacs>indutny: I think the docs are a bit confusing as well, saying that it defaults to "buffer"
20:54:16  <mbalho>it is i, maxogden the bearded guy who takes compromising photos of isaacs cats
20:54:24  <isaacs>indutny: does it work if you pass "buffer" as a string?
20:55:18  * dshaw_quit (Quit: Leaving.)
20:55:33  <bnoordhuis>mbalho: hey, are you in amsterdam tomorrow?
20:55:46  <mbalho>yep and saturday and sunday
20:55:49  <TooTallNate>piscisaureus_: sorry i figured it out
20:55:53  <TooTallNate>it's port 2222
20:56:21  <TooTallNate>i have it defined as "pi" in my ssh config so I forgot :p
20:57:31  <mbalho>whoa this room is like all the people i talk to in node.js without the dummies
20:57:45  <bentkus>should I concat all results from ares_txt_reply?
20:58:10  <TooTallNate>mbalho: what's "balho"?
20:58:17  <TooTallNate>i can guess what "m" stands for :)
20:58:18  <mbalho>bridger abraham leonidas herb ogden
20:58:28  <mbalho>its a sweet hacker username like rms
20:58:34  <TooTallNate>wow
20:58:57  <mbalho>have you seen my beard lately? i am from a total hippie family
20:59:47  <bnoordhuis>mbalho: drop by the c9 office tomorrow and we'll go out for beers
21:00:11  <mbalho>i get in at like 6pm, how late will you be there?
21:00:27  <bnoordhuis>certainly not before noon
21:00:28  <mbalho>if you have an android you should put @gather on your phone and then make a gathering and i'll show up to it :)
21:00:36  <bnoordhuis>i'm a gentleman and there are standards to uphold
21:00:38  <mbalho>(or iphone)
21:00:47  <piscisaureus_>TooTallNate: not working
21:00:52  <bnoordhuis>oh, i don't have one of those fancy phones
21:01:12  <mbalho>your life is better without one to be honest (im sure you know this)
21:01:30  * dshaw_joined
21:01:34  <bnoordhuis>i do :) i have a nokia 3310 (i think) and it's already probing the limits of my technical abilities
21:02:22  <TooTallNate>piscisaureus_: what happens?
21:02:44  <bnoordhuis>mbalho: can you drop by around 18:45 or 19:00?
21:02:46  <piscisaureus_>TooTallNate:
21:02:48  <piscisaureus_>C:\Users\Bert Belder\.ssh>ssh pi@n8.io
21:02:48  <piscisaureus_>pi@n8.io's password:
21:03:00  <TooTallNate>piscisaureus_: ssh pi@n8.io -p 2222
21:03:03  <bnoordhuis>piscisaureus_: add -p 2222
21:03:27  <piscisaureus_>TooTallNate: yep, works. thanks
21:03:51  <TooTallNate>piscisaureus_: awesome :) just don't kill the web server that's running :)
21:05:41  <piscisaureus_>i'll let bnoordhuis have it's way first
21:06:01  <bnoordhuis>piscisaureus_: no need, i have arm gear myself
21:06:08  <bnoordhuis>no pi though
21:06:31  * paddybyersquit (Quit: paddybyers)
21:07:33  <bentkus>[5~[5~[6~
21:14:20  * mikealquit (Quit: Leaving.)
21:16:57  <mbalho>bnoordhuis: there is some sort of speaker dinner tomorrow that adobe is payin for, wanna go?
21:17:27  <bnoordhuis>mbalho: sure, if the food is good, the booze plentiful and the speakers interesting
21:17:35  <bnoordhuis>where is it and what is it about?
21:18:32  <mbalho>bnoordhuis: all i know is they're supposed to pick me up in a canal boat at 18:45 in front of Hotel Estherea
21:18:47  <bnoordhuis>mbalho: sounds fancy
21:19:11  <indutny>isaacs: yeah, they're
21:19:17  <mbalho>canal boats: transport of the aristocracy
21:19:33  <bnoordhuis>piscisaureus_: want to go? it's not far from the office
21:19:37  <mbalho>"the speakers' dinner will be on a (moving) canal boat
21:19:37  <mbalho>around the same time you land, Martijn and I will keep in touch when
21:19:37  <mbalho>he's driving to Amsterdam. He'll probably drop you off somewhere we
21:19:37  <mbalho>can pick you up with the boat, but we don't know where yet, since that
21:19:39  <mbalho>depends on your exact arrival time."
21:19:41  <mbalho>OOPS
21:19:43  <mbalho>newlines sorry
21:19:52  <piscisaureus_>bnoordhuis: a sec
21:22:49  <mbalho>bnoordhuis: adobe said its cool if you guys wanted to crash the boat party
21:23:06  <bentkus>ok now some working input for ns?!
21:23:12  <bnoordhuis>cool, crashing boats is something i excel at
21:23:14  <bnoordhuis>parties too
21:23:20  <mbalho>hah
21:23:26  <indutny>isaacs: nope, it doesn't
21:23:36  <bnoordhuis>bentkus: any domain, try `dig google.com ns`
21:23:39  <indutny>isaacs: I think I'll fix docs tomorrow
21:23:43  <indutny>feeling sick
21:23:53  <indutny>bnoordhuis: piscisaureus_: mind landing fsevents?
21:24:24  <bnoordhuis>indutny: no, but i can't verify that it actually compiles and/or works :)
21:24:34  <indutny>piscisaureus_: I know who can
21:24:35  <indutny>;)
21:24:44  <indutny>and I think tjfontaine tested it too
21:25:10  * loladiroquit (Quit: loladiro)
21:25:27  <indutny>ok
21:25:30  <indutny>going to sleep
21:25:31  <indutny>ttyl
21:25:32  <piscisaureus_>bnoordhuis: so what's that you want to go to tomorrow?
21:25:50  <bnoordhuis>piscisaureus_: read what mbalho pasted
21:25:52  <piscisaureus_>indutny: I'll land it. I recall TooTallNate saying it works.
21:26:40  * mmaleckiquit (Ping timeout: 246 seconds)
21:26:44  * TooTallNatequit (Ping timeout: 240 seconds)
21:26:50  <bentkus>o yeah, my code already worked, just tried to print out the wrong fields
21:27:18  <bentkus>what about SOA and PTR?
21:27:38  <piscisaureus_>ah
21:27:51  <piscisaureus_>mbalho: bnoordhuis: +1 for crashing the adobe party
21:28:10  <mbalho>sweeet
21:29:55  <bnoordhuis>bentkus: soa=any domain, ptr is for reverse lookups
21:30:03  <bnoordhuis>try e.g. `dig 8.8.8.8.in-addr.arpa. ptr`
21:30:43  * TooTallNatejoined
21:31:13  <bentkus>thanks
21:32:22  * dshaw_quit (Quit: Leaving.)
21:32:47  * dshaw_joined
21:39:50  <AutowiredConstru>http://www.infoq.com/presentations/Node-js-Windows-Azure
21:41:57  <bnoordhuis>yeees?
21:41:57  * travis-cijoined
21:41:57  <travis-ci>[travis-ci] joyent/libuv#675 (master - f8e7513 : Fedor Indutny): The build passed.
21:41:57  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/57e6113683bb...f8e7513a0618
21:41:57  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/2426530
21:41:57  * travis-cipart
21:44:34  <bnoordhuis>is the CIA bot MIA again?
21:44:49  <piscisaureus_>yeah it's dead
21:45:02  <piscisaureus_>neither node nor uv commits show up
21:46:05  <bentkus>too bad
21:46:06  <bentkus>:(
21:50:27  * rendarquit
21:52:57  <piscisaureus_>TooTallNate's pi is not in a hurry
21:52:58  <AutowiredConstru>the bot is not powered by libuv nor node.js so no wonder why it fails so hardly
21:59:37  <bentkus>xD
21:59:47  <bentkus>is it powered by ruby?
21:59:55  <bentkus>since it is an github service it has to be
22:01:52  <isaacs>indutny: i'm sorry. get some rest, man.
22:03:04  <saghul>should we have uv_udp_connect and uv_udp_send / sendto ?
22:03:26  <saghul>to avoid passing the address and port every time
22:03:48  <piscisaureus_>saghul: if we'd do that I'd rather rename uv_tcp_connect to uv_connect
22:03:53  <piscisaureus_>saghul: and use uv_write
22:04:16  <piscisaureus_>saghul: i could live with that if we made udp an uv_stream_t
22:04:43  <saghul>piscisaureus_but udp is not really a tream, is it? (conceptually, that is)
22:04:50  <saghul>s/tream/stream/
22:04:58  <piscisaureus_>saghul: well, it isn't atm but we could make it so
22:05:18  <saghul>piscisaureus_ interesting
22:05:45  <piscisaureus_>saghul: but - I think this is not really a priority for anyone atm.
22:06:09  <saghul>piscisaureus_ yeah, I was just wondering :-)
22:14:08  <piscisaureus_>bnoordhuis: speaking of crashing parties... some of these election parties look interesting
22:14:12  <piscisaureus_>theres free booz too
22:14:59  <bnoordhuis>the heresy that is udp streams again? >:(
22:15:51  <isaacs>piscisaureus_: is openssl 1.0.1c done merging into master?
22:16:08  <isaacs>or is there still stuff left to do with it
22:16:10  <piscisaureus_>isaacs: yeah, although we need to investigate a test failure on arm
22:16:17  <isaacs>feh. arm.
22:16:26  <piscisaureus_>isaacs: well, we need to enable ASM code
22:16:29  <isaacs>whatever, it's not like anyone actually uses those ;)
22:16:42  <piscisaureus_>isaacs: but the state of affairs is the same as it was in that regard
22:16:51  <isaacs>ok
22:17:01  <piscisaureus_>isaacs: but yeah the upgrade as such is done
22:17:06  <isaacs>you had some ideas about enabling asm code on 0.8 as well? or am i remembering it wrong?
22:17:18  <piscisaureus_>isaacs: better even. I have a patch :-)
22:17:23  <isaacs>huzzah!
22:17:27  <isaacs>let's do it.
22:17:41  <isaacs>and indutny's session thing
22:17:47  <isaacs>and call it 0.8.10
22:18:03  <piscisaureus_>isaacs: https://github.com/joyent/node/pull/3981
22:18:10  <piscisaureus_>isaacs: not sure if it's completely done
22:19:05  <saghul>cc1: error in backend: Global variable 'use_emfile_trick.12635' has an invalid section specifier '.data.read_mostly': mach-o section specifier requires a segment and section separated by a comma.
22:19:11  <saghul>Anyone seen this? ^
22:19:18  <bentkus>vote for the republicans
22:19:28  <saghul>That is OSX 10.6.8, gcc 4.2.1
22:20:27  <piscisaureus_>saghul: that looks like a compiler bug to me (or rather, an incompatibility between llvm and backend)
22:20:53  <piscisaureus_>saghul: you can search the code for read_mostly (it's a define somewhere) and make it empty
22:21:35  <saghul>piscisaureus_ will do
22:21:50  <piscisaureus_>saghul: if that fixes it then we have to berate bnoordhuis
22:22:08  <saghul>I'll add a && !defined(__APPLE__)
22:22:17  <saghul>though it may work with clang
22:22:36  <piscisaureus_>i doubt it
22:23:31  <bnoordhuis>piscisaureus_: berate apple you mean?
22:24:00  <piscisaureus_>bnoordhuis: apple too, yeah
22:24:04  <bnoordhuis>saghul: __read_mostly is defined in src/unix/internal.h
22:24:13  <piscisaureus_>ah
22:24:22  <piscisaureus_>the pi finally managed to compile openssl
22:24:49  <bnoordhuis>i compiled node on my nokia n900 a while ago
22:24:54  <bnoordhuis>took the better part of the night
22:25:05  <isaacs>piscisaureus_, bnoordhuis: Do we enable tls compression by default?
22:25:10  <bnoordhuis>it kind of surprised me that it didn't run out of memory halfway through
22:25:13  <piscisaureus_>isaacs: no it's off
22:25:15  <bnoordhuis>isaacs: no, it's disabled
22:25:16  <isaacs>kewl.
22:25:23  <isaacs>thanks. i thought so, but watned to sanity check.
22:26:04  <piscisaureus_>hmm
22:26:08  <piscisaureus_>the pi also has no screen command
22:29:21  <saghul>bnoordhuis I modified line 31 by adding && !defined(__APPLE__) to it but it got worse: https://gist.github.com/3710435
22:29:35  * saghulis clueless now :-S
22:32:20  <piscisaureus_>saghul: I think you *dont* want the #define __attribute__(y) line
22:32:34  <bnoordhuis>saghul: what piscisaureus_ said
22:32:59  <AutowiredConstru>i think what bnoordhui s said is correct
22:33:31  <saghul>piscisaureus_ bnoordhuis you got it, works now
22:33:55  <bnoordhuis>piscisaureus_: are you going to land the openssl upgrade?
22:34:02  <bnoordhuis>oh, you did
22:34:03  <piscisaureus_>bnoordhuis: it already did
22:34:04  <bnoordhuis>bloody CIA
22:34:06  <piscisaureus_>cia is dead
22:34:11  <saghul>but you guys compiled libuv just fine with ML, right?
22:34:14  <bnoordhuis>we hardly knew ye
22:34:37  <bnoordhuis>saghul: yes
22:35:01  * loladirojoined
22:35:10  <saghul>bnoordhuis sigh, so seems to be a llvm-gcc-4.2.1 issue or something
22:35:30  <saghul>bnoordhuis I trust you guys know better than I here ;-)
22:35:30  <piscisaureus_>saghul: ML ?
22:35:40  <saghul>piscisaureus_ Mountain Lion
22:35:43  <piscisaureus_>oh
22:35:43  <bnoordhuis>piscisaureus_: mountain lion, not the programming language
22:35:49  <saghul>which uses clang
22:36:08  <piscisaureus_>saghul: ah, I use ML but I have gcc installed
22:36:18  <piscisaureus_>not sure what is being used actually
22:36:34  <piscisaureus_>saghul: I though ML == Marc L.
22:36:42  <saghul>xD
22:37:17  <piscisaureus_>saghul: I'd be fine with landing that patch, ideally you'd make it not overly broad
22:37:41  <piscisaureus_>saghul: is it possible to detect clang or a particular clang version range
22:38:06  <saghul>piscisaureus_ let me try to write it nicely :-)
22:38:13  <piscisaureus_>er, replace clang by llvm :-)
22:38:17  <saghul>piscisaureus_ cc uses clang on ML, btw
22:38:21  <bnoordhuis>saghul, piscisaureus_: i'm fine with removing __read_mostly
22:38:30  <bnoordhuis>it's a micro-optimization
22:38:46  <piscisaureus_>I'd call it a nano optimization
22:38:52  <bnoordhuis>pico optimization
22:38:59  <saghul>bnoordhuis femto
22:39:02  <piscisaureus_>atto
22:41:01  <saghul>bnoordhuis would you like a patch that reverts 5f8185ab and the usage of __read_mostly?
22:41:41  <bnoordhuis>saghul: sure
22:41:51  <saghul>ok, coming
22:41:54  <isaacs>does test/pummel/test-debugger-debug-brk.js fail for anyone else?
22:42:25  <bnoordhuis>isaacs: yes
22:42:29  <isaacs>k
22:42:30  <bnoordhuis>it's been broken for a while
22:42:40  <isaacs>yeah, seems that way
22:42:58  <isaacs>0.8 merge
22:43:06  * isaacschanged nick to CIA
22:43:36  * CIAchanged nick to Guest97662
22:43:42  <Guest97662>https://github.com/joyent/node/commit/bb207c2827cd758dd273adb28badeb9ec2586939
22:43:49  * Guest97662changed nick to CIA-12342312
22:43:51  <CIA-12342312>https://github.com/joyent/node/commit/bb207c2827cd758dd273adb28badeb9ec2586939
22:43:54  * CIA-12342312changed nick to isaacs
22:49:58  * TheJHquit (Ping timeout: 246 seconds)
22:55:10  <saghul>indutny, bnoordhuis src/unix/fsevents.c:173: warning: overflow in implicit constant conversion
22:55:25  <bnoordhuis>hmm
22:55:58  * loladiroquit (Quit: loladiro)
22:57:00  * piscisaureus__joined
22:57:03  * piscisaureus_quit (Read error: Connection reset by peer)
22:59:05  <piscisaureus__>bnoordhuis: isaacs: I am going to land the openssl-asm patch on 0.8, unless you speak up now
22:59:13  <piscisaureus__>(after rebasing, that is)
23:02:44  * TooTallNatequit (Quit: Computer has gone to sleep.)
23:10:25  * loladirojoined
23:10:37  <bentkus>duck duck go
23:14:02  <piscisaureus__>HA
23:17:32  * CIA-131joined
23:18:09  <piscisaureus__>bnoordhuis: isaacs: nvm. actually I found another issue
23:24:00  <CIA-131>node: Felix Böhm master * rf5c68b2 / benchmark/url.js : bench: add url parser benchmark - http://git.io/otlhnA
23:24:00  <CIA-131>node: Ben Noordhuis master * r10ba95c / benchmark/url.js : bench: improve url parser benchmark - http://git.io/U-Au9w
23:24:37  <bnoordhuis>piscisaureus__: what is it?
23:24:55  <piscisaureus__>bnoordhuis: well, the asm version of cpuid wasn't actually used
23:25:05  <piscisaureus__>bnoordhuis: because of that google patch
23:25:17  <bnoordhuis>i know
23:25:29  <piscisaureus__>bnoordhuis: so instead we'd just assume that the processor has no capabilities :-)
23:25:39  <piscisaureus__>bnoordhuis: but this is trivial to fix
23:25:53  <bnoordhuis>yes - that's why i mentioned it yesterday :)
23:26:08  <piscisaureus__>I just did, actually
23:26:15  <piscisaureus__>will push
23:26:58  * stagasquit (Quit: ChatZilla 0.9.88-rdmsoft [XULRunner 1.9.0.17/2009122204])
23:27:40  <CIA-131>node: Bert Belder openssl-asm * rbaefd24 / (4 files in 3 dirs): openssl: use optimized asm code on x86 and x64 (+5 more commits...) - http://git.io/tOrpIQ
23:29:26  * saghulquit (Ping timeout: 268 seconds)
23:31:27  * saghuljoined
23:39:28  <piscisaureus__>^-- not correct yet...
23:41:14  <bentkus>does node.js not use optimized ssl stuff?
23:43:27  <bnoordhuis>bentkus: ask that question again in 30 minutes
23:44:21  <bentkus>only master will
23:44:44  <bentkus>but speed improvements are always awesome
23:44:48  <bnoordhuis>master is all that matters
23:45:00  <bnoordhuis>0.8 is so last tuesday
23:45:17  <bentkus>i have 0.9
23:46:51  <bentkus>and there is already 0.9.1
23:47:59  <bentkus>I wonder whats unstable about the 0.9 branch
23:48:05  <CIA-131>libuv: saghul master * r3f2479e / (4 files): unix: remove __read_mostly macro - http://git.io/QLIcgw
23:48:09  <bnoordhuis>saghul: ^
23:48:26  <bnoordhuis>hmm, wait
23:48:48  <CIA-131>libuv: Saúl Ibarra Corretgé master * rb7047d6 / (4 files): unix: remove __read_mostly macro - http://git.io/MGLZ9Q
23:49:46  * travis-cijoined
23:49:46  <travis-ci>[travis-ci] joyent/libuv#678 (master - 3f2479e : saghul): The build passed.
23:49:46  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/f8e7513a0618...3f2479ebee05
23:49:46  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/2427767
23:49:46  * travis-cipart
23:50:29  * travis-cijoined
23:50:29  <travis-ci>[travis-ci] joyent/libuv#679 (master - b7047d6 : Saúl Ibarra Corretgé): The build passed.
23:50:29  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/3f2479ebee05...b7047d6701db
23:50:29  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/2427771
23:50:29  * travis-cipart
23:52:11  <CIA-131>node: Bert Belder openssl-asm * r5c79a82 / (4 files in 3 dirs): openssl: use optimized asm code on x86 and x64 - http://git.io/HwPfbw
23:52:40  <piscisaureus__>^-- ok, that's correct. Unfortunately AES becomes 40% slower on my machine (using x64)
23:52:54  <piscisaureus__>so let's not land that yet
23:53:14  <bentkus>aes is old school
23:53:18  <bnoordhuis>wut
23:53:20  <bnoordhuis>how come?
23:53:29  <piscisaureus__>if only I knew
23:56:19  <bentkus>what's unstable in 0.9?
23:56:41  <bnoordhuis>bentkus: it's where new features get landed
23:57:04  <bentkus>like the the 40% aes decrease?
23:57:11  <bnoordhuis>for example
23:57:36  <bentkus>can't wait to try that one out
23:57:52  <piscisaureus__>bentkus: this asm patch is targeted for 0.8