00:14:22  <isaacs>igorzi: pushed: https://github.com/isaacs/npm/commit/cc0a00e68fb798c4daea9faf2e0b9508f5c15662
00:15:05  <isaacs>nice big red section there :)
00:22:27  * mikealjoined
00:46:28  <piscisaureus_>igorzi: hey, yt?
00:48:08  * mmaleckichanged nick to mmalecki[zzz]
01:04:32  * mikealquit (Quit: Leaving.)
01:06:27  * orlandovftwquit (Ping timeout: 252 seconds)
01:14:14  * saghulquit (Ping timeout: 245 seconds)
01:15:01  * AvianFlujoined
01:20:19  <TooTallNate>isaacs: is the npm in 0.6.13 gonna do the node-gyp thing?
01:27:11  <piscisaureus_>igorzi: did the SDK problem that Interlocked intrinsics are not enabled ever get publicly acknowledged?
01:34:12  * abraxasjoined
01:35:43  <isaacs>TooTallNate: yes.
01:47:10  * perezdquit (Quit: perezd)
01:47:37  <isaacs>kohai: ++
01:47:37  <kohai>I am Kohai, semi-useful communications-facilitating pseudointelligence!
01:47:38  <kohai>kohai has 10 beers
01:47:42  <isaacs>kohai: ++
01:47:44  <kohai>I am Kohai, semi-useful communications-facilitating pseudointelligence!
01:47:44  <kohai>kohai has 11 beers
01:47:46  <isaacs>kohai: ++
01:47:47  <kohai>I am Kohai, semi-useful communications-facilitating pseudointelligence!
01:47:47  <kohai>kohai has 12 beers
01:50:22  <TooTallNate>mjr_: for god sake! don't ctrl+c twice in the repl in production!
01:50:56  <mjr_>does that do what I'm just now realizing it might do?
01:51:20  <TooTallNate>https://github.com/joyent/node/blob/master/lib/repl.js#L139
01:51:35  <isaacs>lol
01:51:39  <isaacs>i think i put that code there
01:51:40  <TooTallNate>p.s. i'm working on some patches
01:51:45  <TooTallNate>https://github.com/TooTallNate/node/compare/repl
01:51:58  <TooTallNate>isaacs: you did
01:52:05  <isaacs>TooTallNate: so, what would be really ideal here is to somewhat separate out the "tty" portion of readline/repl
01:52:39  <isaacs>TooTallNate: so, maybe, ^C^C will emit some event that says "kill this thing", but if that's a socket, then it just disconnects.
01:52:58  * mikealjoined
01:53:01  <isaacs>TooTallNate: and then the default repl in node.js can process.exit there
01:53:09  <TooTallNate>isaacs: i just made REPLServer an acutal eventemitter
01:53:13  <TooTallNate>and have it emit "exit"
01:53:14  <isaacs>perfect
01:53:51  * mikealquit (Client Quit)
01:54:02  <isaacs>i'd say (what it sounds like you're already doing) a good constraint here shoudl be that you can use it over a TTY or socket nicely, but also, there's no observable difference in the normal "node <noargs>" behavior we have today
01:54:17  <mjr_>Man, I love repl.
01:54:24  <isaacs>and where "node <noargs>" kills the process, it should kill the connection
01:54:29  <mjr_>So amazing to be able to change your program while it is running.
01:54:36  <isaacs>mjr_: yeah, it's magic.
01:55:02  <TooTallNate>isaacs: i agree with your constraint
01:55:20  <TooTallNate>isaacs: side note, there should be a `node --repl` flag
01:55:33  <TooTallNate>so we can spawn node processes in the repl
01:55:39  <mjr_>We have solved a couple of the hairiest of problems using REPL sockets, and also implemented a very complicated admin feature that way.
01:55:45  <TooTallNate>i know indexzero wanted that back
01:56:39  * mikealjoined
01:56:57  * brsonquit (Ping timeout: 264 seconds)
02:01:18  <piscisaureus_>people are idiots
02:01:50  <piscisaureus_>I have seen like a million tweets about node.cs
02:02:00  <piscisaureus_>that project doesn't even work and has been abandonded for almost a year now
02:02:11  <benvie>yeah I was wondering about that
02:02:19  <benvie>unless we both missed a beat that isn't news worthy
02:02:20  <piscisaureus_>it binds to libev
02:02:47  <piscisaureus_>so even if it would work it would be as slow as node 0.4.4 on mingw
02:02:52  <piscisaureus_>and let me tell you, that was slow
02:02:56  <benvie>anything that doesn't involve libuv isn't worth talking about in relation to node
02:03:15  <piscisaureus_>well, thank you for the compliment, although I may not agree
02:03:31  <benvie>well not ANYTHING
02:03:31  <piscisaureus_>but at least it should involve a sane implementation :-)
02:03:41  <benvie>but if it involes the non-js part
02:04:05  <piscisaureus_>if people do node.cs with whatever .net offers out of the box, that could also be nice
02:04:13  <piscisaureus_>but libev, c'mon
02:05:20  <benvie>node has a lot of quite good js in it and a decent amount of pretty not great js in it, and then it has bindings to make it all work, and then it has libuv that helps it tell a story that's not "so I was on unix and I wanted,,,"
02:06:37  * orlandovftwjoined
02:08:09  <piscisaureus_>igorzi: still not there?
02:08:24  <piscisaureus_>WHERE IS IGORZI
02:08:28  <piscisaureus_>IMMA FIRE HIM
02:08:36  <isaacs>TooTallNate: what would --repl do?
02:08:44  <isaacs>TooTallNate: force it to be in the repl even if there's args?
02:08:51  <benvie>I would guess it's like -i in other stuff
02:08:57  <benvie>interactive modes
02:09:00  * perezdjoined
02:09:01  <isaacs>yeah
02:09:06  <isaacs>but, you can just run "node" to get that
02:09:13  <benvie>well what if you do
02:09:15  <TooTallNate>isaacs: force it to do repl even if stdio isn't a tty
02:09:22  <isaacs>oh, ok, i get it
02:09:23  <isaacs>yeah
02:09:30  <isaacs>sure, patch welcome.
02:09:32  <isaacs>should be -i, though
02:09:35  <benvie>node -e "require('child_process').exec(\'node\')"
02:09:41  <benvie>is closer
02:09:52  * perezdquit (Client Quit)
02:10:32  <benvie>and yeah -i is that thing I believe in perl, python, php, and a few other things so it has a strong tradition of being -i =D
02:10:32  <TooTallNate>benvie: ^ what's that supposed to do?
02:10:40  <benvie>nothing useful right now
02:10:51  <benvie>basically just saying
02:11:13  <benvie>there's times you run it through some interface that's not obviously interactive, but want node to treat it that way besides
02:11:20  * dshaw_quit (Ping timeout: 255 seconds)
02:11:23  <isaacs>benvie: right
02:11:28  <benvie>which is what you're patching right?
02:12:02  <benvie>err
02:12:12  <benvie>what someone theoretically would patch if they chose to do so
02:12:21  <TooTallNate>i would think it would be a quick patch, not too high of a priority to me
02:12:38  <TooTallNate>this repl stuff is more fun :)
02:12:48  <benvie>yeah I realized that after the fact =D
02:23:06  * mikealquit (Quit: Leaving.)
02:24:45  <TooTallNate>anyone wanna review? https://github.com/joyent/node/pull/2919
02:25:14  <TooTallNate>that's just the start, i'll work on refactoring tty stuff next
02:25:45  <benvie>I love repls
02:25:54  <benvie>I want my whole computer to be a repl
02:26:11  <benvie>a javascript repl ;)
02:33:15  <bnoordhuis>https://github.com/joyent/node/pull/2919 <- what's up with github?
02:36:35  <mjr_>We are currently making 24,000 HTTP GET requests / sec from node into our riak clusters.
02:38:58  <TooTallNate>bnoordhuis: seems back up now
02:39:02  * pieternquit (Quit: pietern)
02:39:19  <bnoordhuis>still not working for me
02:39:25  <bnoordhuis>oh well, i'll look at it tomorrow
02:39:37  <bnoordhuis>btw, that's an awful lot of http requests, mjr_
02:39:41  * bnoordhuissigns off
02:40:21  <mjr_>Yeah, I think we'd do a lot more if we weren't brutalizing the disk drives so much.
02:41:39  * orlandovftwquit (Ping timeout: 244 seconds)
02:45:20  * bnoordhuisquit (Ping timeout: 272 seconds)
02:52:28  * igorzi_joined
02:52:41  <igorzi_>piscisaureus_: whatup?
03:02:26  <piscisaureus_>igorzi_: did this SDK issue that intrinsics are not used ever get publicly acknowledged?
03:02:53  <piscisaureus_>igorzi_: I'd like to send a patch to v8 for review but it would help if there's a bug report or something that I can point to
03:06:45  <igorzi_>piscisaureus_: no, i don't think there's a bug report or anything like that out there
03:06:51  <igorzi_>piscisaureus_: but let me check
03:11:06  * dshaw_joined
03:18:33  * mikealjoined
03:55:09  * igorzi_quit (Quit: Page closed)
04:02:03  * mikealquit (Quit: Leaving.)
04:11:19  * piscisaureus_quit (Ping timeout: 244 seconds)
04:13:25  * orlandovftwjoined
04:35:16  <CIA-99>node: isaacs master * rbcb0cc0 / (251 files in 41 dirs): (log message trimmed)
04:35:16  <CIA-99>node: Upgrade npm to 1.1.6
04:35:16  <CIA-99>node: * Fix #472 Adds support for os/cpu fields in package.json (Adam Blackburn)
04:35:16  <CIA-99>node: * Don't use existing pkg if explicit
04:35:16  <CIA-99>node: * Install missing deps, even if they are in bundleDependencies
04:35:16  <CIA-99>node: * bundle node-gyp
04:35:16  <CIA-99>node: * Automatically node-gyp build stuff with a bindings.gyp
04:49:51  * travis-cijoined
04:49:51  <travis-ci>[travis-ci] joyent/node#581 (master - bcb0cc0 : isaacs): The build is still failing.
04:49:51  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/compare/9701f1c...bcb0cc0
04:49:51  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/852213
04:49:51  * travis-cipart
04:56:49  * isaacsquit (Remote host closed the connection)
05:28:03  * mikealjoined
05:47:49  * Raynosquit (Excess Flood)
05:47:50  * mrb_bkquit (Excess Flood)
05:47:54  * mikealquit (Excess Flood)
05:47:54  * arlolraquit (Excess Flood)
05:47:55  * dshaw_quit (Excess Flood)
05:47:55  * philipsquit (Excess Flood)
05:47:55  * coderarityquit (Excess Flood)
05:48:03  * mikealjoined
05:48:03  * arlolrajoined
05:48:04  * dshaw_joined
05:48:17  * Raynosjoined
05:48:43  * philipsjoined
05:48:57  * coderarityjoined
05:49:05  * dshaw_quit (Excess Flood)
05:49:05  * mikealquit (Excess Flood)
05:49:08  * arlolraquit (Excess Flood)
05:49:13  * dshaw_joined
05:49:16  * mikealjoined
05:49:33  * arlolrajoined
05:50:36  * mikealquit (Client Quit)
05:59:55  * mrb_bkjoined
06:16:20  * Raynosquit (Excess Flood)
06:16:20  * mrb_bkquit (Excess Flood)
06:16:25  * coderarityquit (Excess Flood)
06:16:25  * dshaw_quit (Excess Flood)
06:16:26  * arlolraquit (Excess Flood)
06:16:26  * philipsquit (Excess Flood)
06:16:47  * dshaw_joined
06:16:48  * Raynosjoined
06:17:03  * arlolrajoined
06:18:13  * philipsjoined
06:18:25  * coderarityjoined
06:24:28  <igorzi>isaacs: hey, yt?
06:30:25  * mrb_bkjoined
06:34:03  * natahnjoined
06:34:08  * TooTallNatequit (Read error: Connection reset by peer)
06:34:20  * natahnquit (Client Quit)
06:34:25  * Raynosquit (Excess Flood)
06:36:04  * toothrquit (Ping timeout: 240 seconds)
06:37:07  * igorziquit (Ping timeout: 245 seconds)
06:37:32  * AvianFluquit (Ping timeout: 244 seconds)
06:38:03  * einarosquit (Ping timeout: 244 seconds)
06:39:45  * einarosjoined
06:41:42  * toothrjoined
06:42:28  * mralephjoined
06:43:34  * Raynosjoined
06:43:42  * mralephquit (Remote host closed the connection)
06:43:44  * Raynosquit (Excess Flood)
06:50:40  * AvianFlujoined
06:53:27  * Raynosjoined
06:53:38  * Raynosquit (Excess Flood)
06:54:16  * paddybyersjoined
07:03:24  * mikealjoined
07:04:22  * Raynosjoined
07:36:29  * mjr__joined
07:39:08  * mjr_quit (Ping timeout: 244 seconds)
07:39:08  * mjr__changed nick to mjr_
07:44:00  * dshaw_quit (Quit: Leaving.)
08:04:58  * stephankquit (Quit: *Poof!*)
08:06:51  * orlandovftwquit (Ping timeout: 265 seconds)
08:37:01  * rendarjoined
08:52:16  * paddybyersquit (Read error: Connection reset by peer)
08:52:59  * paddybyers_joined
08:56:12  * paddybyers_quit (Client Quit)
08:57:53  * dshaw_joined
09:29:35  * mmalecki[zzz]changed nick to mmalecki
09:36:07  * mikealquit (Ping timeout: 252 seconds)
09:38:10  * mmaleckichanged nick to mmalecki[away]
09:40:25  * mikealjoined
09:43:01  * rendarquit (Ping timeout: 244 seconds)
09:44:47  * dshaw_quit (Quit: Leaving.)
09:47:36  * rendarjoined
09:49:19  * indutny_sleepingchanged nick to indutny
09:50:19  * ryahquit (*.net *.split)
09:50:19  * creationixquit (*.net *.split)
09:55:49  * ryahjoined
09:55:50  * creationixjoined
10:00:27  * mmalecki[away]changed nick to mmalecki
10:39:36  * mmaleckichanged nick to mmalecki[away]
10:42:55  * theColejoined
10:51:15  * theColequit (Quit: theCole)
10:56:34  * abraxasquit
10:59:52  * paddybyersjoined
11:07:24  * saghuljoined
11:07:42  * saghulquit (Client Quit)
11:08:09  * saghuljoined
12:23:36  * bnoordhuisjoined
12:38:27  * theColejoined
12:41:44  * theColequit (Client Quit)
14:01:14  <bnoordhuis>ircretary: tell piscisaureus_ does msvc have some kind of facility to always #include a file? something akin to gcc's -include foo.h?
14:01:14  <ircretary>bnoordhuis: I'll be sure to tell piscisaureus_
14:47:06  * piscisaureus_joined
14:47:29  <piscisaureus_>bnoordhuis: yes, there is. but why?
14:47:50  <bnoordhuis>piscisaureus_: smjs needs it
14:51:05  <piscisaureus_>'VCCLCompilerTool': { 'ForcedIncludeFiles': 'foo.h;bar.h' }
14:51:27  <piscisaureus_>bnoordhuis: there probably is a gyp shortcut for it, e.g. msv_forced_include_files or something
14:51:38  <piscisaureus_>or maybe gyp has a provision for it anyway
14:52:11  <bnoordhuis>okay, thanks
14:54:41  * isaacsjoined
15:07:18  <isaacs>Whoops. https://github.com/joyent/node/commit/bcb0cc0b184c61313541002e504948ed21f59dad
15:07:29  <isaacs>should really change the commit messages to isaacs/npm#1234
15:07:36  <isaacs>luckily both bugs were already closed :)
15:07:42  * theColejoined
15:44:32  * dvv1joined
15:44:39  * dvv1part
15:51:49  * theColequit (Quit: theCole)
15:58:01  * mmalecki[away]changed nick to mmalecki
16:00:28  * dshaw_joined
16:02:40  * mikealquit (Quit: Leaving.)
16:13:23  * mikealjoined
16:19:48  * isaacstopic: test, please: http://nodejs.org/dist/v0.7.6/node-v0.7.6-RC0.tar.gz
16:19:50  <isaacs>test, please: http://nodejs.org/dist/v0.7.6/node-v0.7.6-RC0.tar.gz
16:21:13  * pieternjoined
16:21:56  <CIA-99>node: isaacs v0.7.6-release * re4fc2cb / (194 files in 16 dirs): Upgrade v8 to 3.9.17 - http://git.io/pe3Nmw
16:21:56  <CIA-99>node: isaacs v0.7.6-release * r4e85257 / deps/v8/build/common.gypi : Patches floating on V8 - http://git.io/qUOkug
16:21:56  <CIA-99>node: Christian Ress v0.7.6-release * rc15e690 / lib/zlib.js : Added destroy method to Zlib object in zlib.js module - http://git.io/Vgjk_Q
16:21:57  <CIA-99>node: isaacs v0.7.6-release * r943c396 / doc/api/readline.markdown : Fix merge conflict in doc/api/readline.markdown - http://git.io/WlazYg
16:21:58  <CIA-99>node: isaacs v0.7.6-release * r4ec3bc5 / (7 files in 6 dirs): (log message trimmed)
16:21:58  <CIA-99>node: 2012.03.13, Version 0.7.5 (unstable)
16:21:58  <CIA-99>node: * Upgrade v8 to 3.9.17
16:21:58  <CIA-99>node: * Upgrade npm to 1.1.6
16:21:58  <CIA-99>node: - Add support for os/cpu fields in package.json (Adam Blackburn)
16:21:59  <CIA-99>node: - Automatically node-gyp packages containing a binding.gyp
16:21:59  <CIA-99>node: - Fix failures unpacking in UNC shares
16:25:00  * orlandovftwjoined
16:33:03  * pieternquit (Quit: pietern)
16:34:08  * pieternjoined
16:34:21  * brsonjoined
16:36:27  * AndreasMadsenjoined
16:36:40  * travis-cijoined
16:36:40  <travis-ci>[travis-ci] joyent/node#582 (v0.7.6-release - 4ec3bc5 : isaacs): The build failed.
16:36:40  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/compare/e4fc2cb^...4ec3bc5
16:36:40  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/856229
16:36:40  * travis-cipart
16:38:44  <AndreasMadsen>issacs, no because I like to be a pain, but these cluster patches ... when do you plan to review them?
16:38:51  <isaacs>AndreasMadsen: this week.
16:38:57  <isaacs>AndreasMadsen: i hope to get them into 0.7.7
16:39:04  <isaacs>AndreasMadsen: please be a pain. it's appreciated :)
16:39:12  <AndreasMadsen>isaacs: thanks
16:39:41  <isaacs>AndreasMadsen: i had planned to get to them for 0.7.6, but didn't want to delay the release any longer, and got very distracted by the npm security issue
16:40:18  <AndreasMadsen>isaacs: I understand - just to be a pain did you agree on something with -> https://github.com/joyent/node/pull/2650/files
16:40:51  <isaacs>AndreasMadsen: yes, it should be an 'emit' not a throw.
16:41:13  <isaacs>AndreasMadsen: bnoordhuis is a bit stricter on this point. and surely, it woudl be nice to be strict like that, but for now, we cannot ensure that.
16:41:29  <isaacs>AndreasMadsen: emitting an error throws if it's not handled anyway.
16:42:10  <AndreasMadsen>isaacs: I also vote for emit for that reason.
16:44:33  <AvianFlu>holy crap, down to only 2 failed tests on linux! https://gist.github.com/0d122fc33c238b2994c0
16:44:37  <AvianFlu>(debian 6)
16:52:00  * dshaw_quit (Quit: Leaving.)
16:52:37  <AndreasMadsen>AvianFlu: I'm down to one error on mac, nice job guys.
16:54:23  * igorzijoined
16:59:07  * mikealquit (Quit: Leaving.)
17:03:47  <isaacs>found an npm bug.
17:03:56  <isaacs>one of these days, i'll do it perfectly, and rc0 will be the release :)
17:07:10  <AndreasMadsen>isaacs: perfection is an illusion
17:07:13  <AvianFlu><troll> that's gonna be the same day you release that ignorefile lib, I bet! </troll>
17:07:32  <isaacs>AvianFlu: hahah
17:07:35  <isaacs>that's coming, dude, i swear
17:07:46  <isaacs>it took me 2 years to actually deliver a working js glob implementation, but i did it.
17:07:58  <AvianFlu>it's cool, I found the funniest workaround ever - I just tell people to .npmignore their .gitignore
17:08:20  <isaacs>you know you don't have to do that any moer.
17:08:29  <isaacs>any .npmignore (even a blank one) will keep npm from reading .gitignore
17:08:34  <AvianFlu>I'm aware
17:08:40  <AvianFlu>but it's more fun and easier to remember :D
17:08:42  <isaacs>and .gitignore is never put into a package (it's renamed to .npmignore if present)
17:09:03  <AvianFlu>that last part I hadn't noticed, but cool
17:09:51  * isaacstopic: test, please: http://nodejs.org/dist/v0.7.6/node-v0.7.6-RC1.tar.gz
17:12:09  <isaacs>also, with this release: `npm install jsdom` Just Works, on Windows and everything
17:12:37  <isaacs>Let the binding.gyp pull requests begin!
17:14:10  * orlandovftwquit (Ping timeout: 260 seconds)
17:14:42  * kuebkjoined
17:19:16  <AvianFlu>awesome
17:27:04  * dshaw_joined
17:33:03  <piscisaureus_>bnoordhuis: isaacs: igorzi: so the call was moved to 7pm CET (in 30 minutes)
17:33:09  <piscisaureus_>?
17:33:14  <isaacs>yeah
17:33:32  <isaacs>piscisaureus_: daylight savings time
17:33:59  <piscisaureus_>huh
17:34:50  <mmalecki>piscisaureus_: usa is in dst and Europe isn't
17:35:00  <piscisaureus_>europe will be in 3 weeks I think
17:35:05  <piscisaureus_>last sunday of march
17:35:07  <mmalecki>which constantly confuses me, because standup happens hour before
17:35:18  <mmalecki>s/before/earlier/
17:36:33  <creationix>yeah, I was an hour early to cloud9 standup yesterday very confused
17:37:01  <piscisaureus_>creationix: well you can sleep an hour longer for the next couple of weeks
17:37:02  <piscisaureus_>win!
17:38:03  <creationix>not if I still want to get off at the same time in the afternoon
17:38:38  <mmalecki>timezones have been always too complicated for me
17:49:48  * AndreasMadsenquit (Remote host closed the connection)
18:00:33  * TooTallNatejoined
18:03:32  * orlandovftwjoined
18:03:45  * wankdankerjoined
18:04:16  <isaacs>skype?
18:04:24  <isaacs>bnoordhuis, igorzi, piscisaureus_
18:05:44  <bnoordhuis>isaacs: online
18:09:39  <TooTallNate>mjr_: my ocd kicked in and i had to do an overhaul of repl, tty, and readline
18:09:50  <TooTallNate>mjr_: https://github.com/joyent/node/pull/2919 and https://github.com/joyent/node/pull/2922
18:09:56  <TooTallNate>i meant to /cc you on those
18:12:24  <bnoordhuis>piscisaureus_: why is gyp + msbuild complaining that setup_env.bat is missing?
18:13:07  <bnoordhuis>(this is for smjs btw) i don't see that file anywhere, it's not in node/ or gyp/
18:16:08  * `3rdEdenjoined
18:16:23  * kuebkquit
18:17:27  * sh1mmerquit (Quit: sh1mmer)
18:18:14  <wankdanker>Linux 3.0.0-16-generic - node-v0.7.6-RC1.tar.gz - [02:02|% 100|+ 372|- 2] (test-debugger-repl and test-debugger-repl-utf8 failed)
18:18:49  <benvie>TooTallNate: yeah I went through the same process
18:18:57  <benvie>TooTallNate: that area is rich for refactoring
18:23:30  <bnoordhuis>piscisaureus_: nm, found it. apparently gyp expects a cygwin shell if your gyp file defines actions
18:23:37  <TooTallNate>wankdanker: i think that one is flaky, it failed for me too
18:23:42  <TooTallNate>but when i run it manually it works fine
18:24:15  <TooTallNate>benvie: well hopefully by v0.8 well be able to make sweet repls over all kinds of sockets :)
18:26:09  <isaacs>https://gist.github.com/2030486
18:26:11  <isaacs>piscisaureus_: ^
18:26:15  <isaacs>bnoordhuis: ^
18:26:52  <benvie>once your done with that refactoring I have some stuff I'll update to work with it and send some pull requests
18:28:03  <benvie>let making the keybindgs not hard coded for starters
18:28:38  <benvie>er like
18:28:55  <TooTallNate>benvie: RE the window size thing
18:29:04  <TooTallNate>now readline has a default when none is specified
18:29:06  <TooTallNate>80x30
18:29:19  <benvie>yeah, I was thinking about
18:29:23  <TooTallNate>but it checks for "stream.rows" and "stream.columns"
18:29:35  <benvie>isn't there a control protocol as part of ansi escapes?
18:29:39  <TooTallNate>so if you have a proper telnet server that knows the client's screen size, you could set those attributes
18:29:45  <benvie>I know tty.js has stuff for that
18:29:58  <TooTallNate>control like what?
18:30:11  <benvie>like you can send an escape sequence that indicates what the window size is
18:30:22  <benvie>so it's sent formatted correctly from server to client
18:30:34  <TooTallNate>that's what telnet is for
18:30:55  <benvie>I guess that's more a whole separate module then
18:30:56  <TooTallNate>specifically NAWS
18:31:01  <TooTallNate>indeed
18:31:17  <TooTallNate>http://www.faqs.org/rfcs/rfc1073.html
18:31:21  <benvie>well
18:31:30  <benvie>at least it's possible to build that on top of this more easily now!
18:31:35  <TooTallNate>exactly
18:31:43  <benvie>quite a bit more easily
18:32:02  <TooTallNate>telnet also has commands for putting the client in raw mode and back
18:32:21  <benvie>before it wasn't really a natural extension to add that but it definitely seems so now
18:32:22  <TooTallNate>so really telnet is the "prime" candidate for this new api
18:32:50  * TooTallNatesad there's no good telnet server for node :(
18:33:04  <benvie>that'll be high on my list of things to do if it doesn't get done otherwise, because it's key to my 256 color aspirations on windows
18:33:27  * mikealjoined
18:33:59  <benvie>the importance of 256 colors can't be understated
18:34:01  <TooTallNate>what terminal can do 256 on windows?
18:34:11  <TooTallNate>i wanna see yoshi on windows :p
18:34:13  <benvie>mintty, putty
18:34:18  * indutnychanged nick to indutny_sleeping
18:34:44  <TooTallNate>i could never figure out how to make putty open a local cmd
18:34:47  <benvie>and a bit further down my list is to make a modified compilation of putty set up to be a node module
18:35:27  <benvie>stripped down to just the essentials to use it as a tty interface basically
18:35:53  <tjfontaine>aside from tty.js being slightly averse to the multi user concept, I'm quite happy with it for communicating to our legacy vt220 app
18:35:58  <benvie>maybe not putty but probably putty since there doesn't seem to be a ton of alternatives there
18:36:21  <benvie>and mintty has too much reliance on the mingw stack
18:36:28  * hij1nx_joined
18:36:44  * hij1nx_changed nick to hij1nx
18:38:30  <benvie>oh hey
18:38:43  <benvie>tj had the same thought as much, but for tty.js
18:38:46  <benvie>on the custom keybindings
18:39:06  <tjfontaine>benvie: ya, absolutely required for my use case
18:40:13  <benvie>over time starting with these changes I can migrate the stuff I did for my custom readline/repl to node core hopefully. Just the basic stuff, but I think keybindings is one of those
18:40:35  <benvie>because otherwise you have to redo the whole module if you want to change any of them
18:41:09  <benvie>well not the whole module, but the large chunk that is ttyWrite
18:42:17  <piscisaureus_>TooTallNate: you can remove the isTTY check from readline as well - but take care to handle undefined width/height properly
18:42:48  <TooTallNate>piscisaureus_: yes, there's already defaults for that
18:43:04  <TooTallNate>piscisaureus_: should I add a 4th argument for "enabled"?
18:43:10  <TooTallNate>to readline.Interface
18:44:30  <piscisaureus_>piscisaureus_: what does "enabled" do?
18:44:50  * `3rdEdenquit (Quit: Leaving...)
18:44:50  <TooTallNate>piscisaureus_: readline has 2 modes, normal, and "enabled" (could use a better name)
18:45:12  <TooTallNate>piscisaureus_: normal is line-buffered, "enabled" is with all the ansi tty bells and whistles
18:45:19  <piscisaureus_>ah, right
18:45:34  <TooTallNate>normal mode is also required for NODE_NO_READLINE, for people who use rlwrap
18:45:57  <TooTallNate>(but really, they shouldn't use that, cause then you don't get tab completion :) )
18:46:27  <piscisaureus_>TooTallNate: do it if you really care. But 4 arguments is close to the limit, you should probably move towards an options object then
18:46:47  <TooTallNate>piscisaureus_: i agree, that's what i recommended in the pull request
18:46:56  <TooTallNate>but figured that belonged in another pull
18:47:20  <piscisaureus_>TooTallNate: if you sqash the last 2 commits from 2919 I think we can just land that.
18:47:28  <TooTallNate>ok one sec
18:48:13  <TooTallNate>piscisaureus_: but the bottom example here is the "desired" API https://github.com/joyent/node/pull/2922#issuecomment-4468777
18:48:57  <piscisaureus_>TooTallNate: yeah that definitely looks better
18:49:19  <piscisaureus_>TooTallNate: although I'd like to support the old api too for a while - in a deprecated fashion
18:49:37  <TooTallNate>piscisaureus_: that shouldn't be a problem
18:50:12  <mjr_>TooTallNate: I thoroughly support your OCD if it gets us better net repls.
18:51:20  <TooTallNate>:D
18:51:28  <bnoordhuis>piscisaureus_: i get a lot of warnings like this: definition of dllimport function not allowed
18:51:35  <mmalecki>isaacs: hey, it's possible that you have a regression in npm
18:51:54  <piscisaureus_>bnoordhuis: can you paste the full error?
18:52:05  <bnoordhuis>piscisaureus_: it's exactly that
18:52:17  <bnoordhuis>and here's the code that triggers it: #define MOZ_IMPORT_DATA(x) __declspec(dllimport) x
18:52:33  <mmalecki>isaacs: when we have a bundled package add git+ssh url in dependencies, it tries to install from git+ssh
18:52:40  <piscisaureus_>bnoordhuis: ah monkeyluv
18:52:46  <mmalecki>isaacs: even tho the package is in dependencies
18:52:52  <mmalecki>er, node_modules
18:52:55  <bnoordhuis>yep
18:53:00  <isaacs>mmalecki: are you running "npm install" manually?
18:53:06  <piscisaureus_>bnoordhuis: #define MOZ_IMPORT_DATA(x) __declspec(dllimport) is probably wrong
18:53:14  <piscisaureus_>bnoordhuis: should be #define MOZ_IMPORT_DATA(x) __declspec(dllexport)
18:53:16  <isaacs>mmalecki: or you mean as a dep of dep?
18:53:23  <mmalecki>isaacs: yeah, manually
18:53:27  <piscisaureus_>bnoordhuis: (that is probably what is intended)
18:53:33  <mmalecki>isaacs: sec, I'll pm you the package.json
18:53:40  <piscisaureus_>bnoordhuis: so probably you are missetting a #define somewhere
18:54:04  <isaacs>k
18:54:13  <tjfontaine>glorious irc colors, I thought piscisaureus_ was sending blank lines :)
18:54:23  <piscisaureus_>oh, did I use colors
18:54:31  <bnoordhuis>piscisaureus_: not me, spidermonkey
18:54:35  <bnoordhuis>i haven't touched any of that code
18:54:58  <piscisaureus_>bnoordhuis: you probably have to set a define like MOZ_BUILDING_SHARED_LIBRARY in the gyp file
18:55:12  <piscisaureus_>bnoordhuis: can you point me to the file that contains that MOZ_IMPORT_DATA declaration?
18:55:25  <bnoordhuis>piscisaureus_: it's in mfbt/Types.h
18:55:46  <bnoordhuis>there's a MOZ_EXPORT_DATA macro too btw, uses dllexport
18:56:07  <TooTallNate>piscisaureus_: ok https://github.com/joyent/node/pull/2919 is updated
18:57:35  <bnoordhuis>piscisaureus_: https://github.com/bnoordhuis/smjs/compare/windows <- here's what i have so far
18:57:38  <piscisaureus_>bnoordhuis: add a 'IMPL_MFBT=1' define to gyp
18:59:42  <CIA-99>node: Nathan Rajlich master * r884a209 / doc/api/repl.markdown : docs: document the repl "exit" event (+7 more commits...) - http://git.io/n6bdBQ
19:00:52  <piscisaureus_>TooTallNate: there ya go
19:01:07  <TooTallNate>piscisaureus_: beautiful, thanks!
19:01:33  <piscisaureus_>TooTallNate: thanks yourself. Now you gotta finish the job :-)
19:01:59  <TooTallNate>so basically, removing the isTTY from readline?
19:02:10  <TooTallNate>anything else that other one needs?
19:02:31  <piscisaureus_>TooTallNate: isTTY has to go.
19:02:43  <TooTallNate>piscisaureus_: also hold off on https://github.com/joyent/node/pull/2907, i think that problem is that i wasn't invoking gyp with -G msvs_version=2010
19:02:51  <piscisaureus_>TooTallNate: and make "enabled" an option
19:02:59  <TooTallNate>piscisaureus_: ok sounds good
19:03:11  <TooTallNate>piscisaureus_: turn to an Options object on this pull or a new one?
19:03:22  <piscisaureus_>TooTallNate: I don't mind :_)
19:03:52  <TooTallNate>ok cool, that'll be ready in a lil bit
19:04:17  <TooTallNate>gotta do some real work first :D
19:04:19  * `3rdEdenjoined
19:04:24  <piscisaureus_>TooTallNate: I still have to review the patches so it's possibly I'll have some comments as well.
19:04:27  <piscisaureus_>Will do that now.
19:04:50  <piscisaureus_>TooTallNate: what is the real work you are doing?
19:04:59  <TooTallNate>LearnBoost
19:05:12  <piscisaureus_>ah, right
19:05:14  <TooTallNate>just some debugging right now, final touches on a new feature
19:05:39  <isaacs>mmalecki: so, what behavior do you see when you run `npm install`
19:05:49  <isaacs>mmalecki: i see it download the git url, but not actually install over the existing copy.
19:07:10  <mmalecki>isaacs: pm, again :)
19:09:55  <bnoordhuis>piscisaureus_: didn't fix it
19:10:05  <bnoordhuis>and fsck it, i'll leave it up to you to fix the windows build :)
19:11:18  * mikealquit (Quit: Leaving.)
19:13:15  <piscisaureus_>bnoordhuis: should I clone from you or from tim?
19:14:26  * travis-cijoined
19:14:27  <travis-ci>[travis-ci] joyent/node#583 (master - 884a209 : Nathan Rajlich): The build is still failing.
19:14:27  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/compare/bcb0cc0...884a209
19:14:27  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/857386
19:14:27  * travis-cipart
19:16:19  <bnoordhuis>piscisaureus_: from me
19:25:18  <piscisaureus_>bnoordhuis: "D:\smjs\.\setup_env.bat" wordt niet herkend als een interne
19:25:18  <piscisaureus_>of externe opdracht, programma of batchbestand.
19:25:23  <piscisaureus_>bnoordhuis: did you get around that?
19:27:34  <piscisaureus_>bnoordhuis: hmm, nvm, it was the cygwin flag
19:28:38  <piscisaureus_>d:\smjs\js\src\jsapi.h(47): fatal error C1083: Cannot open include file: 'mozil
19:28:38  <piscisaureus_>la/StandardInteger.h': No such file or directory [D:\smjs\smjs.vcxproj]
19:30:04  <piscisaureus_>bnoordhuis: are you sure you committed all your changes?
19:30:10  <bnoordhuis>piscisaureus_: yes
19:30:15  <bnoordhuis>do symlinks work on windows?
19:30:21  <piscisaureus_>bnoordhuis: mfbt is not on the include path
19:30:42  <bnoordhuis>piscisaureus_: it is indirectly
19:30:51  <bnoordhuis>i'll explain but let's move over to #luvmonkey
19:30:55  <piscisaureus_>kk
19:36:00  * `3rdEdenquit (Quit: Leaving...)
19:47:05  * `3rdEdenjoined
20:03:33  <CIA-99>node: Artur Adib v0.7.6-release * r1cecfee / (3 files in 2 dirs): Doc highlight fix - http://git.io/ITpdZQ
20:03:33  <CIA-99>node: Trent Mick v0.7.6-release * r1cdadb1 / tools/installer.js : issue #2697: have 'make install' set npm shebang to use its node - http://git.io/xYllXA
20:03:33  <CIA-99>node: isaacs v0.7.6-release * r8b79a31 / tools/installer.js : shebang should make file executable as well - http://git.io/zMrVTw
20:03:34  <CIA-99>node: isaacs v0.7.6-release * r10dc129 / (181 files in 26 dirs): Update npm to 1.1.8 - http://git.io/8_2a2A
20:03:35  <CIA-99>node: isaacs v0.7.6-release * rf06abda / (7 files in 6 dirs): (log message trimmed)
20:03:35  <CIA-99>node: 2012.03.13, Version 0.7.6 (unstable)
20:03:35  <CIA-99>node: * Upgrade v8 to 3.9.17
20:03:35  <CIA-99>node: * Upgrade npm to 1.1.8
20:03:35  <CIA-99>node: - Add support for os/cpu fields in package.json (Adam Blackburn)
20:03:35  <CIA-99>node: - Automatically node-gyp packages containing a binding.gyp
20:03:36  <CIA-99>node: - Fix failures unpacking in UNC shares
20:06:49  <TooTallNate>geez, a lot of stuff in this release, eh?
20:13:31  * `3rdEdenquit (Quit: Leaving...)
20:13:47  <bnoordhuis>it's been a while since the last one
20:18:02  <isaacs>yeah
20:18:19  <isaacs>i'll try to get bac on the weekly or biweekly release schedule.
20:18:30  <isaacs>check this out: https://github.com/mozilla/pdf.js/pull/1320
20:18:40  <isaacs>bnoordhuis, piscisaureus_: check out the comments the bot posts.
20:18:50  * travis-cijoined
20:18:50  <travis-ci>[travis-ci] joyent/node#584 (v0.7.6-release - f06abda : isaacs): The build is still failing.
20:18:50  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/compare/4ec3bc5...f06abda
20:18:50  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/857967
20:18:50  * travis-cipart
20:19:30  * isaacstopic: test, please: http://nodejs.org/dist/v0.7.6/node-v0.7.6-RC2.tar.gz
20:20:22  <bnoordhuis>isaacs: nifty
20:22:52  <TooTallNate>isaacs: if a module has both a wscript and a binding.gyp, does the gyp stuff take precendence?
20:23:29  <isaacs>TooTallNate: yesskr
20:23:34  <TooTallNate>cool
20:24:35  * AndreasMadsenjoined
20:25:03  * igorziquit (Ping timeout: 245 seconds)
20:29:13  <piscisaureus_>isaacs: cool. want
20:40:45  * igorzijoined
20:51:26  <isaacs>ok, last chance: anything that ought to be in https://github.com/joyent/node/commit/v0.7.6-release and is missing?
20:51:53  <isaacs>the mac binary fatso is building, then i'm gonna drop the email etc.
20:52:09  * isaacstopic: test, please: http://nodejs.org/dist/v0.7.6/node-v0.7.6.tar.gz
21:02:36  <piscisaureus_>isaacs: lgtm
21:03:12  <isaacs>sweet
21:04:40  <CIA-99>node: isaacs master * r73c8746 / (387 files in 51 dirs): Merge branch 'v0.7.6-release' - http://git.io/JbX7RQ
21:04:41  <CIA-99>node: isaacs master * rbda08bf / src/node_version.h : Now working on 0.7.7 - http://git.io/dqydLw
21:04:43  <CIA-99>node: isaacs master * rf06abda / (7 files in 6 dirs): (log message trimmed)
21:04:43  <CIA-99>node: 2012.03.13, Version 0.7.6 (unstable)
21:04:43  <CIA-99>node: * Upgrade v8 to 3.9.17
21:04:43  <CIA-99>node: * Upgrade npm to 1.1.8
21:04:43  <CIA-99>node: - Add support for os/cpu fields in package.json (Adam Blackburn)
21:04:43  <CIA-99>node: - Automatically node-gyp packages containing a binding.gyp
21:04:44  <CIA-99>node: - Fix failures unpacking in UNC shares
21:04:44  <CIA-99>node: isaacs master * r10dc129 / (181 files in 26 dirs): Update npm to 1.1.8 - http://git.io/8_2a2A
21:05:26  <TooTallNate>14 new contributors!!
21:05:32  <TooTallNate>that's a lot
21:06:31  <mmalecki>TooTallNate: ++, you're rocking this release!
21:06:31  <kohai>TooTallNate has 9 beers
21:06:40  <TooTallNate>mmalecki: thanks :D
21:06:43  <TooTallNate>mmalecki++
21:06:43  <kohai>mmalecki has 70 unicorns
21:07:13  <mmalecki>:)
21:07:37  <TooTallNate>hrtime() is pretty badass
21:08:05  <TooTallNate>i'm curious to see interesting things done with it other than super-precise benchmarks
21:08:07  <mmalecki>indeed!
21:10:04  <isaacs>yeah, this is The Too Tall Release, for sure.
21:14:35  <ryah>isaacs: the link for x64 windows files is wrong on the blog post
21:14:51  <AndreasMadsen>TooTallNate: ++ for awesome work. Hopefully 0.7.7 is going to be my release - I have 5 pull request wanting to be merged.
21:14:51  <kohai>TooTallNate has 10 beers
21:15:13  <TooTallNate>thanks guys <3
21:16:31  * mmaleckideploys to Travis
21:18:16  <TooTallNate>and the wave of new twitter followers begins :p
21:18:17  <TooTallNate>thanks isaacs
21:19:49  <isaacs>ryah: d'oh. thanks. fixed in the email, but forgot the re-generate the blog.html file. fixed now.
21:20:46  <isaacs>AndreasMadsen: How many "steps" do you think are left in cluster2?
21:21:04  <AndreasMadsen>isaacs: non or one, depending on a dissussio
21:21:05  <AndreasMadsen>n
21:21:29  <isaacs>AndreasMadsen: ok, col
21:21:31  <isaacs>*cool
21:22:41  <isaacs>TooTallNate: I think i asked before, but i forget now, what happens when you do `node-gyp yzxyzyx` and you don't have the MSVS whatevers in the right place?
21:22:50  <isaacs>TooTallNate: does it fetch them for you, or just abort in fury?
21:23:35  <TooTallNate>it just fails for now, i'm gonna update it to open up a web page or something
21:24:18  * travis-cijoined
21:24:18  <travis-ci>[travis-ci] joyent/node#585 (master - bda08bf : isaacs): The build is still failing.
21:24:18  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/compare/884a209...bda08bf
21:24:18  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/858541
21:24:18  * travis-cipart
21:24:42  <isaacs>TooTallNate: rad.
21:24:58  <isaacs>TooTallNate: so, we're in this weird state now, where npm bundles node-gyp and node bundles npm..
21:25:11  <isaacs>TooTallNate: to make npm use your dev copy of node-gyp, you can do this:
21:25:33  <isaacs>npm explore npm -g -- npm link ~/path/to/node-gyp
21:25:36  <TooTallNate>hahaha, recursive bundling
21:25:54  <TooTallNate>isaacs: ok thanks, i'll have to remember that :D
21:26:11  <isaacs>TooTallNate: it sets up the PATH env so that you are guaranteed to get the one it provides.
21:26:12  <TooTallNate>isaacs: the bundled node-gyp only kicks in during 'install'?
21:26:17  <isaacs>TooTallNate: yes.
21:26:22  <TooTallNate>ok makes sense
21:26:33  <isaacs>TooTallNate: but any `node-gyp` command will use the bundled one. lib/utils/lifecycle.js forces the PATH that way.
21:26:58  <TooTallNate>isaacs: you mean after using the installer?
21:26:58  <isaacs>and (for now at least, maybe temporarily) it rewrites any blahblahb.gyp to binding.gyp
21:27:04  <isaacs>TooTallNate: yeah
21:27:35  <TooTallNate>isaacs: ya i'm forcing binding.gyp for now but it's totally artificial, we could lift that restriction
21:27:50  <isaacs>TooTallNate: https://github.com/isaacs/npm/blob/master/lib/utils/lifecycle.js#L68-75
21:28:00  <isaacs>TooTallNate: or we could force it to be something else.
21:28:35  <isaacs>TooTallNate: ideally, node-gyp will be basically a subcomponent of npm, and mostly just used by the build server to pre-generate the binaries
21:28:49  <isaacs>but like, several parts of that sentence don't exist yet.
21:28:50  <isaacs>:)
21:29:41  <TooTallNate>right, it's probably ok for now, but eventually we may want only npm to use it internally, and not be in the resulting PATH
21:30:02  <TooTallNate>isaacs: what about any bugs in the current bundled version? can users still update it through npm?
21:30:18  <TooTallNate>%s/bugs/new features/
21:30:36  <isaacs>TooTallNate: for now, it just is what it is :)
21:30:46  <isaacs>TooTallNate: but... yeah..
21:30:57  <isaacs>we should work out an easy way to update it at the top level
21:31:14  <isaacs>TooTallNate: you could install a new one with: npm explore npm -g -- npm install node-gyp
21:31:26  <isaacs>TooTallNate: that'd install node-gyp into npm's node_modules folder
21:31:40  <TooTallNate>that's easy enough to tell people for now :)
21:40:35  <TooTallNate>bnoordhuis: you should school this guy https://github.com/TooTallNate/node.bcrypt.js/commit/59b5c2980b65ada54bd988ba05755223579dabeb
21:41:55  <mjr_>Man, crypto people are so sure of themselves.
21:42:38  <mjr_>You'd think that of all people, they'd understand the limitations of their understanding, and the value of tradeoffs.
21:45:22  <bnoordhuis>TooTallNate: he's right in that #include <foo.h> will first search the user-supplied include path for foo.h
21:46:06  <bnoordhuis>the problems start when you want to include a specific version and instead get a system-wide one
21:46:09  <TooTallNate>right, so the change is purely cosmetic?
21:46:49  <TooTallNate>bnoordhuis: i mean that's what's happening here; it's picking up the system openssl rather than the bundled one
21:47:48  <bnoordhuis>TooTallNate: that shouldn't happen if -I/path/to/node/deps/openssl/openssl/include/ is specified
21:48:22  <bnoordhuis>-I comes before system headers
21:48:32  <TooTallNate>oh ok, so then it was cosmetic then...
21:48:35  <isaacs>C++ has a crappy module system.
21:48:35  <kohai>C has 10 beers
21:48:37  <tjfontaine>unless the configure/build finds the system and puts it in -I first?
21:49:06  <bnoordhuis>tjfontaine: yeah... it's a problem on freebsd for example because we -I/usr/local/include there
21:49:53  <TooTallNate>bnoordhuis: i think the problem was coming from node-waf anyways
21:50:06  <TooTallNate>i think gyp does the right thing regardless
21:50:49  <bnoordhuis>TooTallNate: re cosmetic: the #include "node_foo.h" change was cosmetic yes because the odds of node's headers being installed system-wide are slim
21:50:55  <bnoordhuis>zlib.h is another story though
21:51:33  <tjfontaine>TooTallNate: https://github.com/TooTallNate/node.bcrypt.js/blob/master/wscript#L23 seems quite likely to fubar with the include path
21:54:01  <TooTallNate>tjfontaine: good find!
21:54:17  <TooTallNate>bnoordhuis: well we do #inlcude "zlib.h" in node_zlib.cc
21:54:23  <TooTallNate>bnoordhuis: what about node_crypto though?
21:54:35  <TooTallNate>still using #include <openssl/rand.h>
21:54:52  <bnoordhuis>right, that should be fixed in due time
21:55:04  <TooTallNate>ok
21:55:34  <TooTallNate>i think what's hard about this particular module is you don't know whether or not node was built with the bundled openssl or the system one :\
21:56:02  <TooTallNate>in the binding.gyp file I'm assuming it was bundled, and use those headers, but that's not necessarily correct
21:56:09  <bnoordhuis>we can add a define for that in the gyp file
21:56:14  <bnoordhuis>HAVE_SYSTEM_OPENSSL=1 or something
21:56:35  <bnoordhuis>though that may not help node-gyp
21:56:44  <TooTallNate>bnoordhuis: but since it's specific to the version of node that is running, it would need to be exposed in node somewhere
21:57:10  <TooTallNate>i proposed `process.shared.zlib === false` the other day
21:57:19  <TooTallNate>probably a better API could be thought of
21:58:35  * hij1nxquit (Quit: hij1nx)
22:02:13  * AndreasMadsenquit (Remote host closed the connection)
22:02:26  * rendarquit
22:05:01  * isaacsquit (Read error: Connection reset by peer)
22:05:36  * isaacsjoined
22:06:55  <isaacs>i like process.shared.zlib === false
22:07:14  <isaacs>in general, stuff relating to the way that node was built should be on the process object, i think
22:07:34  <TooTallNate>isaacs: thanks, i can do it, should be a small one
22:15:52  <TooTallNate>there's no way to link node to a shared libuv right?
22:17:44  <isaacs>no
22:18:01  <isaacs>speaking of which, we should probably update openssl
22:18:07  <isaacs>but the prospect scares me a little bit
22:19:34  * paddybyersquit (Quit: paddybyers)
22:23:44  * AvianFluquit (Read error: Connection reset by peer)
22:24:15  * AvianFlujoined
22:36:30  * paddybyersjoined
22:37:08  <TooTallNate>isaacs: does `npm explore` use readline?
22:37:17  <isaacs>no
22:37:31  <isaacs>TooTallNate: it just opens a subshell
22:37:33  <TooTallNate>how do you do the ^D thing?
22:37:41  <TooTallNate>or is that part of bash?
22:37:41  <isaacs>a bash child process
22:37:47  <isaacs>bash uses gnu readline
22:37:53  <isaacs>(i presume)
22:38:12  <TooTallNate>ok, so you're not the one handling the ^D then, ok just curious
22:46:18  <bnoordhuis>isaacs: i'll update openssl some time next week
22:46:32  <isaacs>bnoordhuis: thanks.
22:46:35  <isaacs>TooTallNate: this is handy: https://github.com/isaacs/dotfiles/blob/master/.extra.bashrc#L527-529
22:46:43  <bnoordhuis>it's been on my mental TODO list for weeks
22:47:09  <bnoordhuis>hence i don't feel bad for putting it off for another week
22:47:14  <TooTallNate>isaacs: what is that?
22:48:05  <isaacs>TooTallNate: i put a . in my shell prompt for each SHLVL i'm in
22:48:15  <isaacs>TooTallNate: so when i'm in a subshell there's a . there
22:49:25  <TooTallNate>oh that's cool
22:49:33  <TooTallNate>so you can see visually when you're in a subshell?
22:49:43  <isaacs>nave puts 2 dots, because it's a subshell called from within a script, which itself is a SHLVL=1
22:49:59  <mmalecki>shellception.
22:53:37  * mrb_bkquit (Ping timeout: 252 seconds)
22:53:51  * Raynosquit (Ping timeout: 260 seconds)
22:54:26  * coderarityquit (Ping timeout: 246 seconds)
23:10:00  * Raynosjoined
23:12:18  * paddybyersquit (Quit: paddybyers)
23:13:12  * sh1mmerjoined
23:20:09  <TooTallNate>yup, rather simple https://github.com/joyent/node/pull/2927
23:43:41  * xaqjoined
23:46:34  * CoverSlidejoined
23:51:01  * igorziquit (Quit: Page closed)
23:56:33  <piscisaureus_>start .