00:09:58  * AvianFluquit (Ping timeout: 244 seconds)
00:14:16  * AvianFlujoined
00:26:00  <isaacs>piscisaureus_: i think i'm seeing this file locking thing.
00:26:03  <isaacs>this is really strange.
00:26:38  <piscisaureus_>isaacs: good, it's nice that you can actually reproduce it
00:26:47  <isaacs>piscisaureus_: i'm getting access denied for a long time after node is exited, though
00:26:55  <isaacs>i can't even delete the file in explorer
00:26:59  <isaacs>this is super odd
00:27:01  <piscisaureus_>isaacs: huh
00:27:10  <piscisaureus_>isaacs: do you have zombie processes left?
00:27:20  <piscisaureus_>(try `tasklist | grep node`)
00:27:50  <isaacs>noppe
00:27:51  <isaacs>nope
00:28:13  <piscisaureus_>isaacs: if you have file locking issues after node exited is can't really be related
00:28:42  <piscisaureus_>isaacs: but if you're interested, download "process explorer"
00:28:57  <isaacs>oh, wth, now i can delete them...
00:29:00  <isaacs>this is so weird
00:29:04  <isaacs>windows is like the moon, man.
00:29:36  <isaacs>hmm…. it looks like it's failing on utimes
00:29:37  <piscisaureus_>isaacs: file locking is really annoying, especially these cases where you are like "wtf why is this file locked"
00:29:49  <isaacs>maybe i should just skip that on windows?
00:30:05  <piscisaureus_>isaacs: it's probably a bug... does the utimes target exist?
00:30:11  <isaacs>yeah
00:31:15  <isaacs>yeah, so, the call to fs.utimes is raising an EACCES
00:31:21  <isaacs>and then i can't delete the file for a few minutes
00:31:27  <piscisaureus_>hmm
00:33:16  <piscisaureus_>huh
00:35:08  <isaacs>ok, so, skipping utimes makes it break in a new way, which is clearly just a bug somewhere.
00:35:27  <piscisaureus_>isaacs: utimes just works fine here
00:35:41  <isaacs>yeah, i can do it just fine one at a time
00:35:56  <isaacs>but for some reason, laying out a few hundred files and utimesing all of them ends up raising an eacces
00:35:59  <isaacs>??
00:39:04  <piscisaureus_>isaacs: hmm weird
00:39:18  <isaacs>it didn't on my other laptop
00:39:26  <isaacs>which had the same windows vm, but a different version of vmware.
00:39:33  <isaacs>so it may well be a vmware bug, i duno
00:39:59  <piscisaureus_>isaacs: are you installing to a virtual file system or to a mapped folder / drive?
00:40:09  <isaacs>piscisaureus_: no, i am installing *from* one, though
00:40:22  <isaacs>but it is just doing the cache add portion, which only writes to the home dir
00:40:38  <isaacs>utimes isn't importnat.
00:40:40  <piscisaureus_>isaacs: it would be interesting if you could log all the file operations and see where it breaks
00:40:53  <piscisaureus_>so I actually have an idea what npm is doing :-)
00:41:07  <isaacs>i'm adding // XXX comments where i put in kludges for windows :)
00:41:44  <piscisaureus_>isaacs: isn't it possible to just hack fs and log all the file ops?
00:41:49  <isaacs>but i do sometimes add a debug output in graceful-fs and log it all, yes
00:41:52  <isaacs>it's just super noisy
00:42:05  <isaacs>i wrap fs for my own reasons anyway
00:42:12  <piscisaureus_>isaacs: just do it once, and gist the output
00:42:20  <isaacs>mostly so i can just throw a zillion open operations at it, and never get emfiled
00:42:35  <isaacs>sure, just a second
00:42:53  <piscisaureus_>k
00:46:56  <isaacs>how do you direct stderr to a file in windows?
00:53:19  <piscisaureus_>isaacs: node foo.js 2>out.txt
00:53:25  <isaacs>yeah, that's what i thought
00:53:57  <isaacs>this is weird though… when i do that, it fails to load stuff, i think because it's using the fs module to write its writes
00:54:03  * AvianFluquit (Quit: Leaving)
00:55:08  <piscisaureus_>isaacs: yeah.
00:55:16  <piscisaureus_>isaacs: if you have mingw, pipe through cat
00:55:43  <piscisaureus_>isaacs: node foo.js 2>&1 | cat > a.txt
00:56:39  <piscisaureus_>(works in cmd as well(
00:57:59  <isaacs>weird, that's failing too.
00:58:13  <isaacs>oh well. graceful-fs isn't really intended to be DTrace
00:58:17  <isaacs>:)
00:59:09  <piscisaureus_>ok
00:59:47  * AvianFlujoined
01:00:51  <bnoordhuis>ffs, why has apple fucked up large file support so bad?
01:01:11  * bnoordhuisis debugging #2061 again
01:01:44  <piscisaureus_>isaacs: node -e "require('child_process').spawn('node', ['foo.js']).stderr.pipe(requ
01:01:44  <piscisaureus_>ire('fs').createWriteStream('out.txt'))"
01:01:56  <piscisaureus_>isaacs: if that doesn't work, I'm lost :-)
01:02:29  <isaacs>piscisaureus_: we can try this a bit later. i've gotta get ready for class.
01:02:34  <piscisaureus_>yeah
01:02:41  <isaacs>(ie, hurriedly do my homework right before class like i'm still in 8th grade)
01:02:46  <piscisaureus_>flex your fingers
01:03:03  <isaacs>but, the good news is, it's working on linux and win, with a few bugs, but mostly there.
01:03:06  <isaacs>thursday looks good.
01:03:26  <piscisaureus_>nice
01:03:29  <piscisaureus_>isaacs++
01:03:34  <piscisaureus_>it has to be said
01:04:03  <isaacs>nevermind, not working on windows.
01:04:06  <isaacs>but this is a different problem.
01:04:08  <isaacs>:)
01:04:13  <isaacs>i spoke too soon. cache adding works.
01:04:17  <isaacs>installing is pretty easy compared to that
01:04:28  <isaacs>it's just failing to remove its temp folder after installing
01:04:53  <piscisaureus_>isaacs: *shrug* not the worst that could happen
01:05:11  <piscisaureus_>isaacs: probably really a file locking issue. You can't delete a directory if files in it are still open
01:05:35  <isaacs>yeah
01:05:44  <isaacs>i'm sure it's something dumb. it always is.
01:05:55  <isaacs>(that is, "something where it's not working because i'm being dumb")
01:06:12  <piscisaureus_>isaacs: it could be something dumb on "our" side too
01:06:17  <piscisaureus_>that is, the libuv side
01:06:21  <isaacs>sure
01:06:40  <isaacs>i'll try it with other versions and stuff
01:06:45  <isaacs>ok, tomorrow
01:06:48  <piscisaureus_>yeah. later
01:06:51  <piscisaureus_>good luck with class
01:15:25  <CIA-95>node: Ingmar Runge v0.6 * r2603832 / (src/node_crypto.cc test/simple/test-crypto-ecb.js):
01:15:25  <CIA-95>node: node_crypto: use EVP_Cipher*_ex methods, not 'obsolete' versions
01:15:25  <CIA-95>node: This also fixes an issue that made blowfish's ECB mode unusable. - http://git.io/s1ckkg
01:15:25  <CIA-95>node: Russ Bradberry v0.6 * r754e23d / src/node.cc : fixed problem with rss memory reporting incorrect number over 4 gigs - http://git.io/RAeBoQ
01:17:31  * AvianFlu_joined
01:20:14  * AvianFluquit (Ping timeout: 244 seconds)
01:23:25  * AvianFlu_changed nick to AvianFlu
01:26:30  * mikealquit (Quit: Leaving.)
01:31:05  * ericktquit (Ping timeout: 244 seconds)
01:31:19  <CIA-95>node: Ben Noordhuis master * rbbd976b / (common.gypi deps/uv/uv.gyp node.gyp):
01:31:19  <CIA-95>node: build: always compile with large file support
01:31:19  <CIA-95>node: Fixes a nasty sizeof(struct stat) mismatch bug that caused random crashes and
01:31:19  <CIA-95>node: silent memory corruption. - http://git.io/YA3ERw
01:41:06  * AvianFlu_joined
01:43:48  * mikealjoined
01:44:28  * AvianFluquit (Ping timeout: 260 seconds)
01:46:52  * dapquit (Quit: Leaving.)
02:05:31  * AvianFlu_changed nick to AvianFlu
02:08:07  * pieternquit (Quit: pietern)
02:09:50  * AvianFlu_joined
02:10:52  * AvianFluquit (Ping timeout: 244 seconds)
02:12:14  * AvianFlujoined
02:13:55  * jmp0joined
02:16:02  * AvianFlu_quit (Ping timeout: 260 seconds)
02:17:19  * AvianFluquit (Quit: Leaving)
02:17:28  * mikealquit (Quit: Leaving.)
02:17:46  * benviejoined
02:24:08  * isaacsquit (Quit: isaacs)
02:25:00  * mikealjoined
02:34:55  * piscisaureus_quit (Quit: ~ Trillian Astra - www.trillian.im ~)
02:43:52  * ericktjoined
03:05:07  * indexzerojoined
03:07:58  * brsonquit (Quit: leaving)
03:17:49  * mikealquit (Quit: Leaving.)
03:31:03  * bnoordhuisquit (Ping timeout: 240 seconds)
05:43:18  * mikealjoined
05:48:41  * mikealquit (Quit: Leaving.)
05:48:50  * mikealjoined
05:55:17  * felixgejoined
06:20:00  * ericktquit (Quit: erickt)
06:23:46  * mikealquit (Quit: Leaving.)
06:45:38  * mikealjoined
06:55:05  * isaacsjoined
07:12:26  * mikealquit (Quit: Leaving.)
07:19:39  <sh1mmer>ow
07:19:43  <sh1mmer>0.7 hurts me
07:30:00  * mikealjoined
07:34:03  * mralephjoined
07:35:01  * sh1mmerquit (Quit: sh1mmer)
08:35:53  * kuebkjoined
09:14:30  * eddybjoined
09:14:32  <eddyb>yay
09:14:37  <eddyb>managed to fix my build system
09:14:43  <eddyb>Makefile > gyp
09:14:57  <eddyb>for anything else than gyp was created for
09:15:21  <eddyb>[00:19|% 100|+ 173|- 0]: Done
09:15:25  <eddyb>and no test errors
09:15:31  <eddyb>are they all silently failing?
09:17:55  <eddyb>test-all gives errors, test-uv doesn't
09:22:10  * mralephquit (Quit: Leaving.)
09:23:11  <eddyb>and I lost mraleph
09:40:29  * isaacsquit (Quit: isaacs)
10:10:12  * piscisaureus_joined
11:09:02  * felixgequit (Quit: felixge)
11:55:52  * piscisaureus_quit (Ping timeout: 255 seconds)
11:58:26  * felixgejoined
12:06:49  * piscisaureus_joined
12:46:35  * bnoordhuisjoined
13:16:49  <CIA-95>node: Nathan Rajlich v0.6 * rb204006 / (lib/util.js test/simple/test-util-inspect.js):
13:16:49  <CIA-95>node: util: ensure that the .inspect function isn't the one being executed
13:16:49  <CIA-95>node: Fixes #2225. - http://git.io/eiThZA
13:27:25  <piscisaureus_>bnoordhuis: about uv_chdir
13:27:28  <piscisaureus_>er
13:27:34  <piscisaureus_>uv_getcwd
13:27:42  <piscisaureus_>who should alloc the memory ... libuv or the user
13:27:56  <piscisaureus_>I think on windows the path can be very long (like 32k characters)
13:28:28  <CIA-95>node: Ben Noordhuis master * rf44d0b9 / (src/node_crypto.cc test/simple/test-crypto.js):
13:28:29  <CIA-95>node: crypto: throw exception on unknown digest method
13:28:29  <CIA-95>node: Fixes #2227. - http://git.io/nGSTvA
13:28:34  <bnoordhuis>piscisaureus_: the user should pass in a buffer
13:28:44  <bnoordhuis>maybe we can supply a hint about the preferred buffer size
13:29:08  <piscisaureus_>bnoordhuis: then on windows we would hint to send a 64k buffer
13:29:19  <piscisaureus_>otherwise it could - in theory - fail
13:30:36  <bnoordhuis>piscisaureus_: so be it
13:30:54  <piscisaureus_>bnoordhuis: maybe we should have UV_MAX_PATH :-)
13:31:23  <bnoordhuis>piscisaureus_: come to think of it... it's theoretically possible for a unix path to be megabytes long
13:31:35  <bnoordhuis>i mean, there is no mandated upper limit
13:31:46  <piscisaureus_>bnoordhuis: yeah ok but no unices actually do that
13:32:00  <piscisaureus_>but on windows it *is* actually possible to have a path that long
13:32:02  <bnoordhuis>not any i care about anyway
13:32:30  <piscisaureus_>(although I just ccreated a path that long, then I was unable to remove it with explorer and cmd)
13:32:36  <bnoordhuis>heh
13:36:56  <CIA-95>libuv: Bruce Mitchener v0.6 * rd513d9b / (22 files in 4 dirs): Fix typos. - http://git.io/ovyzFA
13:38:15  <bnoordhuis>^ notice that none of those typos were in my code >:)
13:44:55  <piscisaureus_>bnoordhuis: I think our american friends made just as much typos as I did
13:45:45  <bnoordhuis>piscisaureus_: shameful, isn't it?
13:45:56  <piscisaureus_>well
13:46:00  <piscisaureus_>who cares
13:46:09  <bnoordhuis>bruce mitchener does
13:46:41  <piscisaureus_>as long as the comments are understandable it is not much of an issue
13:49:06  <CIA-95>libuv: Bruce Mitchener v0.6 * rb684deb / include/uv.h : Note that uv_is_active works on timers. - http://git.io/SJJZZg
13:50:47  <piscisaureus_>bnoordhuis: https://github.com/runner-mei/libuv/commit/f56d9e0699b477a74a58430c8b8551ca66eab28b
13:50:52  <piscisaureus_>what do you think of it?
13:50:59  <piscisaureus_>there's some style issues
13:51:49  <bnoordhuis>yes, function naming
13:51:59  <bnoordhuis>c++ / c99 comments
13:52:08  <piscisaureus_>bnoordhuis: but, conceptually?
13:52:40  <piscisaureus_>bnoordhuis: does it matter that he reallocs an array with loop pointers
13:52:59  <piscisaureus_>It's now what I would do (I have an irrational love for linked lists) but I think it's fine, right?
13:53:31  <piscisaureus_>s/now/not/
13:55:13  <bnoordhuis>piscisaureus_: well... i'd rather see a ngx_queue_t there
13:55:22  <bnoordhuis>and he needs to handle malloc failure
13:57:18  <piscisaureus_>agreed
13:57:44  <piscisaureus_>although I think this is fine for a while (I mean, how many loops are we going to have anyway?)
13:58:01  <bnoordhuis>+ signal_data_t* data = (signal_data_t*)((uv_async_t*)handle)->data;
13:58:01  <bnoordhuis>+ free(data);
13:58:14  <bnoordhuis>^ this is funky - he shouldn't touch handle->data
13:59:13  <piscisaureus_>bnoordhuis: it doesn't matter - it's an *internal* watcher
13:59:19  <piscisaureus_>that is not exposed to the user
14:00:28  <bnoordhuis>someone is going to overlook that handle->data is used and you'll get all kinds of hard to track down bugs
14:02:02  <piscisaureus_>bnoordhuis: he is also initializing the async watcher from the signal handler
14:02:05  <piscisaureus_>that's not right
14:03:48  <bnoordhuis>re handle->data, he should embed the struct and use container_of
14:03:55  <piscisaureus_>yrah
14:04:04  <piscisaureus_>yarrr
14:04:41  <bnoordhuis>yo ho and a bottle of rum!
14:05:55  <bnoordhuis>okay, edje de gekste expects me at 5 near the openbare bibliotheek
14:06:18  <bnoordhuis>apparently there's an italian restaurant there with a good view
14:07:16  <bnoordhuis>piscisaureus_: ^
14:07:34  <piscisaureus_>bnoordhuis: you mean, vapiano's?
14:07:45  <bnoordhuis>i think that's what he called it yes
14:08:03  <piscisaureus_>bnoordhuis: I don't think I will be that early
14:08:16  <piscisaureus_>I haz work to do
14:08:33  <bnoordhuis>oh, is it that day of the week again?
14:08:50  <bnoordhuis>you can join later if you want
14:09:12  <piscisaureus_>i think i will
14:17:13  <piscisaureus_>bnoordhuis: what's eddy's last name?
14:17:31  <bnoordhuis>piscisaureus_: dunno
14:18:06  <piscisaureus_>bnoordhuis: edje?
14:18:12  <piscisaureus_>bnoordhuis: wwaar ken je hem van?
14:18:29  <bnoordhuis>piscisaureus_: oh edje
14:18:33  <bnoordhuis>ex-nrc collega
14:18:42  <bnoordhuis>of collega, hij was een freelancer
14:19:39  <piscisaureus_>dus, ed _wie_?
14:19:49  <bnoordhuis>piscisaureus_: edwin duinkerken
14:24:45  <bnoordhuis>piscisaureus_: https://github.com/joyent/node/pull/2231/files#r260880 <- sensible?
14:26:45  <piscisaureus_>bnoordhuis: is dat Edwin Duinkerken uit utrect,?
14:26:49  <piscisaureus_>web dev?
14:26:54  <bnoordhuis>piscisaureus_: the one and only
14:27:22  <piscisaureus_>ah
14:27:25  <piscisaureus_>die ken ik niet
14:27:39  <bnoordhuis>straks wel
14:27:49  <piscisaureus_>muhahaha
14:29:11  <piscisaureus_>bnoordhuis: punicode upgrade - fine
14:29:35  <bnoordhuis>i was mostly concerned with the html changes
14:29:41  <piscisaureus_>but the html change has a semicolon smell to it
14:29:49  <bnoordhuis>but it's fine according to the people in #node.js
14:30:35  <bnoordhuis>hmm, there's a lot of discussion about it now actually
14:30:37  <bnoordhuis>okay, won't merge
14:30:48  <piscisaureus_>bnoordhuis: fine go ahead
14:31:05  <piscisaureus_>bnoordhuis: but this doesn't get you an AUTHORS entry :-)
14:31:39  <piscisaureus_>it's a super pointless change :-)
14:32:09  <bnoordhuis>yeah - but maybe i'll leave it in anyway, saves me having to rebase out the commit
14:41:43  <CIA-95>node: Mathias Bynens v0.6 * rcf89bee / lib/punycode.js : punycode: Update to v0.2.1 - http://git.io/mwP4HA
14:41:44  <CIA-95>node: Mathias Bynens v0.6 * r456dafc / doc/template.html : (log message trimmed)
14:41:45  <CIA-95>node: Docs template: HTML cleanup
14:41:45  <CIA-95>node: Remove references to `/favicon.ico` as it’s the default in all browsers. More
14:41:45  <CIA-95>node: info: http://mathiasbynens.be/notes/rel-shortcut-icon
14:41:45  <CIA-95>node: `type=text/css` and `media=all` are the implied defaults for `<link
14:41:46  <CIA-95>node: rel=stylesheet>` (and `<style>`), so omit these attributes.
14:41:46  <CIA-95>node: Similarly, `type=text/javascript` is the default for `<script>` elements, so
14:52:22  <CIA-95>node: Ben Noordhuis v0.6 * r3ca0517 / wscript :
14:52:22  <CIA-95>node: build: add --without-npm flag to configure script
14:52:22  <CIA-95>node: Fixes #2200. - http://git.io/KeSVbQ
14:54:44  * ericktjoined
15:01:16  * bnoordhuisquit (Ping timeout: 255 seconds)
16:17:13  * kuebkpart
16:43:34  * dapjoined
16:54:35  * indexzeroquit (Quit: indexzero)
16:57:59  * isaacsjoined
17:01:16  * sh1mmerjoined
17:13:01  * indexzerojoined
17:45:53  * piscisaureus__joined
17:48:21  * piscisaureus_quit (Ping timeout: 244 seconds)
17:51:53  * piscisaureus__quit (Quit: ~ Trillian Astra - www.trillian.im ~)
17:53:27  * piscisaureus_joined
17:56:07  * brsonjoined
18:04:17  * piscisaureus_quit (Ping timeout: 244 seconds)
18:24:22  * mikealquit (Quit: Leaving.)
18:28:13  * mikealjoined
18:30:27  * mikealquit (Client Quit)
18:30:40  * mikealjoined
18:31:09  * pieternjoined
18:56:32  * felixgequit (Quit: felixge)
19:10:25  * creationixjoined
19:11:11  * AvianFlujoined
19:12:56  * indexzeroquit (Quit: indexzero)
19:26:23  * AvianFlu_joined
19:27:29  * AvianFluquit (Disconnected by services)
19:27:33  * AvianFlu_changed nick to AvianFlu
19:27:56  * mikealquit (Quit: Leaving.)
19:33:35  * ericktquit (Quit: erickt)
20:04:13  * AvianFlu_joined
20:06:34  * AvianFluquit (Ping timeout: 244 seconds)
20:13:12  * AvianFlujoined
20:14:19  * AvianFlu_quit (Ping timeout: 244 seconds)
20:38:13  <isaacs>ok, npm is working again on windows, but this whole "installing into %ProgFiles%" thing is way more obnoxious than i originally thought.
20:38:19  <isaacs>i don't know how anyone is not freaking out more about this.
20:38:45  <creationix>everything in windows is obnoxious
20:38:57  <isaacs>i dunno about *everything*
20:39:02  <isaacs>the startup chime is quite nice.
20:39:15  <isaacs>and there's that grassy field. i always think that would be a lovely place to take a nap.
20:39:21  <eddyb>hi guys
20:39:41  <eddyb>isaacs: IT'S A LIE
20:39:53  <eddyb>you can have the chime and the background on linux, too
20:47:26  * ericktjoined
20:54:56  <creationix>eddyb: I agree with Isaacs, there is something almost intoxicating about the green field that only true windows can pull off
20:55:21  <eddyb>waaait
20:55:26  <eddyb>is the the XP green field?
20:55:37  <eddyb>and I thought he had some sense in him...
20:55:47  <eddyb>my grub screen looks better than that
20:56:16  <creationix>of course, I haven't used windows since XP, so it may be even better now
20:56:39  <eddyb>creationix: it's gone
20:57:02  <eddyb>win7 has some cool landscapes instead
21:05:33  * creationixpart
21:09:58  <CoverSlide>gnome has a nice close-up pic of grass that i like, even when I'm not using gnome
21:18:45  * eddybquit (Quit: Konversation terminated!)
21:19:09  * felixgejoined
21:20:17  <indutny>isaacs: hi! yt?
21:20:20  <isaacs>hi
21:43:09  * felixgequit (Quit: felixge)
21:44:35  * ljacksonjoined
21:46:48  <ljackson>anyone have a porting guide for libev code into nodejs libuv ? looking for ev_io to I guess uv_handle_t or uv_stream_t ?
21:54:48  <indutny>isaacs: sorry, 4g internet is lagging. https://github.com/isaacs/npm/issues/1818
21:54:54  <indutny>npm is publishing private data
21:55:28  <isaacs>indutny: don't start npmignore lines wiht /
21:55:33  <isaacs>indutny: that is indeed a bug.
21:55:43  <indutny>isaacs: well, I did that now :)
21:55:53  <isaacs>indutny: but, i'm not going to start parsing the git index any time soon, i don't think. that's a lot of work and complexity
21:56:02  <isaacs>maybe a 2.0 thing, i dunno. but not soon
21:56:34  <indutny>isaacs: are pull requests welcome?
21:56:59  <isaacs>indutny: you can always request :) but i'm not very interested in supporting a git-aware feature.
21:58:48  <indutny>isaacs: we can just exec some git commands inside module directory
21:59:06  <indutny>or you can implement global hooks
21:59:31  <indutny>and I'll just wrote module for myself :P
21:59:37  * pieternquit (Ping timeout: 260 seconds)
22:03:44  * mralephjoined
22:10:45  * mikealjoined
22:57:29  * pieternjoined
23:28:14  * mralephquit (Quit: Leaving.)
23:34:13  * mralephjoined
23:38:08  * mikealquit (Quit: Leaving.)
23:50:11  * mralephquit (Quit: Leaving.)
23:54:48  * paddybyersquit (Quit: paddybyers)