00:01:14  * c4miloquit (Remote host closed the connection)
00:06:21  * mikealquit (Quit: Leaving.)
00:17:14  * mikealjoined
00:28:30  <deoxxa>i seem to be getting a crash in uv_fs_read that originates inside libeio
00:29:04  <deoxxa>i'll paste a gdb thingy
00:30:08  <deoxxa>https://gist.github.com/2715167
00:31:50  * isaacsquit (Remote host closed the connection)
00:33:13  * irajoined
00:35:56  <deoxxa>ah, looks like i have a race condition elsewhere
00:54:01  * dapquit (Quit: Leaving.)
00:57:38  * elijah-mbpquit (Read error: Connection reset by peer)
01:18:23  * TooTallNatequit (Quit: Leaving...)
01:33:44  * bnoordhuisjoined
01:35:16  * c4milojoined
01:35:16  * c4miloquit (Remote host closed the connection)
01:37:23  * elijah-mbpjoined
01:48:44  * abraxasjoined
01:49:03  * abraxasquit (Read error: Connection reset by peer)
01:49:31  * abraxasjoined
01:54:09  * iraquit (Quit: Leaving...)
01:57:00  * brsonquit (Ping timeout: 252 seconds)
01:58:26  * mikealquit (Quit: Leaving.)
02:01:29  <deoxxa>what does the "result" field in a uv_fs_t response from uv_fs_stat mean?
02:02:18  * mikealjoined
02:02:57  <deoxxa>it seems that when it's non-zero (but not -1?), the ptr field is garbage
02:03:13  <deoxxa>i.e. segfault upon attempted access
02:06:56  * mmaleckijoined
02:08:08  * mmaleckiquit (Read error: Connection reset by peer)
02:09:37  * mmaleckijoined
02:09:44  * mmaleckiquit (Client Quit)
02:09:52  * mmaleckijoined
02:24:57  * ericktquit (Quit: erickt)
02:31:02  <bnoordhuis>deoxxa: 0 usually means success except with uv_fs_readdir, in that case it's the number of entries
02:31:05  <bnoordhuis>and -1 means error
02:31:16  <deoxxa>and anything else means...
02:31:38  <bnoordhuis>oh, and readlink and friends set it to the buffer's length
02:32:10  <deoxxa>i'm getting 135329328. looks like bad data to me...
02:32:18  <deoxxa>i thought it might have been some kind of bitset
02:32:30  <bnoordhuis>the ptr field is only relevant for some fs calls, like stat and lstat
02:33:03  <deoxxa>yep, this is uv_fs_fstat
02:33:22  <deoxxa>"result" was 135329328 (or another large number), "ptr" was some random memory address
02:33:48  <mmalecki>bnoordhuis: jeez, do you ever sleep?
02:34:32  <bnoordhuis>mmalecki: sleep is for the weak
02:35:48  <deoxxa>uv_fs_open sets result to the fd, yes?
02:36:05  <bnoordhuis>deoxxa: sanity check, do you compile with _LARGEFILE_SOURCE and _FILE_OFFSET_BITS=64?
02:36:37  <bnoordhuis>struct stat is kind of a wtf on unices, it changes size depending on those defines
02:37:18  <bnoordhuis>i don't want to know how many hours of debugging are the result of that stupid, stupid idea
02:37:36  <deoxxa>hmm
02:37:39  * deoxxachecks
02:37:59  <deoxxa>i'll have to finish this silly refactoring i'm doing to try to isolate the problem first though
02:39:24  * TooTallNatejoined
02:43:04  * TooTallNatequit (Client Quit)
02:46:47  * mikealquit (Quit: Leaving.)
02:50:35  * TooTallNatejoined
02:56:07  * isaacsjoined
02:56:54  <isaacs>bnoordhuis: Thoughts on https://github.com/joyent/node/pull/3271?
02:57:26  * mmaleckiquit (Read error: Connection reset by peer)
02:57:35  <bnoordhuis>isaacs: i have some doubts about the last commit
02:57:42  <isaacs>bnoordhuis: yeah, me too
02:58:01  <bnoordhuis>otherwise lgtm. haven't actually run it though
02:58:06  <isaacs>bnoordhuis: not even so much that it lies, as that it's actually slower in anything except benchmarks.
02:58:18  <isaacs>granted, measuring a single fs.readfile is tricky
02:58:26  <isaacs>but that array nonsense is work, and it's more than 0
02:58:57  <isaacs>but if you read the same file a thousand times, boy, it goes SOOOOO fast
02:59:22  <bnoordhuis>heh
02:59:40  <bnoordhuis>still, i would move to keep that one out for now
02:59:45  * philipsquit (Excess Flood)
03:02:49  * philipsjoined
03:03:28  <isaacs>k, i'm convinced.
03:03:42  <isaacs>i should write a "win at benchmarks" module.
03:04:01  <isaacs>it hacks fs and http to cache everything in memory, memoizes all buffer creations, etc.
03:04:42  <deoxxa>oh so you mean level the playing field
03:04:48  <isaacs>hahah
03:04:55  <isaacs>k, i'm gonna land the other two, then
03:05:02  <deoxxa>since that's what every other benchmark does to prove that node's "not that fast"
03:05:05  <isaacs>(the benchmark and the non-cheating faster fs.readFile)
03:05:28  <deoxxa>sorry, "prove"
03:05:53  <CIA-155>node: isaacs master * ree6c118 / benchmark/fs-readfile.js : benchmark for fs.readfile - http://git.io/MkkeZg
03:05:53  <CIA-155>node: isaacs master * r1a2255a / (lib/fs.js test/simple/test-fs-sync-fd-leak.js): Faster fs.readFile and fs.readFileSync - http://git.io/NbUH4Q
03:06:26  * elijah-mbpquit (Remote host closed the connection)
03:06:51  <isaacs>ok, i gotta go pack. g'nite.
03:06:59  <deoxxa>o/
03:07:09  * isaacsquit (Remote host closed the connection)
03:21:46  * mikealjoined
03:24:32  * perezdquit (Quit: perezd)
03:26:41  <deoxxa>huh. so after uv_fs_stat (not fstat), result == 0, but ptr == nil
03:26:46  <deoxxa>does this seem right..?
03:27:39  <deoxxa>oh my
03:27:52  <deoxxa>i had uv_fs_req_cleanup in the wrong place
03:27:56  <deoxxa>i'm so dumb sometimes
04:04:07  * mmaleckijoined
04:06:34  <CIA-155>node: Kevin Bowman master * rccc4e54 / doc/api/fs.markdown : doc: document fs 'rs+' open mode - http://git.io/loIm2A
04:07:49  * mmaleckiquit (Client Quit)
04:08:02  * mmaleckijoined
04:08:48  * mmaleckiquit (Client Quit)
04:14:28  * mmaleckijoined
04:16:41  * mmalecki_joined
04:16:51  * mmaleckiquit (Client Quit)
04:19:47  <CIA-155>node: Ben Noordhuis master * r18b94ea / doc/api/fs.markdown : doc: improve fs.open() docs - http://git.io/GI0RSQ
04:21:08  * mmalecki_quit (Client Quit)
04:21:16  * mmaleckijoined
04:35:22  * ericktjoined
04:35:29  * mmaleckiquit (Read error: Connection reset by peer)
05:12:38  <CIA-155>libuv: Ben Noordhuis master * r9efa8b3 / (60 files in 8 dirs): unix, windows: rework reference counting scheme - http://git.io/mxGiRQ
05:12:46  <bnoordhuis>^ there it is, people
05:13:41  <ggreer>m-m-m-m-megadiff
05:14:47  * travis-cijoined
05:14:47  <travis-ci>[travis-ci] joyent/libuv#287 (master - 9efa8b3 : Ben Noordhuis): The build is still failing.
05:14:47  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/07622e7...9efa8b3
05:14:47  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/1352917
05:14:47  * travis-cipart
05:16:41  * ericktquit (Quit: erickt)
05:34:24  * mikealquit (Quit: Leaving.)
05:35:53  * mikealjoined
05:41:01  * mikealquit (Quit: Leaving.)
05:55:24  * mikealjoined
06:01:00  * paddybyersjoined
06:02:57  * mikealquit (Quit: Leaving.)
06:27:28  * xaqjoined
06:30:29  * xaqquit (Remote host closed the connection)
06:31:54  * xaqjoined
06:36:53  * elijah-mbpjoined
07:35:13  <saghul>bnoordhuis ++ :-)
07:37:25  * xaq_joined
07:38:09  * xaqquit (Ping timeout: 244 seconds)
07:41:24  * mmaleckijoined
07:54:35  * xaq_quit (Remote host closed the connection)
08:06:43  * rendarjoined
08:10:34  * paddybyers_joined
08:12:21  * paddybyersquit (Read error: Operation timed out)
08:12:22  * paddybyers_changed nick to paddybyers
08:23:14  * felixgejoined
08:23:14  * felixgequit (Changing host)
08:23:14  * felixgejoined
08:28:54  * TooTallNatequit (Quit: Leaving...)
08:34:39  * ericktjoined
08:55:50  * ericktquit (Quit: erickt)
09:20:19  * felixgequit (Quit: felixge)
10:02:17  * mmaleckiquit (Ping timeout: 240 seconds)
10:02:34  * mmaleckijoined
10:15:17  * bnoordhuisquit (Ping timeout: 240 seconds)
10:38:28  * AndreasMadsenjoined
10:39:03  * AndreasMadsenquit (Client Quit)
10:39:08  * AndreasMadsenjoined
10:39:11  * AndreasMadsenquit (Client Quit)
10:46:02  * TheJHjoined
10:50:54  * avsejpart ("Ex-Chat")
11:03:40  * mmaleckiquit (Ping timeout: 272 seconds)
11:12:17  * mmaleckijoined
11:41:48  * irajoined
12:29:25  * abraxasquit (Read error: Connection reset by peer)
12:39:18  * iraquit (Quit: Textual IRC Client: http://www.textualapp.com/)
12:44:09  * mmaleckiquit (Quit: leaving)
12:50:01  * irajoined
13:02:16  * c4milojoined
13:37:28  * mmaleckijoined
13:54:28  * ericktjoined
14:09:25  * mattstevensjoined
14:27:00  * mmaleckiquit (Quit: Lost terminal)
15:00:12  * toothrquit (Ping timeout: 252 seconds)
15:04:02  * toothrjoined
15:11:25  <creationix>bnoordhuis: refcount refactor! :)
15:16:57  * elijah-mbpquit (Remote host closed the connection)
15:32:03  * kohaijoined
15:35:28  <saghul>pyuv is working just fine after I adapted it due to the changes because of the refcount refractor, great work guys! :-)
15:37:19  * creationixcan't wait to update luvit, candor.io, and luvmonkey
15:57:51  * ericktquit (Quit: erickt)
16:02:35  * iraquit (Quit: Textual IRC Client: http://www.textualapp.com/)
16:06:20  * dapjoined
16:11:36  * ericktjoined
16:22:54  * mikealjoined
16:28:47  * ericktquit (Quit: erickt)
16:33:08  * felixgejoined
16:33:08  * felixgequit (Changing host)
16:33:09  * felixgejoined
16:34:03  * igorzi_joined
16:34:47  * benviequit
16:38:29  * benviejoined
16:50:25  * mikealquit (Quit: Leaving.)
16:51:27  * elijah-mbpjoined
16:57:18  * mikealjoined
16:57:28  <CIA-155>libuv: Igor Zinkovsky master * r60af28a / (5 files in 4 dirs): windows: support junctions with uv_fs_synlink - http://git.io/zxUY4Q
16:59:33  * travis-cijoined
16:59:33  <travis-ci>[travis-ci] joyent/libuv#288 (master - 60af28a : Igor Zinkovsky): The build is still failing.
16:59:33  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/9efa8b3...60af28a
16:59:33  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/1357814
16:59:33  * travis-cipart
17:05:13  * brsonjoined
17:13:26  * mikealquit (Quit: Leaving.)
17:17:57  * mikealjoined
17:22:14  * AndreasMadsenjoined
17:34:34  * isaacsjoined
17:35:55  * mikealquit (Quit: Leaving.)
17:38:59  * isaacsquit (Read error: Connection reset by peer)
17:49:13  * mikealjoined
18:05:10  * mjr_joined
18:10:43  * mikealquit (*.net *.split)
18:10:44  * ljacksonquit (*.net *.split)
18:11:18  * ljacksonjoined
18:11:19  * mikealjoined
18:12:24  * igorzi__joined
18:14:00  * igorzi_quit (Ping timeout: 245 seconds)
18:26:43  * `3rdEdenjoined
18:45:38  * ericktjoined
18:45:49  * mattstevensquit (Quit: mattstevens)
18:51:36  * ljacksonquit (*.net *.split)
18:51:36  * mikealquit (*.net *.split)
18:52:21  * ljacksonjoined
18:52:21  * mikealjoined
18:59:33  * mikealquit (Quit: Leaving.)
19:02:08  * mikealjoined
19:02:11  * irajoined
19:06:01  * ericktquit (Read error: Connection reset by peer)
19:09:42  * ericktjoined
19:19:28  * ericktquit (Remote host closed the connection)
19:19:50  * ericktjoined
19:26:16  * toothrquit (Ping timeout: 240 seconds)
19:27:09  * mikealquit (Quit: Leaving.)
19:27:31  * toothrjoined
19:32:03  * erickt_joined
19:36:10  * ericktquit (Ping timeout: 252 seconds)
19:36:11  * erickt_changed nick to erickt
19:47:33  * mikealjoined
19:54:21  * pieternjoined
19:56:03  * xaqjoined
20:09:19  * `3rdEdenquit (Quit: Linkinus - http://linkinus.com)
20:12:01  * c4miloquit (Remote host closed the connection)
20:34:18  * elijah-mbpquit (Ping timeout: 272 seconds)
20:44:08  * brsonquit (Ping timeout: 252 seconds)
20:51:46  * elijah-mbpjoined
21:01:09  * bnoordhuisjoined
21:02:19  * brsonjoined
21:12:16  * mmaleckijoined
21:22:51  * loladirojoined
21:29:21  * AndreasMadsenquit (Remote host closed the connection)
21:30:09  * loladiroquit (Remote host closed the connection)
21:32:17  * loladirojoined
21:38:42  * mmaleckiquit (Quit: Reconnecting)
21:40:31  * mmaleckijoined
21:41:11  * felixgequit (Quit: felixge)
21:47:54  * mmaleckiquit (Read error: Connection reset by peer)
21:48:36  * rendarquit
21:54:32  * c4milojoined
22:40:14  <mjr_>I just noticed that console.log() is blocking the event loop. When did that change?
22:43:29  <AvianFlu>0.5 sometime
22:44:04  <mjr_>Huh. I wonder why I never noticed it until now.
22:44:25  <mjr_>Is there a way to get the old buffering behavior if you know what you are doing?
22:44:35  * mikealquit (Quit: Leaving.)
22:51:48  <AvianFlu>dunno, but I think even process.send blocks because of it
22:51:58  * paddybyersquit (Quit: paddybyers)
22:52:29  <mjr_>This is kind of bad for us because it means that logging back pressure can block the event loop
23:04:37  * mikealjoined
23:05:30  <pquerna>stderr blocks, stdout shouldn't?
23:05:46  <pquerna>or maybe that was nuked in 0.5. bleh
23:06:21  <pquerna>They are blocking in the case that they refer to regular files or TTY file descriptors. In the case they refer to pipes, they are non-blocking like other streams.
23:16:47  * elijah-mbpquit (Remote host closed the connection)
23:21:05  <CIA-155>libuv: Ben Noordhuis master * r073a48d / (include/uv-private/uv-unix.h src/unix/core.c src/unix/loop.c): unix: back out new idle watcher for now - http://git.io/7DrqSA
23:21:05  <CIA-155>libuv: Ben Noordhuis master * r538a562 / (test/test-list.h uv.gyp test/test-tcp-connect-timeout.c): test: add tcp connect timeout test - http://git.io/LyJeTw
23:21:06  <CIA-155>libuv: Ben Noordhuis master * r7447048 / test/test-list.h : test: enable tcp_connect_error_fault test - http://git.io/pybZMw
23:21:06  <CIA-155>libuv: Ben Noordhuis master * re10a87a / src/unix/stream.c : unix: streamline shutdown req cleanup logic - http://git.io/WJ_wlA
23:21:06  <CIA-155>libuv: Ben Noordhuis master * re82d464 / src/unix/stream.c : unix: fix close-before-connect bug - http://git.io/lBjDrg
23:21:08  <mjr_>pquerna: I just tested, and it does seem like console.log will block the event loop
23:21:23  <bnoordhuis>yes, it's sync
23:21:30  <bnoordhuis>i guess pquerna already explained it
23:23:15  * travis-cijoined
23:23:16  <travis-ci>[travis-ci] joyent/libuv#289 (master - e82d464 : Ben Noordhuis): The build is still failing.
23:23:16  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/60af28a...e82d464
23:23:16  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/1361519
23:23:16  * travis-cipart
23:23:26  <igorzi__>bnoordhuis: when are you planning to integrate the uv refcount rework into node?
23:23:49  <bnoordhuis>igorzi__: working on it. there are 5 or 6 tests failing
23:24:03  <bnoordhuis>nothing serious, probably me just not properly updated the relevant code in node
23:24:07  <bnoordhuis>*updating
23:25:47  * AvianFluquit (Quit: Leaving)
23:26:33  <igorzi__>bnoordhuis: k, thx
23:27:28  <igorzi__>bnoordhuis: also, do have the list of failing tests on unix handy? (prior to refcount rework)
23:27:40  <bnoordhuis>igorzi__: sneak preview in case you're interested: https://github.com/bnoordhuis/node/compare/rr
23:27:46  <bnoordhuis>igorzi__: in node or libuv?
23:27:51  <igorzi__>bnoordhuis: node
23:28:02  <bnoordhuis>with node, everything was passing
23:28:09  <igorzi__>bnoordhuis: there are many windows tests failing (25+). i want to see how many of them are windows-specific
23:28:14  <igorzi__>bnoordhuis: ok
23:47:14  * loladiroquit (Quit: loladiro)
23:48:04  * loladirojoined
23:56:41  * elijah-mbpjoined