00:01:59  <bnoordhuis>cool
00:09:30  * brsonquit (Ping timeout: 244 seconds)
00:09:43  * brsonjoined
00:10:40  <bnoordhuis>piscisaureus: https://github.com/bnoordhuis/libuv/commit/cf9c9a7
00:13:50  <piscisaureus>bnoordhuis: commented
00:13:56  <piscisaureus>bnoordhuis: I had 1 comment
00:14:29  <bnoordhuis>"You should probably for invocations of close_cb." <- wut?
00:14:47  <piscisaureus>bnoordhuis: grep close_cb -r src/win
00:15:15  <bnoordhuis>so what does an endgame entail?
00:15:44  <piscisaureus>bnoordhuis: close, or an evaluation if close is possible, or shutdown, or flush
00:16:33  <bnoordhuis>okay... so why is there no common exit function? i mean something like uv__finish_close
00:16:49  <piscisaureus>bnoordhuis: e.g. https://github.com/bnoordhuis/libuv/blob/master/src/win/tcp.c#L168-195
00:16:55  <piscisaureus>bnoordhuis: just because
00:18:29  <piscisaureus>bnoordhuis: probably because
00:18:29  <piscisaureus> if (handle->close_cb) {
00:18:29  <piscisaureus> handle->close_cb((uv_handle_t*)handle);
00:18:29  <piscisaureus> }
00:18:36  <piscisaureus>is the only code that would be shared
00:22:50  <piscisaureus>smartos is very barebones
00:23:07  <piscisaureus>I installed it in a vm but it's very hard to set it up to work conveniently
00:24:02  * loladiroquit (Quit: loladiro)
00:31:32  <bnoordhuis>piscisaureus: https://github.com/bnoordhuis/libuv/commit/33e5fd1
00:32:58  <bnoordhuis>^ i'll squash that commit btw
00:33:42  <piscisaureus>bnoordhuis: lgtm
00:34:22  * loladirojoined
00:35:28  <CIA-155>libuv: Ben Noordhuis master * r122cd94 / src/unix/internal.h : unix: remove unused flag UV__PENDING - http://git.io/uf_uiQ
00:35:29  <CIA-155>libuv: Ben Noordhuis master * r171ad85 / (20 files in 6 dirs): unix, windows: add uv_walk() - http://git.io/XcaOhA
00:35:30  <CIA-155>libuv: Ben Noordhuis master * r71a66c1 / src/unix/uv-eio.c : unix: mark eio poll handles as internal - http://git.io/wOntag
00:35:34  <bnoordhuis>saghul: ^
00:36:15  * mmaleckiquit (Quit: Lost terminal)
00:37:13  * travis-cijoined
00:37:13  <travis-ci>[travis-ci] joyent/libuv#339 (master - 71a66c1 : Ben Noordhuis): The build passed.
00:37:13  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/9d26f49...71a66c1
00:37:13  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/1472246
00:37:13  * travis-cipart
00:41:44  <piscisaureus>hopefully for saghul he sleeps
00:46:25  <bnoordhuis>node must be getting big, i'm getting a lot of node.js headhunter spam lately
00:47:06  <piscisaureus>yeah winter is coming
00:47:12  <piscisaureus>although sometimes...
00:47:16  * c4miloquit (Remote host closed the connection)
00:47:32  <piscisaureus>I got this from flipboard:
00:47:33  <piscisaureus>I came across your personal website and was impressed by your work and interests in web/mobile apps. I'm leading an exclusive search for a talented Front End Software Engineer and/or Web Developer with skills in front end design.
00:48:13  <piscisaureus>I mean, srsly :-(
00:48:39  <bnoordhuis>hah yeah, i got that one too
00:49:12  <piscisaureus>I should've created a colorful site with a lot of comic sans ms
00:52:49  * pieternquit (Quit: pietern)
00:53:05  * xaqquit (Remote host closed the connection)
01:00:05  <piscisaureus>3 am
01:00:07  <piscisaureus>time to go home
01:00:13  <piscisaureus>bnoordhuis: time to go zzzzzzz
01:04:54  <piscisaureus>bnoordhuis: hey, btw, https://github.com/joyent/node/commit/dfcdd5b
01:05:10  <piscisaureus>bnoordhuis: isaacs: you are aware that this probably doesn't work, right :-)
01:05:38  <piscisaureus>because it requires alignment of the read buffer, and the file read offset, and the number of bytes read
01:06:38  <piscisaureus>hmm, nvm
01:07:19  * c4milojoined
01:15:20  * mjr_quit (Quit: mjr_)
01:15:44  <piscisaureus>so painful that windows just supports that just fine, but O_SYNC is not defined :-(
01:15:55  <piscisaureus>we gotta have UV_O_SYNC
01:16:36  <piscisaureus>anyway, go-home time
01:17:46  * brsonquit (Quit: leaving)
01:26:24  * c4miloquit (Remote host closed the connection)
01:28:02  * AvianFluquit (Quit: Leaving)
01:30:21  * bitprobejoined
01:32:03  * piscisaureusquit (Quit: ~ Trillian Astra - www.trillian.im ~)
01:32:25  * bnoordhuisquit (Ping timeout: 245 seconds)
01:46:40  * mikealquit (Quit: Leaving.)
01:51:32  * c4milojoined
01:54:01  * c4miloquit (Remote host closed the connection)
01:54:13  * bbenviechanged nick to benvie
01:54:37  * benviequit
01:54:51  * benviejoined
01:57:50  * abraxasjoined
02:02:46  * TooTallNatequit (Ping timeout: 244 seconds)
02:04:46  * toothrquit (Quit: here we are)
02:56:51  * xaqjoined
03:14:24  * theColejoined
03:18:15  * TooTallNatejoined
03:18:21  * TooTallNatequit (Client Quit)
03:31:12  * xaqquit (Remote host closed the connection)
03:38:02  * iraquit (Quit: Computer has gone to sleep.)
03:41:51  * seebees1joined
03:43:55  * toothrjoined
03:49:00  * bitprobequit (Quit: Computer has gone to sleep.)
04:01:51  * irajoined
04:11:40  * iraquit (Quit: Computer has gone to sleep.)
04:23:00  * theColequit (Quit: theCole)
04:47:27  * theColejoined
04:52:10  * milanijoined
04:52:44  <milani>hey what is this new thing in libuv? uv: add interface for joining external event loops
05:44:06  * paddybyersjoined
05:46:35  * milanipart ("Ex-Chat")
05:51:13  * avalanche123quit (Quit: Computer has gone to sleep.)
05:51:13  * avalanch_changed nick to avalanche123
06:00:52  * loladiroquit (Quit: loladiro)
06:07:27  * bulatshakirzyanojoined
06:55:52  * isaacsquit (Remote host closed the connection)
06:56:44  * isaacsjoined
06:56:50  * theColequit (Quit: theCole)
07:03:59  * isaacsquit (Ping timeout: 244 seconds)
07:22:26  * theColejoined
07:38:50  * rendarjoined
08:43:50  * mikealjoined
09:20:25  <einaros>how's Joyent's policy on people using the node.js logo for posters promoting community user groups?
09:28:04  * abraxasquit (Remote host closed the connection)
09:28:41  * abraxasjoined
09:32:50  * abraxasquit (Ping timeout: 245 seconds)
09:52:49  * theColequit (Ping timeout: 244 seconds)
09:52:54  * theCole_joined
11:17:37  * paddybyersquit (Quit: paddybyers)
11:21:33  * paddybyersjoined
11:41:24  * irajoined
11:44:21  * iraquit (Client Quit)
12:38:04  * AndreasMadsenjoined
12:41:04  <saghul>ircretary tell bnoordhuis and piscisauseus that they rock :-)
12:41:05  <ircretary>saghul: I'll be sure to tell bnoordhuis
12:41:10  <saghul>ircretary thanks
12:41:10  <ircretary>saghul: You're welcome :)
12:49:07  * bnoordhuisjoined
12:51:14  * irajoined
13:05:06  * AndreasMadsenquit (Remote host closed the connection)
13:16:18  * piscisaureus_joined
13:18:59  * piscisaureus_changed nick to piscisaureus
13:29:26  * paddybyersquit (Quit: paddybyers)
13:43:16  * AndreasMadsenjoined
13:43:59  * AndreasMadsenquit (Remote host closed the connection)
13:49:15  * paddybyersjoined
13:56:14  * c4milojoined
14:08:58  * xaqjoined
14:24:31  * loladirojoined
14:49:53  * bitprobejoined
15:09:39  * paddybyersquit (Quit: paddybyers)
15:09:44  * felixgejoined
15:09:44  * felixgequit (Changing host)
15:09:45  * felixgejoined
15:20:29  * isaacsjoined
15:28:35  <indutny>bnoordhuis: heya
15:29:35  * mmaleckijoined
15:31:55  * mmalecki_joined
15:32:20  * mmaleckiquit (Client Quit)
15:32:37  * mmaleckijoined
15:33:29  <indutny>ircretary: please tell bnoordhuis to ping me
15:33:29  <ircretary>indutny: I'll be sure to tell bnoordhuis
15:44:05  <isaacs>wow, the v8 in 0.7.9 is weird
15:44:12  * AvianFlujoined
15:44:14  <isaacs>upgrading it fixes a bunch of oddball issues.
15:44:37  * mmalecki_quit (Disconnected by services)
15:45:19  <mmalecki>YES.
15:47:12  * rphillipsquit (Ping timeout: 248 seconds)
15:47:44  * pquernaquit (Ping timeout: 248 seconds)
15:48:07  <isaacs>mmalecki: 3.11.7 looks good, though
15:48:16  <isaacs>mmalecki: i'll land this once indutny's stuff gets in
15:49:10  <mmalecki>indutny: cool, when's that?
15:49:13  <mmalecki>er, isaacs
15:58:02  * paddybyersjoined
15:58:06  * loladiroquit (Quit: loladiro)
16:00:38  <isaacs>mmalecki: soon. definitely before 0.7.10 (which may be 0.8.0)
16:02:18  <mmalecki>isaacs: ok, cool
16:03:03  <mmalecki>isaacs: btw, let's wait for AvianFlu's detached spawn stuff, ok?
16:03:34  <isaacs>yeah, it'd be goodto get that in
16:04:13  <AvianFlu>https://github.com/AvianFlu/libuv/compare/detached2
16:04:20  <AvianFlu>much simpler than last time
16:05:13  <AvianFlu>might still want another test, and maybe some stupid-proofing
16:05:26  * loladirojoined
16:06:49  <isaacs>AvianFlu: yeah, that is a much smaller patch than i remember
16:09:39  * loladiroquit (Ping timeout: 246 seconds)
16:21:47  * xaqquit (Remote host closed the connection)
16:33:07  <indutny>isaacs: I think we only need bnoordhuis' sign off on my commit to pull it
16:33:29  <indutny>bnoordhuis: https://github.com/indutny/node/commit/434b81926258c90d1714546eeb154271c44cb8fd ;)
16:33:35  <indutny>mmalecki: ^
16:38:00  * mmalecki_joined
16:38:07  * felixgequit (Quit: felixge)
16:40:10  * mmaleckiquit (Ping timeout: 252 seconds)
16:40:56  <isaacs>indutny: great.
16:41:23  <isaacs>indutny: so, you've been in that part of libuv a lot recently. Wanna review AvianFlu's detached2 branch? https://github.com/AvianFlu/libuv/compare/detached2
16:41:48  <indutny>isaacs: sure
16:52:45  * mmalecki_changed nick to mmalecki
17:02:55  * mmaleckiquit (Quit: Reconnecting)
17:03:04  * AlbireoXquit (Read error: Connection reset by peer)
17:03:08  * mmaleckijoined
17:07:30  * xaqjoined
17:08:18  <indutny>isaacs: lgtm
17:08:24  <indutny>isaacs: though not sure about windows part
17:08:28  <isaacs>k
17:08:34  <isaacs>piscisaureus: ping
17:08:34  <indutny>isaacs: better ask piscisaureus or igorzi about it
17:08:41  <isaacs>igorzi: ping
17:09:39  * ericktjoined
17:15:46  * TheJHjoined
17:37:51  <piscisaureus>indutny: hey
17:38:05  <piscisaureus>indutny: why is UV_IGNORE defined as 0 ?
17:38:37  <piscisaureus>indutny: how can you distinguish the ignore case from the case where we have to grab an fd out of an existing uv pipe?
17:40:54  <indutny>piscisaureus: well, because we're doing following check flags == UV_IGNORE
17:41:09  <indutny>piscisaureus: there're no such case
17:41:10  <piscisaureus>indutny: yes
17:41:21  <indutny>piscisaureus: you either pass fd or pipe without fd
17:41:27  * AndreasMadsenjoined
17:41:31  <indutny>passing pipe with fd doesn't really make sense
17:41:32  <indutny>isn't it?
17:41:52  <piscisaureus>indutny: well if the user has an existing stream
17:41:56  <piscisaureus>like process.stdin
17:42:01  <piscisaureus>how do we pass that to the child?
17:42:39  <indutny>we're taking handle's fd and passing it
17:43:51  <piscisaureus>indutny: but certain handle types don't have an FD, like tty
17:44:09  <indutny>piscisaureus: how are you going to share it with child then?
17:44:30  <indutny>no FD - no way to share
17:44:39  <piscisaureus>indutny: rip the windows HANDLE out of th uv_tty_t and share that
17:45:10  <indutny>piscisaureus: hm... that what I was asking you some time before. Is FD everything that I need to share streams on windows
17:45:27  <indutny>piscisaureus: ok, so we should move that API into uv
17:45:37  <indutny>extraction of fd
17:45:43  <piscisaureus>yeah, exactly
17:45:49  <piscisaureus>lemme finish this windows thing first
17:45:52  <piscisaureus>before we move on
17:46:05  * seebees1quit (Ping timeout: 244 seconds)
17:46:23  <AndreasMadsen>isaacs: the v0.7 cluster mode has a Worker.uniqueID property, personally I don't care but I has been suggested that it should be called cluster.workerID. I opened an issue about it two months ago (https://github.com/joyent/node/issues/3007) and since 0.8 it about to land I would just like to make sure that some decision is made.
17:48:17  <indutny>piscisaureus: ok
17:48:21  <indutny>piscisaureus: just ping me when you done
17:50:25  * seebeesjoined
17:54:30  <isaacs>AndreasMadsen: *shrug*.
17:54:38  <isaacs>AndreasMadsen: is it a big deal? seems ok as uniqueID
17:54:52  <isaacs>AndreasMadsen: but then, why not just worker.id and cluster.id?
17:55:57  <AndreasMadsen>isaacs: no it is not a big deal, but the name was chosen from a different perspective and better names do exist.
17:56:32  <isaacs>AndreasMadsen: it seesm like just calling it "id" would be best.
17:56:44  * AlbireoXjoined
17:57:10  <AndreasMadsen>isaacs: Sure, will make a patch soon
17:57:14  <mmalecki>worker.clusterWorkerUniquePerWholeClusterIntegerId
17:57:24  <mmalecki>no need to thank, I can even submit a patch
17:58:22  <AndreasMadsen>mmalecki: I would go with a big D at the end ID looks better than Id
17:59:34  <AvianFlu>worker.clusterUniqueIDManyAreLikeItButThisOneIsMine
18:00:21  <mmalecki>hahahaha
18:00:23  <mmalecki>lol
18:00:35  <mmalecki>AndreasMadsen: yeah, looks more enterprise
18:01:23  <piscisaureus>cluster.isaacs
18:02:10  <AndreasMadsen>just for the record the cluster object don't have a ID
18:02:18  <piscisaureus>oh
18:02:24  <piscisaureus>let's add that then
18:02:40  <piscisaureus>maybe cluster.ggreer
18:02:52  <piscisaureus>or cluster.indutny
18:03:27  <indutny>I like this
18:03:38  <indutny>lets rename node
18:03:47  <indutny>to ben's node
18:03:57  <AndreasMadsen>mmalecki : more enterprise is worker.cloud_ID
18:03:58  <indutny>or nodehuis
18:04:09  <indutny>bnoordhuis: no offense meant
18:04:42  <mmalecki>lol, a cluster node which'd spawn VMs at some provider and deploy code there
18:04:42  * AlbireoX_joined
18:04:52  <mmalecki>AvianFlu: mad science project! ^
18:05:05  * AlbireoXquit (Read error: Connection reset by peer)
18:05:18  * brsonjoined
18:05:20  <indutny>mmalecki: omg
18:06:03  <piscisaureus>require('indutny').isaacs('www.google.com', function() {
18:06:04  <AndreasMadsen>can you not just add all the coreteam member names to the constant module and get over with it (remember to document it)
18:06:21  <mmalecki>indutny: that could totally work.
18:07:34  <indutny>piscisaureus: hahahaha
18:11:24  * mmaleckiquit (Quit: leaving)
18:12:45  <igorzi>piscisaureus indutny: i was also just thinking about this today.. currently there's actually no way to pass in an existing uv_stream_t :)
18:13:04  <igorzi>piscisaureus indutny: we should probably just change UV_IGNORE to UV_STREAM (or something like that)
18:13:25  <igorzi>isaacs: pong
18:15:13  <piscisaureus>igorzi: yep, I was already doing that
18:15:56  <igorzi>piscisaureus: ok cool
18:18:58  <isaacs>igorzi, piscisaureus: can you guys review https://github.com/AvianFlu/libuv/compare/detached2
18:19:11  <isaacs>especially the windows bits
18:19:37  * mmaleckijoined
18:20:46  <AndreasMadsen>^-- that made my day :D
18:21:51  <piscisaureus>AvianFlu: hey, I noticed that the unix implementation no longer double forks.
18:22:12  <piscisaureus>AvianFlu: doesn't that mean the we can just keep the exit callback like it used to ?
18:22:16  <AvianFlu>piscisaureus, you're correct. previously, I couldn't get the parent to exit without the double fork
18:22:31  <AvianFlu>now, apparently thanks to the refcount refactor, it behaves like I expect with only one fork
18:23:03  <AvianFlu>well, it seemed to me that I had to bypass the child watcher, or else the parent would just end up waiting for the child still
18:23:13  <AvianFlu>I might be wrong, but I think it was necessary?
18:23:25  <piscisaureus>AvianFlu: ah right
18:23:58  <piscisaureus>AvianFlu: that's correct... however it's sorta lame., since we could also just uv_unref the process handle now that the refcount refactor has landed
18:24:23  <AvianFlu>I'd be fine with that
18:24:35  <piscisaureus>Maybe we could have a node binding for that or something
18:24:48  <AvianFlu>so instead of bypassing the ev child watcher stuff, just unref the process handle at the end?
18:24:56  <piscisaureus>AvianFlu: yeah
18:25:03  <AvianFlu>ok
18:25:10  * xaqquit (Read error: Connection reset by peer)
18:25:15  <piscisaureus>AvianFlu: also, setsid() cannot fail :-)
18:25:34  * xaqjoined
18:25:37  <piscisaureus>unless the process was a group leader already, in which case you got exactly what you wanted ;_)
18:25:50  <AvianFlu>various manpages made it sound like there were a few various super-obscure things worth checking for
18:25:54  <AvianFlu>but yeah, you're right
18:27:08  * rphillipsjoined
18:28:01  * pquernajoined
18:28:50  * pquernaquit (Changing host)
18:28:50  * pquernajoined
18:33:27  * theCole_quit (Quit: theCole_)
18:44:10  <creationix>has anyone seen a bug in 0.6.18 where cleared timeouts still fire?
18:53:02  * loladirojoined
19:01:19  <AvianFlu>so piscisaureus, to be clear, uv_spawn with detached set is just gonna start the child and return? you want to just ignore the callback?
19:03:16  * `3rdEdenjoined
19:04:38  <piscisaureus>AvianFlu: well I want to to behave like a normal child process
19:04:56  <piscisaureus>so it should just fire ' exit' when it actually exits
19:05:22  * TooTallNatejoined
19:06:57  <AvianFlu>and then the unref will prevent it from keeping the parent up, if you're trying to leave it in the background after the parent exits?
19:07:08  <piscisaureus>yep
19:07:15  <AvianFlu>and you just might not see exit, if it doesn't exit
19:07:17  <AvianFlu>that makes sense
19:07:26  <AvianFlu>cool
19:08:04  <piscisaureus>AvianFlu: I am actually in doubt whether I would want to unref automatically,
19:09:24  <piscisaureus>AvianFlu: maybe we should just have a node binding for uv_ref/uv_unref now
19:10:33  <AvianFlu>so detached would just setsid or set the windows flags, and then a libuv user would worry about unrefing for themselves, and we'd do that part in node for node users?
19:10:47  <piscisaureus>yes
19:11:00  <piscisaureus>unref() is easy now, right :-)
19:11:03  <AvianFlu>certainly makes sense
19:11:05  <AvianFlu>and yes!
19:11:09  <AvianFlu>much less voodoo.
19:20:25  * arlolrajoined
19:24:45  <bnoordhuis>indutny: pong
19:24:52  * irachanged nick to ira_afk
19:35:15  <piscisaureus>I have to run
19:35:22  <piscisaureus>bnoordhuis: coming to 020 tomorrow?
19:35:42  <bnoordhuis>piscisaureus: yep
19:35:58  <mmalecki>bnoordhuis: answer to my pm, would you?
19:35:59  <piscisaureus>bnoordhuis: time?
19:36:14  <bnoordhuis>piscisaureus: same as always, around noon
19:36:19  <bnoordhuis>piscisaureus: read: 1 am-ish
19:36:25  <bnoordhuis>mmalecki: pm?
19:36:26  <bnoordhuis>oh
19:36:37  * loladiroquit (Quit: loladiro)
19:36:54  <bnoordhuis>piscisaureus: make that 1 pm-ish :)
19:37:03  <piscisaureus>heh
19:37:15  <piscisaureus>I was already wondering :p
19:37:36  <isaacs>creationix: no, that sounds weird
19:37:39  <isaacs>creationix: got a test?
19:38:32  * piscisaureusquit (Quit: ~ Trillian Astra - www.trillian.im ~)
19:41:10  * ira_afkchanged nick to ira
19:42:05  <indutny>bnoordhuis: pang
19:42:21  <bnoordhuis>indutny: what's up?
19:42:35  * loladirojoined
19:42:36  <indutny>bnoordhuis: well, nothing now
19:42:49  <indutny>bnoordhuis: but I wanted to ask you to review that https://github.com/indutny/node/compare/feature-child-process-stdio
19:43:07  <indutny>bnoordhuis: now I'm waiting for bert to finish his API changes to uv_spawn's options
19:43:17  <indutny>bnoordhuis: and mirror them in unix side
19:43:30  <indutny>after that I'll make changes to that node.js branch
19:43:34  <bnoordhuis>oh, so i don't have to do anything?
19:43:36  <indutny>so no need in reviewing it now
19:43:40  <bnoordhuis>cool
19:43:43  * bnoordhuisleans back again
19:43:47  <indutny>:)
19:53:58  * AndreasMadsenquit (Remote host closed the connection)
20:00:41  <creationix>bnoordhuis, luvmonkey!
20:07:00  * loladiroquit (Quit: loladiro)
20:20:41  * stephankjoined
20:23:50  * theColejoined
20:32:12  * mjr_joined
20:34:51  * mikealquit (Quit: Leaving.)
20:41:21  * toothrquit (Ping timeout: 240 seconds)
20:43:04  * toothrjoined
20:48:38  * piscisaureus_joined
20:53:50  * `3rdEdenquit (Quit: Leaving...)
21:09:56  * piscisaureus_quit (Read error: Connection reset by peer)
21:10:26  * piscisaureus_joined
21:17:43  <creationix>I can't believe all the anger over the nextTick proposal
21:19:05  <tjfontaine>opinions and assholes
21:21:40  <bnoordhuis>piscisaureus_: why do we have uv_timer_again?
21:22:21  <bnoordhuis>actually... the whole concept of repeating timers seems kind of silly, the user's callback can restart them
21:22:47  <piscisaureus_>bnoordhuis: yes
21:22:56  <piscisaureus_>bnoordhuis: it's because of libev
21:23:05  <creationix>I know in browsers setInterval acts different than a recursive setTimeout
21:23:21  <creationix>but only under cpu load
21:23:36  <bnoordhuis>piscisaureus_: i don't mind stripping that out, i'm working on timers now anyway
21:23:38  <creationix>and node acts rather different than most browsers in this area too
21:24:27  <piscisaureus_>bnoordhuis: not now, please
21:25:06  <bnoordhuis>aah?
21:25:19  <piscisaureus_>bnoordhuis: I mean, don't strip timers now :-(
21:25:27  <piscisaureus_>bnoordhuis: we have bigger fish to fty
21:25:29  <piscisaureus_>*fry
21:26:18  <bnoordhuis>what fish?
21:26:56  * elijah-mbpquit (Ping timeout: 276 seconds)
21:28:00  <piscisaureus_>er, finish 0.8 ?
21:29:04  * elijah-mbpjoined
21:29:04  <piscisaureus_>bnoordhuis: git show abd2d1c
21:29:18  <piscisaureus_>bnoordhuis: at some point there were no timers
21:29:26  <piscisaureus_>I'd have to dig up why we changed it tho
21:30:25  <bnoordhuis>piscisaureus_: next question, why are the timeout and repeat signed integers?
21:31:01  <piscisaureus_>umm well just because
21:31:14  <piscisaureus_>setting a negative timeout does the same as setting a 0 timeout iirc
21:31:27  <piscisaureus_>repeat should always be > 0
21:31:53  <bnoordhuis>we might as well make them unsigned
21:32:09  <bnoordhuis>maybe unsigned 32 bits to save some space. then again...
21:32:30  <piscisaureus_>ah shit
21:32:38  <piscisaureus_>the timer api was discussed on convore
21:32:50  <bnoordhuis>convore... name rings a bell
21:32:55  <piscisaureus_>we experimented with doing design discussions on convore for a while
21:33:08  <piscisaureus_>but convore is dead
21:33:32  <bnoordhuis>oh, that convore
21:33:43  <piscisaureus_>https://convore.com/node/oio-timer-api/?created=true
21:33:43  <piscisaureus_>https://convore.com/nodejs-7/liboio-design/?created=true
21:33:45  <indutny>piscisaureus_: how is it going with stdio?
21:33:54  <indutny>;)
21:34:01  <piscisaureus_>steady
21:34:06  <piscisaureus_>getting impatient eh :-)
21:34:54  <indutny>ok, going to sleep
21:34:55  <indutny>ttyl :)
21:34:59  <bnoordhuis>sleep tight, fedor
21:35:24  <piscisaureus_>indutny: https://github.com/piscisaureus/libuv/commit/d0b36ce79d71ab3531d592bd4283f4a36be094f1
21:35:35  <piscisaureus_>indutny: this attempt will work :-)
21:35:38  <indutny>:)
21:36:47  <indutny>piscisaureus_: oook, I see
21:36:48  <indutny>:)
21:36:48  <piscisaureus_>bnoordhuis: If you care you can add something like timer_req
21:36:51  <indutny>thank you
21:36:52  <indutny>ttyl
21:37:34  <piscisaureus_>bnoordhuis: I think killing uv_timer on such a short notice is kinda rude
21:37:52  <piscisaureus_>bnoordhuis: but uv_timer could just use uv_timeout_req under the hood, then
21:38:12  <piscisaureus_>bnoordhuis: oh you know, there is one issue with it...
21:38:30  <piscisaureus_>bnoordhuis: we don't have ways to cancel reqs atm but you do want to cancel timers every now and then
21:40:40  * dapjoined
21:43:12  * elijah-mbpquit (Ping timeout: 246 seconds)
21:50:47  * elijah-mbpjoined
21:54:18  <bnoordhuis>piscisaureus_: okay, i'll let it live for now
21:56:18  * mikealjoined
21:59:17  * bitprobequit (Quit: Computer has gone to sleep.)
21:59:48  * c4miloquit (Remote host closed the connection)
22:02:50  * piscisaureus_quit (Ping timeout: 245 seconds)
22:02:53  * rendarquit
22:04:15  * theColequit (Quit: theCole)
22:08:42  * piscisaureus_joined
22:10:08  * TheJHquit (Read error: Operation timed out)
22:19:10  * paddybyers_joined
22:19:27  * mmalecki_joined
22:21:20  * mmaleckiquit (Ping timeout: 240 seconds)
22:21:42  * paddybyersquit (Ping timeout: 260 seconds)
22:21:43  * paddybyers_changed nick to paddybyers
22:22:50  * mmalecki_changed nick to mmalecki
22:25:42  * mmalecki_joined
22:28:16  * mmaleckiquit (Ping timeout: 248 seconds)
22:30:29  * piscisaureus_quit (Quit: ~ Trillian Astra - www.trillian.im ~)
22:50:25  * mmalecki_changed nick to mmalecki
23:02:20  * xaqquit (Remote host closed the connection)
23:14:54  * mikealquit (Quit: Leaving.)
23:15:03  * benviequit
23:34:04  * paddybyersquit (Quit: paddybyers)
23:34:46  * mikealjoined
23:49:21  * arlolraquit (Quit: Linkinus - http://linkinus.com)
23:54:59  <bnoordhuis>i think fs.watchFile is broken...
23:55:31  <bnoordhuis>`node -e 'require("fs").watchFile(".", console.error)'` blocks with v0.6, immediately quits with master
23:56:17  <bnoordhuis>it's using the iowatcher bindings...