00:02:07  * paddybyersquit (Quit: paddybyers)
00:30:13  <CIA-111>node: Ben Noordhuis v0.6 * r6df7bdd / (lib/child_process.js test/simple/test-child-process-fork.js):
00:30:13  <CIA-111>node: child_process: make .send() throw if message is undefined
00:30:13  <CIA-111>node: JSON.stringify(undefined) returns "undefined" but JSON.parse() doesn't know how
00:30:13  <CIA-111>node: to parse that. - http://git.io/MJlyPg
00:37:48  * travis-cijoined
00:37:48  <travis-ci>[travis-ci] joyent/node#88 (v0.6 - 6df7bdd : Ben Noordhuis): The build passed.
00:37:48  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/compare/c4d2244...6df7bdd
00:37:48  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/418144
00:37:48  * travis-cipart
00:50:48  <CIA-111>node: Yoshihiro Kikuchi v0.6 * rf2f3028 / (lib/timers.js test/simple/test-timers-zero-timeout.js):
00:50:49  <CIA-111>node: timers: fix memory leak in setTimeout
00:50:49  <CIA-111>node: Closing handle is leaked when setTimeout called with arguments which are
00:50:49  <CIA-111>node: 1. a callback
00:50:49  <CIA-111>node: 2. zero delay
00:50:49  <CIA-111>node: (i.e. setTimeout(function(){}, 0); ) - http://git.io/Jpk6NA
00:58:33  * travis-cijoined
00:58:33  <travis-ci>[travis-ci] joyent/node#89 (v0.6 - f2f3028 : Yoshihiro Kikuchi): The build passed.
00:58:33  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/compare/6df7bdd...f2f3028
00:58:33  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/418173
00:58:33  * travis-cipart
02:15:44  * bnoordhuisquit (Ping timeout: 240 seconds)
02:22:23  * brsonquit (Quit: leaving)
02:30:29  <pquerna>http://journal.paul.querna.org/articles/2011/12/17/technology-cloud-monitoring/
02:30:33  <pquerna>in post 1 of 500
02:33:08  * mraleph1joined
02:33:09  * mralephquit (Read error: Connection reset by peer)
02:39:18  * mralephjoined
02:39:18  * mraleph1quit (Read error: Connection reset by peer)
02:52:16  * mralephquit (Quit: Leaving.)
04:12:03  * ericktjoined
05:07:56  * sh1mmerquit (Quit: sh1mmer)
05:08:30  * sh1mmerjoined
06:33:16  * ericktquit (Quit: erickt)
08:14:35  <ryah>pquerna: nice post
08:16:01  <indutny>ryah: thanks :)
08:16:18  * einarosquit (Ping timeout: 240 seconds)
08:16:19  * einarosjoined
08:17:00  * einarosquit (Ping timeout: 240 seconds)
08:17:15  * einarosjoined
08:40:20  * paddybyersjoined
09:07:10  * mikealjoined
09:35:41  <pquerna>2/500: http://journal.paul.querna.org/articles/2011/12/18/the-switch-python-to-node-js/
09:39:18  * paddybyersquit (Quit: paddybyers)
09:43:12  <indutny>pquerna: nice!
09:43:56  <mmalecki>pquerna: fyi, there's a pretty nice evented python framework: tornado
09:44:08  <mmalecki>pquerna: http://www.tornadoweb.org/
09:44:40  <pquerna>yes, we considered tornado too
09:44:52  <pquerna>but its even a smaller community than gevent and twisted
09:45:33  <mmalecki>yeah, it's pretty painful. I started building my pseudo-startup on top of it but failed - we had to write everything from scratch.
09:46:06  <mmalecki>it has pretty outstanding benchmarks tho
09:48:56  <pquerna>benchmarks don't get users or vc money, most of the time :)
09:50:38  <mmalecki>hah, definitely :). I guess it was my problem
09:51:20  <ryah>wrestling with thread locks now
09:51:23  <ryah>unhappy
09:51:33  <mmalecki>threads :(
09:52:58  <indutny>mmalecki: I've used it for a month or so
09:53:05  <indutny>mmalecki: even built some small app on it
09:53:19  <indutny>mmalecki: but I like node.js much more :)
09:53:24  <indutny>mmalecki: vm is much faster
09:53:39  <mmalecki>indutny: small apps are fine, but big apps using many APIs are a fail, no modules, et.c
09:53:46  <indutny>mmalecki: yeah
09:53:57  <indutny>mmalecki: python package installer is awful too
09:53:59  <indutny>gtg
09:54:08  <mmalecki>indutny: tornado works with PyPy btw. but v8 is still way better.
09:54:59  <indutny>mmalecki: yep, I think so too
09:55:00  <indutny>ttyl
09:55:25  <mmalecki>see you!
10:21:20  * AndreasMadsenjoined
10:21:49  <AndreasMadsen>Sorry I made a mistake in when i configured the master branch there is no problems. I have removed the silent commit, so this pull request is clean. I will base the step 4 pull request on the master branch.
10:21:55  <AndreasMadsen>ryah: ^
10:22:29  * AndreasMadsenquit (Remote host closed the connection)
10:31:57  * paddybyersjoined
10:44:42  * paddybyersquit (Quit: paddybyers)
10:52:01  * mralephjoined
12:31:33  * AndreasMadsenjoined
12:37:08  * felixgejoined
12:37:08  * felixgequit (Changing host)
12:37:08  * felixgejoined
12:37:42  <felixge>https://github.com/felixge/node/commit/3f63c5339d55c058353f091f726aaf173f06bae5
12:37:54  <felixge>anybody objecting against pushing this into master?
12:38:25  <AndreasMadsen>why not v0.6
12:39:37  <felixge>AndreasMadsen: can be 0.6 as well
12:39:40  <felixge>sure
12:40:02  <felixge>also checking why simple/test-pipe.js is racy on OSX
12:40:07  <felixge>sometimes doesn't shut down cleanly
12:40:32  <felixge>checking this on linux now
12:44:54  <AndreasMadsen>Is there a bug in make jslint, try jslinting >> throw { foo: 'bar' }; << :/
12:45:30  <felixge>jslint … why the fuck are we doing that :(
12:46:26  <AndreasMadsen>Apparently we don't: https://github.com/joyent/node/pull/2314
12:46:28  <felixge>well, or even more interesting, why are we pretending to be doing it: Found 1031 errors,
12:46:28  <felixge>:)
12:47:09  <felixge>I need to continue my hacking on syntux: https://github.com/evilhackerdude/syntux
12:47:38  <felixge>instead of being a bitchy little linter, it will allow to transform code to comply with a pre-defined style
12:47:39  <felixge>:)
12:48:28  <felixge>and it's easy to write plugins: https://github.com/evilhackerdude/syntux/blob/master/lib/plugins/quotes.js
12:48:29  <AndreasMadsen>Looks nice, it did take long time to jslint all testcases :)
12:48:57  <felixge>oh I guess that is your patch :)
12:50:57  <AndreasMadsen>One of them, but GitHub count it high, I almost 80% blue now: https://github.com/AndreasMadsen/
12:51:07  <AndreasMadsen>Because of that patch
12:51:49  <AndreasMadsen>Also we should ignore the punycode lib from jslit, there are so many jslint errors and we don't maintain it.
12:53:33  <mmalecki>AndreasMadsen: we should open an issue at punycode and ask them to comply with our jshint settings
12:54:24  <AndreasMadsen>I though that as well, I'm willing to do it if they say yes.
13:01:00  <AndreasMadsen>https://github.com/bestiejs/punycode.js/issues/6
13:02:23  * AndreasMadsenquit (Remote host closed the connection)
13:11:52  * AndreasMadsenjoined
13:12:26  <AndreasMadsen>mmalecki: they have responded: https://github.com/bestiejs/punycode.js/issues/6#issuecomment-3194610 we should ignore it.
13:15:51  <mmalecki>AndreasMadsen: understandable. thanks for opening it :)
13:17:48  <AndreasMadsen>mmalecki: unfortunately I only read Python so I won't make that pull request.
13:19:35  <mmalecki>AndreasMadsen: what do you mean by 'I only read Python'?
13:20:28  <AndreasMadsen>Isn't the node.js linter made in Python, I have not looked intro it so I could be mistaken.
13:21:18  * bradleymeckjoined
13:22:20  <mmalecki>ah, right
13:24:05  * Raynosquit (Read error: Connection reset by peer)
13:27:50  * AndreasMadsenquit (Remote host closed the connection)
13:34:10  * bradleymeckquit (Ping timeout: 240 seconds)
13:59:25  * bnoordhuisjoined
14:39:20  * paddybyersjoined
14:39:30  * Raynosjoined
14:47:37  * indutny_joined
14:47:47  * indutny_quit (Client Quit)
14:49:36  * AndreasMadsenjoined
14:50:49  * indutny_joined
14:51:23  <indutny_>fck, server just hung out
14:51:27  <indutny_>rebooted it
14:51:40  * indutnyquit (Read error: Operation timed out)
14:57:11  * indutny_quit (Quit: leaving)
14:57:17  * indutnyjoined
14:57:23  <indutny>yeah
14:57:42  <AndreasMadsen>What are the current state on #2163 ( https://github.com/joyent/node/pull/2163 )
15:03:12  <CIA-111>node: Felix Geisendörfer v0.6 * rec51bfc / test/simple/test-net-server-listen-remove-callback.js : Use common.PORT to avoid conflicts - http://git.io/s59pWQ
15:10:49  * travis-cijoined
15:10:49  <travis-ci>[travis-ci] joyent/node#90 (v0.6 - ec51bfc : Felix Geisendörfer): The build passed.
15:10:49  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/compare/f2f3028...ec51bfc
15:10:49  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/419502
15:10:49  * travis-cipart
15:25:55  * luxigojoined
15:52:10  <CIA-111>libuv: Ben Noordhuis v0.6 * re9235a3 / src/unix/dl.c : unix: the dl functions don't set errno - http://git.io/-rQAwQ
15:53:30  <CIA-111>libuv: Ben Noordhuis v0.6 * rfeb267e / (include/uv.h src/unix/dl.c):
15:53:31  <CIA-111>libuv: unix: it's legal for dlsym() to return NULL
15:53:31  <CIA-111>libuv: A symbol name can map to NULL. Check dlerror() to see if a real error happened. - http://git.io/psvq-Q
15:53:51  * travis-cijoined
15:53:51  <travis-ci>[travis-ci] joyent/libuv#10 (v0.6 - feb267e : Ben Noordhuis): The build is still failing.
15:53:51  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/d808cf9...feb267e
15:53:51  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/419654
15:53:51  * travis-cipart
15:55:04  <mmalecki>"still failing" makes me a sad panda :(
16:10:25  * felixgequit (Quit: http://www.debuggable.com/)
16:44:30  * ericktjoined
17:20:38  <AndreasMadsen>It it intended that process.platform is 'mac' and not 'darwin' in node v0.7-pre
18:16:59  * luxigoquit (Ping timeout: 240 seconds)
18:18:49  * paddybyersquit (Quit: paddybyers)
18:34:09  * ericktquit (Quit: erickt)
18:42:48  <AndreasMadsen>crisis: the 0.7.0-pre building system don't detect when a file is changed
18:43:11  <AndreasMadsen>^this was not to crisis I'm having a crisis
18:44:46  <AndreasMadsen>also make --assume-new=lib/cluster.js do not change anythinh
18:52:25  <ryah>AndreasMadsen: assume-new?
18:52:51  <AndreasMadsen>ryah: make --help
18:54:37  <ryah>AndreasMadsen: the makefiles should detect change - which they do for me - if they do not then please fix
18:54:59  <AndreasMadsen>ryah: How should i fix it?
18:55:07  <ryah>*shrug*
18:55:08  <ryah>:)
18:55:33  <ryah>AndreasMadsen: makefiles are generatedby tools/gyp_node
18:55:43  <ryah>except the root Makefile
18:56:11  <ryah>AndreasMadsen: when you change lib/cluster it doesn't detect it?
18:56:31  <AndreasMadsen>Right
18:56:49  <AndreasMadsen>But if I do a make clean && make the changes do take effect
18:57:38  <AndreasMadsen>ryah: I only read Python
18:58:08  <ryah>yeah, i also can experience the bug
18:59:23  <AndreasMadsen>ryah: You are confusing me, is that by purpose
18:59:54  <ryah>nope - we should make a ticket for this
19:00:00  <ryah>i can look into it later
19:00:52  <AndreasMadsen>ryah: Thats fine, I have rebased the cluster step 3 commit, but I'm not going to make step 4 or base step 3 on the master before this is fixed.
19:03:08  <AndreasMadsen>ryah: Also is there a API change document from node 0.6 to 0.7
19:04:20  <ryah>AndreasMadsen: not yet
19:13:27  * AndreasM_joined
19:13:45  <AndreasMadsen>ping him ^ if anything :)
19:28:07  <indutny>ryah: hey
19:28:21  <indutny>debugger patches ;)
19:29:46  <indutny>mraleph: nice! http://blog.mrale.ph/post/14403172501/simple-optimization-checklist
19:30:43  <mraleph>indutny: I am glad you liked it.
19:49:39  * txdvquit (Read error: Connection reset by peer)
19:49:59  * txdvjoined
20:01:25  * AndreasM_quit (Remote host closed the connection)
20:03:49  * AndreasM_joined
20:04:53  * paddybyersjoined
20:19:02  <ryah>AndreasMadsen: if you git revert 10d92b3dcd8dbe4e3fe7cce90f7e792fec3364d0
20:19:09  <ryah>AndreasMadsen: it will probably fix the makefile problem
20:22:01  * mikealquit (Quit: Leaving.)
20:23:26  <indutny>mraleph: anyway we shouldn't know too much about v8 internals
20:23:40  <indutny>mraleph: that's definitely a sin for a js developer
20:23:53  <indutny>mraleph: like using browser-specific hacks
20:24:18  <indutny>mraleph: v8 should know about our code, and optimize it well
20:29:29  <mraleph>indutny: I agree that knowing too much about single VM should not be required
20:29:47  <mraleph>but understanding how VMs in general operate is very important.
20:30:04  <mraleph>also understanding algorithms in general
20:34:13  <indutny>mraleph: +1 for last one
20:34:18  <AndreasMadsen>ryah: I'm trying that now
20:34:35  <indutny>mraleph: but understanding vm principals shouldn't be necessary at all
20:34:57  <indutny>mraleph: but, there should always be a right tool for right purpose
20:35:34  <mraleph>ok lets move discussion out of here, it is of topic for #libuv
20:35:36  <indutny>mraleph: so probably you should use rare non-optimizable by current compilers things only with a new compiler ;)
20:35:43  <indutny>mraleph: right, sorry guys
20:49:17  <AndreasMadsen>ryah: It works now thanks
20:50:38  * mikealjoined
21:07:55  * AndreasMadsenquit (Remote host closed the connection)
21:18:34  * AndreasM_quit (Quit: Leaving)
21:36:41  * brsonjoined
21:43:35  * travis-cijoined
21:43:35  <travis-ci>[travis-ci] joyent/node#91 (v0.6 - 4b455ba : Ben Noordhuis): The build passed.
21:43:35  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/compare/ec51bfc...4b455ba
21:43:35  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/420829
21:43:35  * travis-cipart
21:48:54  <bnoordhuis>ryah: are you trying to piss off every admin in the world?
21:49:11  <bnoordhuis>i am of course talking about https://github.com/joyent/node/commit/e61de70c
22:02:27  * sh1mmerquit (Quit: sh1mmer)
22:09:10  <CIA-111>node: Avi Flax master * ree0a7b9 / (lib/util.js test/simple/test-sys.js): util: output Dates with Date.toString not Date.toUTCString - http://git.io/GVhLGA
22:23:16  * travis-cijoined
22:23:16  <travis-ci>[travis-ci] joyent/node#92 (master - ee0a7b9 : Avi Flax): The build is still failing.
22:23:16  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/compare/b084322...ee0a7b9
22:23:16  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/420878
22:23:16  * travis-cipart
22:37:05  * mikealquit (Quit: Leaving.)
23:03:04  <ryah>bnoordhuis: yeah - it's easier to just compile it in
23:04:33  <bnoordhuis>ryah: easier in what respect?
23:09:34  * paddybyersquit (Quit: paddybyers)
23:10:24  <ryah>bnoordhuis: for us
23:11:09  <CIA-111>node: Ryan Dahl isolates2 * rdb6d846 / (5 files in 3 dirs): ... - http://git.io/B5BM9Q
23:11:10  <ryah>we can ensure wpeople hve NPN and SNI
23:11:22  <ryah>dont have to worry about what arch libssl is
23:11:23  <ryah>etc
23:11:32  <ryah>fuck - didn't mean to land that ...
23:11:42  <ryah>oh well
23:12:48  <ryah>bnoordhuis: we really got to get --dest-cpu working...
23:12:54  <ryah>i tried for a bit yesterday but failed
23:13:19  <bnoordhuis>why? it's only relevant to cross-compilers isn't it?
23:14:33  <ryah>what i meant to say is we should get the x64 build working
23:14:47  <ryah>https://gist.github.com/1494780 <-- dest_cpu mostly working
23:15:12  <ryah>well not mostly
23:15:36  <bnoordhuis>mostly not :)
23:16:33  <ryah>yeah :)
23:18:03  <bnoordhuis>x64 builds work for me btw
23:18:13  <bnoordhuis>that is, compiling an x64 build on a x64 host
23:18:15  <ryah>yeah? hm
23:18:34  <bnoordhuis>$ file out/Debug/node
23:18:34  <bnoordhuis>out/Debug/node: ELF 64-bit LSB executable, x86-64, version 1 (SYSV), dynamically linked (uses shared libs), for GNU/Linux 2.6.15, not stripped
23:19:13  <ryah>is host_arch used anywhere? it doesn't detect it correctly for osx
23:19:36  <ryah>yeah seems so...
23:19:37  <ryah>hm
23:22:09  <bnoordhuis>https://github.com/joyent/node/pull/2337 <- windows add-on build scripts, good stuff
23:23:45  <ryah>bash script is not very windows-y
23:23:54  <ryah>oh - he's got a .bat
23:24:09  <ryah>i think we sholdn't ship a "node-gyp"
23:24:15  * travis-cijoined
23:24:15  <travis-ci>[travis-ci] joyent/node#93 (isolates2 - db6d846 : Ryan Dahl): The build is still failing.
23:24:15  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/compare/c04f0f9...db6d846
23:24:15  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/421162
23:24:15  * travis-cipart
23:24:41  <ryah>it would be better if people just generated makefiles and .sln and uploaded them with the project
23:26:18  <ryah>ideally people could just upload binaries...
23:26:27  <ryah>we should think this through
23:26:39  <ryah>rather than give some half solution
23:27:00  <ryah>windows users will not have a compiler
23:27:47  <ryah>it doesn't make sense to ship a node-gyp, gyp, and header files for node installs on windows
23:29:44  <ryah>also - we should think about embedding modules inside the node build
23:29:59  <ryah>or at least have that as a possible feature
23:34:06  <CIA-111>node: Ryan Dahl master * rbb8f734 / Makefile :
23:34:06  <CIA-111>node: Fix make targets so they detect changes
23:34:06  <CIA-111>node: Broken in 10d92b3 - http://git.io/WOHOng
23:37:47  <benvie>I looked into whether it'd even be possible to redistribute the MSVC compiler and linker and whatnot but as far as I could find they don't allow an avenue to get it aside from Visual Studio or at best VC++ Express
23:38:15  <benvie>the next best thing is what you said which is working binaries into module distribution, which was raised a couple days ago here https://github.com/isaacs/npm/issues/1891#issuecomment-3187925
23:38:54  <ryah>been talking with isaacs about it for a while
23:39:24  <benvie>I don't see another option besides requiring visual c++ express to be installed which isn't really an option
23:40:15  <ryah>no, it's not an option
23:40:32  <ryah>or we stop allowing addons
23:40:51  <ryah>which would be nice - but also not very practical
23:41:12  <ryah>i think the only solution is to distribute binaries
23:41:13  <benvie>I think maybe a middle ground would be some kind of developer version/package
23:41:19  <benvie>which is where that node-gyp thing would fit
23:41:48  <benvie>or at least an easy way to find the resources to do that, because it's great if you do have the setup to compile
23:42:25  <ryah>i think we can require people do have the node tree in order to compile modules
23:42:27  <benvie>makes it easy for people to module creators
23:42:32  <benvie>yeah
23:42:34  <ryah>that way we don't have to create a new dist
23:42:59  <benvie>definitely, and if they have that they'll have all the tools at their disposal already
23:44:00  <ryah>just thinking about this problem makes me mad
23:44:12  <ryah>like - how it works currently is so awful
23:44:19  <ryah>mostly because unix is aweful
23:44:40  <ryah>grr..
23:45:08  <ryah>it's amazing how dman complex compiling C code is
23:45:19  <benvie>ha I thought you were going the other way with that, in terms of Windows not being able to compile as a base thing. I had a conversation a while ago about how it's different paradigms
23:45:32  <benvie>people on unix expect to have to compile, it's built in and relatively easy, and it's required
23:45:34  <pquerna>(i don't know if anyone has stats, but it would be interesting to see what % of modules in NPM are compiled, but then have the top 10 list... and just start porting them to pure js)
23:45:42  <benvie>people on windows don't expect to nor have the tools to, but binaries always work
23:45:50  <mmalecki>pquerna: isaacs has some stats on that
23:46:17  <ryah>many of these modules have little bash scripts
23:46:27  <ryah>or - like libxml2 - uses scons
23:46:36  <ryah>all horrible horrible
23:46:55  * travis-cijoined
23:46:55  <travis-ci>[travis-ci] joyent/node#94 (master - bb8f734 : Ryan Dahl): The build is still failing.
23:46:55  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/compare/ee0a7b9...bb8f734
23:46:55  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/421213
23:46:55  * travis-cipart
23:47:02  <pquerna>python seems to do alright in this regards, though people are angry about vairous things in python setuptools/distutils
23:47:04  <benvie>yeah it's funny when looking at porting a module to windows
23:47:20  <benvie>some of them are like all V8 stuff and are no work at all, they just compile no problem
23:47:21  <pquerna>but setup.py pretty much works everywhere, and it was always easy to build an msi from it
23:47:28  <ryah>the only sane options are A) do not allow addons or B) have binary modules in the registry
23:47:29  <benvie>and then others are these horrible messes of platform shit
23:47:38  <mmalecki>pquerna: looks like it's this view. it looks broken tho http://isaacs.ic.ht/_utils/database.html?registry/_design/app/_view/needBuild
23:48:06  <pquerna>mmalecki: get zero :-/
23:48:18  <ryah>we should force uploaded DLLs to not link to external libraries
23:48:24  <benvie>well for npm it's almost there, there just needs to be a way to identify platform specific sub-components like binaries for windows
23:48:24  <mmalecki>pquerna: yeah
23:48:26  <ryah>if we went that route
23:48:43  <pquerna>on windows that seems reasonable.
23:48:43  <mmalecki>pquerna: not sure why, it looks good
23:48:44  <ryah>e.g. node-libxml2 would be forced to come with libxml2
23:48:47  <pquerna>yeah
23:49:21  <benvie>I'm always biased toward the "it just works" route which is why I started using node
23:49:39  <ryah>debian people will complain, of course
23:49:40  <benvie>so I agree with that sentiment
23:49:42  <ryah>but fuck them
23:50:15  <ryah>ok - so we add a script to the node tree for easily building an addon using gyp
23:50:31  <benvie>that's probably one of the most valuable features that node has that I rarely see actually mentioned/advetised
23:50:36  <benvie>it just works
23:50:56  <benvie>in a world of shit that doesn't, it's pretty great
23:52:59  <ryah>the script will run otool -L on the generated addon afterwards - before it's uploaded
23:53:25  <ryah>and check that it only links to a whitelisted set of libraries
23:53:58  <ryah>e.g. libcotherwise error
23:54:07  <mmalecki>ryah: before it's uploaded to npm? I don't think I follow.
23:55:00  <ryah>if we do this right - i think installing binary addons will be quite nice
23:55:11  <ryah>if we do it wrong it will be hell
23:55:32  <ryah>mmalecki: i want to upload .node files to npm reg
23:56:07  <mmalecki>ryah: ok. and you want to stop developers from uploading ones with dynamic libs linked?
23:56:22  <ryah>yeah
23:56:42  <mmalecki>that seems fair. but dev should also be able to only provide source code
23:57:59  <mmalecki>so, if you provide a compiled lib, I should be able to run it ootb