00:00:00  * ircretaryquit (Remote host closed the connection)
00:00:07  * ircretaryjoined
00:01:07  * blobaumjoined
00:09:38  * mikealjoined
00:12:23  * mikolalysenkoquit (Ping timeout: 260 seconds)
00:24:33  * vitorpachecojoined
00:42:16  * mikealquit (Quit: Leaving.)
00:46:26  * jxsonquit (Remote host closed the connection)
00:48:08  * jxsonjoined
00:52:40  * jxsonquit (Remote host closed the connection)
00:56:19  * no9quit (Ping timeout: 264 seconds)
01:09:16  * mikolalysenkojoined
01:19:34  * jcrugzzquit (Quit: Lost terminal)
01:23:39  * jcrugzzjoined
01:44:19  * ralphtheninjaquit (Ping timeout: 264 seconds)
01:47:40  * paul_irishjoined
01:48:31  * yorickquit (Remote host closed the connection)
01:50:21  * thl0joined
01:53:14  <chrisdickinson>thl0: re: the format json stream, you might be interested in https://github.com/chrisdickinson/json-parse-stream
01:53:30  <chrisdickinson>it emits every "node" it encounters, replete with a parent chain
01:53:38  <chrisdickinson>so you can determine how many tabs to insert
01:53:55  <thl0>chrisdickinson: nice!
01:54:19  <thl0>I just do something much simpler i.e. wait til end gets emitted and call JSON.stringify
01:54:24  * chrisdickinsonnods
01:54:49  <thl0>it's just to be a simple cli tool - need it to figure out npm's couch db queries
01:54:55  <mikolalysenko>dumb question: how do I catch when a pipe is finished?
01:55:09  <thl0>so I can curl and pipe it into the formatter
01:55:10  <mikolalysenko>like if I do a.pipe(b), I want to do something when that pipe completes
01:55:13  <chrisdickinson>mikolalysenko: `stream.on('end', fn)` if you're a consumer
01:55:19  * chrisdickinsonnods
01:55:21  <mikolalysenko>no, b is the consumer
01:55:29  <chrisdickinson>mikolalysenko: ah
01:55:40  <mikolalysenko>here is an example, I want to pipe a stream to a file for a test
01:55:50  <mikolalysenko>then when that pipe completes, do something with that file
01:55:52  <chrisdickinson>`a.pipe(b).on('end', function() { /* .. */ })`
01:56:00  <mikolalysenko>that didn't work :(
01:56:10  <chrisdickinson>hm
01:56:10  <mikolalysenko>end never fired for some reason
01:56:17  <chrisdickinson>what kind of stream is a?
01:56:24  <mikolalysenko>a is a custom stream
01:56:30  <mikolalysenko>it comes from the guts of pngstream
01:56:36  <mikolalysenko>and b is a file stream
01:56:39  <chrisdickinson>aah
01:57:00  <chrisdickinson>you might try `on(
01:57:04  <chrisdickinson>'close')`
01:57:11  <mikolalysenko>trying...
01:57:12  <chrisdickinson>… sorry for the weird newline there.
01:58:02  <mikolalysenko>nope. fail
01:58:44  <mikolalysenko>I am half tempted to just do a setTimeout just to get this test working
01:59:59  <mikolalysenko>actually, it did work
02:00:06  <mikolalysenko>I put the close on the wrong stream
02:00:16  <chrisdickinson>ah, cool
02:00:26  <chrisdickinson>fs streams are… odd, occasionally
02:00:41  <mikolalysenko>yeah...
02:00:57  <mikolalysenko>lately I have grown to distrust streams
02:01:29  <mikolalysenko>I still like the concept though
02:01:50  <mikolalysenko>ah well, back to work
02:05:34  * mmckeggquit (Quit: mmckegg)
02:07:37  * jcrugzz_joined
02:10:20  * jcrugzzquit (Ping timeout: 252 seconds)
02:24:13  * jcrugzz_changed nick to jcrugzz
02:26:59  * vitorpachecoquit (Remote host closed the connection)
02:27:07  <substack>https://gist.github.com/substack/944370affa35f14b7e7a/raw/62c1041b92b5708365e8520a742d7809fb528fb1/gistfile1.txt
02:27:23  <substack>isaacs: hoy
02:27:57  <substack>gozala: I have no idea. It works for me.™
02:28:27  <gozala>substack: I have discovered what was the issue and have submitted a bug report
02:28:49  <gozala>substack: I had syntax error in package.json which was not even related
02:29:09  <gozala>it was just hard to discover why things were going wrong
02:29:09  * vitorpachecojoined
02:29:24  * vitorpachecoquit (Read error: Connection reset by peer)
02:29:47  * vitorpachecojoined
02:51:49  * mikealjoined
02:52:08  * jcrugzz_joined
02:55:04  * jcrugzzquit (Ping timeout: 245 seconds)
03:06:33  * thl0quit (Remote host closed the connection)
03:15:18  * jcrugzz_changed nick to jcrugzz
03:17:38  * mikolalysenkoquit (Ping timeout: 256 seconds)
03:28:33  * mikolalysenkojoined
03:34:19  * jxsonjoined
03:48:24  <jjjjohnn1y>It work for... him :<
03:56:50  * defunctzombie_zzchanged nick to defunctzombie
04:05:49  * dguttmanquit (Quit: dguttman)
04:07:19  * jolissquit (Read error: Connection reset by peer)
04:07:33  * jolissjoined
04:13:13  * jxsonquit (Remote host closed the connection)
04:16:21  * dguttmanjoined
04:17:15  * dguttmanquit (Client Quit)
04:22:13  * brianloveswordsquit (Excess Flood)
04:23:42  * brianloveswordsjoined
05:00:45  * defunctzombieis chillin at casa de Raynos
05:01:50  <substack>\o
05:03:40  <isaacs>substack: yo
05:03:52  <isaacs>substack: i'm going to make keepalive off by default.
05:04:01  <isaacs>substack: but, at least, it'll be just a switch
05:04:36  <substack>why's that?
05:04:40  <isaacs>substack: you often have a lot of connections to the same servers, for datas and such. i think, actually, even in the short term, using keepalive a lot might make your problems less bad.
05:04:41  <substack>won't that make everything crazy slow?
05:04:45  <isaacs>substack: nono
05:04:48  <isaacs>substack: i mean, for the client
05:04:58  <isaacs>substack: if you have pending requests, it'll still reuse existing connections like it does
05:05:03  <substack>makes sense
05:05:12  <isaacs>substack: but basically, current behavior by default, + maxSockets = Infinity by default
05:05:29  <substack>yay
05:05:48  <substack>note too that my complaints about streams2 are transitory in nature
05:05:51  <isaacs>substack: but i've been thinking, actually, for you and stackvm stuff. you should maybe be reusing connections much more aggressively.
05:06:03  <substack>meh
05:06:12  <isaacs>substack: yeah, your complaints are that it's annoying to upgrade. you'll get over that :)
05:06:21  <jcrugzz>isaacs: huzzah. ran into the socket pooling issue today with our service registry (nodejitsu)
05:06:26  <isaacs>history will solve that problem :)
05:06:33  <substack>it MIGHT
05:06:40  <substack>unless there are even more changes
05:06:46  <isaacs>substack: it'll solve it in some way.
05:06:51  <substack>some platforms have a perpetual upgrade cycle
05:06:53  <substack>like python 3
05:07:01  <isaacs>substack: either you'll get over it, and upgrade. or node will lose substack. or the world will lose node.
05:07:10  <isaacs>that's how history is solving python3
05:07:12  <isaacs>no one uses it
05:07:16  <isaacs>problem solved :)
05:07:26  <substack>all my new stuff works on both 0.8 and 0.10, it's just some of the old stuff is still broken
05:07:27  <jjjjohnn1y>what did they do?
05:07:42  <isaacs>substack: yeah, that's why i'm suspecting that the first solution will be the one that takes :)
05:07:48  <substack>jjjjohnn1y: massive backwards incompatible changes
05:07:51  <isaacs>ok, it's bedtime for me.
05:08:00  <isaacs>jjjjohnn1y: they fixed all the warts in the language.
05:08:01  <substack>print 'foo' became print('foo')
05:08:08  <substack>pretty much broke every program ever
05:08:13  <substack>and a bunch of other stuff
05:08:14  <isaacs>jjjjohnn1y: problem is, if you remove all the warts from a language, you also remove the language's skin, and it dies horribly.
05:08:33  <jjjjohnn1y>if your skin is covered with warts
05:08:39  <isaacs>jjjjohnn1y: which all languages are.
05:08:55  <isaacs>jjjjohnn1y: they are an interface between broken human brains and the machines that broken human brains built.
05:09:00  <jjjjohnn1y>i guess so, and all kinds of magic fungii
05:09:04  <isaacs>yeah.
05:09:14  <isaacs>bedtime for isaacs.
05:09:17  * isaacs&
05:09:17  * LOUDBOTALIENATES ALIENATILY
05:19:05  * shamaquit (Remote host closed the connection)
05:26:22  <defunctzombie>substack: "Apply a through stream to the output side of another through stream, preserving the first stream as input in the returned stream." wut
05:32:21  * elliottcablequit (Ping timeout: 248 seconds)
05:35:13  * defunctzombiechanged nick to defunctzombie_zz
05:36:15  * jxsonjoined
05:51:57  <substack>defunctzombie_zz: it's like a.pipe(b); return duplexer(a, b)
05:52:11  <substack>except it doesn't get completely stuck
05:52:26  <substack>because duplexer and pipe don't get along when it comes to pause/resume
05:53:39  * jcrugzzquit (Ping timeout: 252 seconds)
05:56:28  * elliottcablejoined
06:05:18  * mikealquit (Quit: Leaving.)
06:13:10  * mikealjoined
06:16:06  * mikealquit (Client Quit)
06:23:07  * mcollinajoined
06:30:46  * mikealjoined
06:41:34  * fallsemoquit (Quit: Leaving.)
06:42:23  * jcrugzzjoined
06:44:19  * mikolalysenkoquit (Ping timeout: 264 seconds)
06:47:09  * jxsonquit (Remote host closed the connection)
06:53:29  * crankquit (Read error: Operation timed out)
07:00:57  * djcoinjoined
07:04:05  * stagasjoined
07:30:45  * dominictarrjoined
07:31:28  <substack>preview of the new server-side rendering http://browserling.com:9500/substack/defined.html
07:32:02  <substack>written such that I can reuse the same rendering logic browser-side for realtime updates when I get to that part
07:33:12  * stagasquit (Ping timeout: 256 seconds)
07:37:58  <juliangruber>substack: looks good!
07:41:32  * stagasjoined
07:50:10  * mikolalysenkojoined
07:54:04  * crankjoined
07:54:44  * mikolalysenkoquit (Ping timeout: 256 seconds)
07:55:54  * jcrugzzquit (Ping timeout: 264 seconds)
08:23:25  * no9joined
08:28:20  <dominictarr>substack: nice!
08:31:15  * timoxleyjoined
08:35:38  * jibayjoined
08:57:16  <substack>squashed the major bugs in hyperstream/hyperglue/trumpet doing this too
09:03:33  * nicholasfjoined
09:04:28  * stagasquit (Read error: Connection reset by peer)
09:07:48  * thl0joined
09:12:33  * thl0quit (Ping timeout: 256 seconds)
09:25:18  * spionjoined
09:27:07  * ins0mniajoined
09:27:15  * ins0mniapart
09:37:05  * ralphtheninjajoined
09:41:22  <jesusabdullah>substack: what's the state of the substackstack?
09:41:47  <jesusabdullah>substack: got testling on 0.10 yet? XD
10:03:40  <juliangruber>subsubstackstack
10:06:34  * timoxleyquit (Quit: Computer has gone to sleep.)
10:14:27  * ins0mniajoined
10:14:35  * timoxleyjoined
10:21:31  * owen1quit (Ping timeout: 264 seconds)
10:23:14  * jolissquit (Quit: joliss)
10:24:33  * ins0mniaquit (Ping timeout: 240 seconds)
10:25:05  * nicholasfquit (Ping timeout: 248 seconds)
10:26:48  * nicholasfjoined
10:33:37  * timoxleyquit (Quit: Computer has gone to sleep.)
11:08:25  * thl0joined
11:13:04  * thl0quit (Ping timeout: 256 seconds)
11:15:37  * ins0mniajoined
11:23:27  <dominictarr>substack: http://www.youtube.com/watch?v=AyJfHU4GoOQ
11:30:44  <mbalho>dominictarr: when do you get into oslo?
11:30:51  <mbalho>dominictarr: oh wait theres a spreadsheet
11:30:56  <dominictarr>mbalho: tomorrow
11:31:38  <mbalho>dominictarr: when do you leave?
11:33:08  <dominictarr>catching the train from gotheburg, will arive mid afternoon
11:33:32  <mbalho>dominictarr: are you staying after the conf for a couple days? i forget
11:33:47  <dominictarr>yes
11:34:12  <dominictarr>fjords!
11:34:33  <mbalho>dominictarr: yes. i havent figured out what im doing for the 3 days after teh conf yet
11:35:00  <mbalho>dominictarr: theres a 7 hour train to bergen that is like the highest elevation train of all time or something
11:35:41  <dominictarr>that sounds good
11:36:24  <dominictarr>and bergen looks pretty epic
11:37:02  <mbalho>yea we should find a place to crash in bergen
11:37:46  <dominictarr>maybe we could hire bicycles in bergen and go explore
11:37:49  <dominictarr>or kayaks
11:38:21  <mbalho>weathers pretty acceptable here BTW
11:38:28  <mbalho>same as it was in portland
11:41:29  <dominictarr>so, airbnb is pretty expensive there
11:41:30  <dominictarr>https://www.airbnb.com/s/Bergen--Norway?checkin=05%2F26%2F2013&checkout=05%2F29%2F2013&guests=2
11:41:39  * guybrushquit (Excess Flood)
11:41:48  <dominictarr>and there is a hostel that is half that price http://www.hostelbookers.com/property/prp/8365/arr/2013-05-26/ngt/4/ppl/1/
11:42:03  * guybrushjoined
11:42:28  <mbalho>http://www.norwaynutshell.com/sitefiles/1/Rutetider/TimetablesHardangerinanutshellroundtripfromOsloviaBergen2013.pdf
11:42:53  * guybrushquit (Excess Flood)
11:43:00  <mbalho>$380 for that tour
11:43:04  * guybrushjoined
11:43:52  * guybrushquit (Excess Flood)
11:43:55  <dominictarr>how long does that take?
11:43:56  <mbalho>dominictarr: that hostel is 50 bucks a night, not bad
11:44:03  * guybrushjoined
11:44:03  <mbalho>dominictarr: 2 days round trip from oslo
11:44:03  <dominictarr>two days
11:44:09  * guybrushquit (Excess Flood)
11:44:34  * guybrushjoined
11:44:39  * guybrushquit (Excess Flood)
11:45:03  * guybrush_joined
11:45:36  <mbalho>ok goin offline, moar planning later
11:49:28  * jibayquit (Read error: Connection reset by peer)
11:50:01  <dominictarr>okay
11:50:02  * kanzurequit (Ping timeout: 252 seconds)
11:50:34  * kanzurejoined
11:54:13  * nicholasfquit (Remote host closed the connection)
12:03:32  * mikolalysenkojoined
12:12:03  <dominictarr>juliangruber: https://github.com/juliangruber/multilevel/pull/14
12:15:33  <djcoin>dominictarr: https://github.com/dominictarr/JSONStream/issues/31 ; i submitted an issue some times ago, mind to take a look ? Basically : JSONStream does not work in old browser (because of typed array), how to do this ? some shim, then which one ? Otherwise, switch to "buffer" module that has been browserified (but could not make it work)
12:16:18  <dominictarr>djcoin: sorry I missed that
12:16:22  <dominictarr>reading now
12:16:39  <djcoin>no problem, don't know if I stated things correctly
12:17:15  <dominictarr>djcoin: why are you using chrome 23?
12:17:22  <djcoin>chromium
12:17:23  <dominictarr>that is really old, isn't it?
12:17:31  <dominictarr>oh, right
12:17:41  <dominictarr>is that different to chrome23?
12:17:45  <dominictarr>or is it the same?
12:18:12  <djcoin>I don't know - now I'm on chromium26 - I guess it's pretty up to date - got it from archlinux repo which are almost bleeding edge
12:18:29  <dominictarr>right - that sounds better
12:18:41  <dominictarr>does chromium 26 work?
12:18:53  <djcoin>nop, I'm missing the slice method
12:19:22  <dominictarr>okay, yes I have that too
12:19:34  * blobaumquit (Read error: Operation timed out)
12:19:35  <djcoin>oh cool
12:19:51  <djcoin>well this is really annoying because it breaks not once but for all subsequent messages in JSONStream
12:20:08  <djcoin>I guess it breaks the parser that's why
12:20:22  * blobaumjoined
12:20:29  <djcoin>I was not able to find proper shim for those typed array
12:20:55  <djcoin>But I think it would have been easy maybe to use the buffer browserifiable module
12:21:29  <djcoin>but still, i was not really patient/able to make it work :|
12:22:09  <djcoin>(+ then it would require an option to specify you want to use buffer rather than typedarray
12:22:11  <djcoin>)
12:22:57  <dominictarr>so, really you should post this issue on jsonparse
12:23:28  <dominictarr>see this issue https://github.com/creationix/jsonparse/pull/12
12:23:33  <djcoin>Well, jsonparse does not force typedarray for the browser
12:23:49  <dominictarr>substack: are you gonna merge this https://github.com/creationix/jsonparse/pull/12
12:24:45  <djcoin>What happens if you can't slice ? It can carry on like this ? oo'
12:24:47  <djcoin>:)
12:25:09  * blobaumquit (Ping timeout: 248 seconds)
12:26:50  <dominictarr>so, I don't use JSONStream in the browser… but substack, and mbalho do
12:27:37  <dominictarr>the problem here is that we need to clone substack, because he is way to busy, and gets too many issues/pull requests
12:27:47  <djcoin>:)
12:28:38  <djcoin>1/ replicate substack 2/ ??? 3/ proffit
12:28:58  * spionbegins writing a generator "task" library that is callback-based for node 0.11.2 --harmony
12:29:23  <spion>wonder if its worth the effort though
12:39:22  <dominictarr>djcoin: so, what we really need here, is for someone to fork jsonparse and use update it so that it's more compatible with browserify
12:39:51  <dominictarr>djcoin: if you are prepared to do this, I will update JSONStream to depend on your version, instead.
12:40:29  * kevino80joined
12:42:19  <djcoin>dominictarr: i'm not sure I can handle this :) But I may investigate it more to find the proper fix and talk with the jsonparse guy maybe
12:42:55  <dominictarr>creationix doesn't use jsonparse anymore, so he doesn't maintain it actively
12:43:06  <dominictarr>he added substack, but substack is busy too
12:43:43  * spion_joined
12:43:54  <dominictarr>djcoin: it's not that complicated really, you won't have to change any parser logic
12:44:16  <dominictarr>just the bit that accesses the elements
12:44:19  * spionquit (Disconnected by services)
12:44:21  <djcoin>yeah I guess so - but I don't know what would be the proper fix
12:44:22  * spion_changed nick to spion
12:44:45  <dominictarr>might be this https://github.com/chrisdickinson/bops
12:44:52  * yorickjoined
12:45:39  <djcoin>heh, why not
12:45:42  <dominictarr>where you use bops.subarray instead of slice
12:47:18  <djcoin>dominictarr: damn seems bleeding edge, how do everyone know about so many projects appearing all the time ? Watching a whole network ? I got to organize myself, RSS and stuff. Been to long !
12:48:33  <dominictarr>djcoin: everyone who writes cool modules hangs out in stackvm
12:49:21  <dominictarr>so, you have already found the source :)
12:49:51  <djcoin>ahah, yeah I hoped so :)
12:50:12  <djcoin>but i lack organization atm, I should fix this
12:58:16  * yorickquit (Quit: No Ping reply in 180 seconds.)
12:58:32  * yorickjoined
13:10:28  * defunctzombie_zzchanged nick to defunctzombie
13:19:55  * defunctzombiechanged nick to defunctzombie_zz
13:23:23  * st_lukejoined
13:25:25  * kevino80quit (Remote host closed the connection)
13:25:27  * tmcwjoined
13:27:06  * thl0joined
13:27:44  * kevino80joined
13:32:53  <creationix>dominictarr: what's the problem with jsonparse?
13:33:36  <dominictarr>doesn't play nicely with typedarrays, apparently
13:33:51  <creationix>spion: did you see my generator gist? I didn't use full promises, but simple continuables (functions that take a callback)
13:34:13  <creationix>dominictarr: yeah, switching to bops should be easy, I did it for my msgpack codec yesterday
13:34:26  <dominictarr>oh, cool.
13:34:36  <dominictarr>it looks like bops is the way of the future
13:34:53  <creationix>it's the best node/browser way to do binary I can think of
13:35:05  <creationix>the only down-side is the slight performance hit vs using native APIs directly
13:35:12  <creationix>though that is very small
13:35:21  <spion>creationix, thats exactly what I wanted to write
13:35:29  <spion>:)
13:36:06  <spion>but also to try it out with various other node libraries to get a better feel of it
13:53:15  <creationix>djcoin: go ahead and switch jsonparse to use bops, I'll take a pull-request for that
13:54:48  <djcoin>creationix: great, I will see what I can do, i'm not having much time either. But i would be glad to contribute in the coming days
13:55:36  * st_lukequit (Remote host closed the connection)
13:56:32  <creationix>dominictarr: bops doesn't work in IE9 or lower, but that's a feature to me
13:56:41  <creationix>the only IE I really care about is the Surface tablet
13:56:45  <creationix>that comes with 10
13:59:15  <dominictarr>life is pain
14:00:32  * tmcwquit (Remote host closed the connection)
14:01:13  * mcollina_joined
14:04:14  * mcollinaquit (Ping timeout: 245 seconds)
14:13:56  * shuaibjoined
14:34:39  * fallsemojoined
14:35:18  * owen1joined
14:45:25  * thl0quit (Ping timeout: 248 seconds)
14:46:27  * jcrugzzjoined
14:49:29  * mikealquit (Quit: Leaving.)
14:51:26  * thl0joined
14:53:36  * st_lukejoined
14:56:09  * dominictarrquit (Quit: dominictarr)
14:57:01  * mikolalysenkoquit (Ping timeout: 276 seconds)
15:00:35  * dguttmanjoined
15:16:58  * shuaibquit (Quit: Textual IRC Client: http://www.textualapp.com/)
15:23:09  * shamajoined
15:44:02  * st_lukequit (Remote host closed the connection)
15:55:17  * nicholasfjoined
15:59:34  * nicholasfquit (Ping timeout: 246 seconds)
16:02:24  * ralphtheninjaquit (Read error: Operation timed out)
16:03:23  * mikolalysenkojoined
16:07:58  * mikolalysenkoquit (Ping timeout: 246 seconds)
16:08:44  * jcrugzzquit (Quit: leaving)
16:09:12  * jcrugzzjoined
16:09:46  * mikealjoined
16:10:11  * calvinfoquit (Ping timeout: 256 seconds)
16:17:58  * calvinfojoined
16:19:53  <juliangruber>pain creates progress
16:23:25  * thl0quit (Remote host closed the connection)
16:25:14  * mikealquit (Quit: Leaving.)
16:26:01  * Jan____changed nick to jan____
16:28:26  * stagasjoined
16:34:40  * st_lukejoined
16:39:14  <creationix>strange, my new library failed on all browsers with testling, but runs fine locally using beefy https://ci.testling.com/creationix/leb128-frame
16:40:22  * jcrugzzquit (Ping timeout: 276 seconds)
16:41:26  <creationix>I wonder if tape or bops is busted
16:41:57  * jcrugzzjoined
16:42:13  <creationix>hmm, not bops, I still run fine with latest from git
16:43:08  <pkrumins>creationix: this is what happened:
16:43:09  <pkrumins>1369154227630 * preprocessing command: browserify test.js -o 1369154165253.694bd1c4.js --debug*
16:43:12  <pkrumins>1369154228333 Error: module "min-stream-helpers" not found from "/home/creationix/repos/e342e7fd6c9fb8d53448386ed0b2f20dd2dba8dd.1369154164796/test.js"
16:43:29  <pkrumins>we've this issue currently that a browserify error wont get displayed on ci page
16:43:43  <creationix>ahh, thanks
16:43:47  * stagasquit (Read error: Connection reset by peer)
16:43:47  <creationix>I missed a dep in my package.json
16:44:02  <pkrumins>it should work when you update the dep!
16:48:27  <creationix>pkrumins: yep, seems to be working now. Should work in everything except old IE
16:48:31  <creationix>since I use bops
16:48:45  <creationix>now to see why travis refuses to run my node tests
16:49:49  <creationix>I guess they were just slow
16:50:21  * calvinfoquit (Ping timeout: 240 seconds)
16:50:22  <creationix>so it appears bops doesn't like node 0.6.x either (hex encoding)
16:50:28  <creationix>oh well, who uses node 0.6 anymore
16:51:10  <pkrumins>nice!
16:54:07  <creationix>pkrumins: how are you guys so much faster than travis
16:54:15  <creationix>I would think cli testing to be a much easier problem
16:54:23  <creationix>do they just have a ton more traffic?
16:54:43  <juliangruber>creationix: they're running ruby
16:54:44  <juliangruber>:D
16:54:48  <creationix>lol
16:55:16  <creationix>as slow as ruby is, it's nothing compared to 7 different full-blown browser instances
16:55:40  <juliangruber>creationix: or 30
16:55:50  <creationix>I just run the new ones
16:57:20  <pkrumins>creationix, we just dont have too many tests running
17:02:47  <juliangruber>when I do process.exit(1) in a master process it works, but when I do it inside a child_process, it says Error: Unknown signal: 1
17:02:55  <juliangruber>I thought, process.exit(1) was good?
17:03:03  * yorickquit (Remote host closed the connection)
17:03:57  * jxsonjoined
17:04:05  * jcrugzzquit (Ping timeout: 248 seconds)
17:04:43  * dguttmanquit (Quit: dguttman)
17:07:56  * calvinfojoined
17:08:02  <st_luke>1 is for failure
17:09:28  <juliangruber>I know
17:09:32  * mikealjoined
17:09:33  <juliangruber>but I found the issue
17:09:46  <juliangruber>I forwarded that to child_process.kill somewhere
17:09:53  <juliangruber>and the stacktrace didn't show that at all
17:10:16  * dguttmanjoined
17:11:14  * dguttmanquit (Client Quit)
17:13:51  * ralphtheninjajoined
17:14:02  * st_lukequit (Read error: Connection reset by peer)
17:17:43  * dominictarrjoined
17:18:54  * mikolalysenkojoined
17:22:12  * stagasjoined
17:28:00  <jjjjohnn1y>http://www.c-span.org/flvPop.aspx?id=10737439742
17:29:51  <dominictarr>they are doing the whole world a favor, not giving that money to those gangsters
17:32:26  <jjjjohnn1y>yeah
17:32:31  <jjjjohnn1y>lol
17:34:06  <jjjjohnn1y>i am surprised that they can just sign away their profits to a foriegn company tho
17:34:13  <jjjjohnn1y>i gotta figure out ho wthey do that
17:36:34  <dominictarr>well, you can start a company in a foriegn company
17:36:47  <dominictarr>or, be a director, or something
17:37:11  <dominictarr>like - I could start a nz company, that you are also a director of.
17:37:16  <dominictarr>something like that
17:38:14  <dominictarr>ralphtheninja: http://www.facebook.com/media/set/?set=a.461630967036.256357.523207036&type=1&l=2f1025e469
17:40:15  * chrisdickinsonwanders in, sees bops, wanders out
17:40:54  <creationix>chrisdickinson: lol
17:44:49  * thl0joined
17:47:17  * no9quit (Ping timeout: 248 seconds)
17:48:33  * shuaibjoined
17:55:47  * nicholasfjoined
17:57:31  * spionquit (Ping timeout: 246 seconds)
18:00:06  * nicholasfquit (Ping timeout: 264 seconds)
18:02:15  * jcrugzzjoined
18:04:48  <mikolalysenko>basic morphological stuff in js: https://github.com/mikolalysenko/ball-morphology
18:16:55  * mikealquit (Ping timeout: 260 seconds)
18:17:13  <juliangruber>touchee substack: I wrote a headless tape runner with smaller abstractions: https://github.com/juliangruber/tape-run
18:17:41  <juliangruber>everybody else: put `browserify test/*.js | tape-run` into your npm test script and get phantomjs tests with proper exit codes
18:18:36  <mikolalysenko>general question: should I be using tape instead of tap in my projects?
18:19:24  <juliangruber>mikolalysenko: tape is for browsers but runs in node too
18:19:32  <mikolalysenko>ok
18:19:36  <mikolalysenko>does tap run in a browser too?
18:19:37  * brianloveswordsquit (Excess Flood)
18:19:52  <juliangruber>I think there were problems, that's why substack wrote tape
18:19:54  <juliangruber>and it should be way lighter too
18:19:59  <mikolalysenko>ok. I am sold
18:20:05  <mikolalysenko>I will use tape instead from now on
18:20:12  <juliangruber>sweet :)
18:20:25  <juliangruber>and you do `var test = require('tape')` instead of `var test = require('tap').test`!
18:20:49  <mikolalysenko>or just require('tape')( ... )
18:20:59  <mikolalysenko>since I usually put all my test cases in one block
18:21:13  * brianloveswords_joined
18:22:16  <juliangruber>yes, works too
18:22:28  <juliangruber>hij1nx dominictarr: https://github.com/meltingice/flickr-store what we were talking about
18:23:48  <dominictarr>haha, yes
18:23:52  <dominictarr>I saw that
18:24:07  <dominictarr>would be cool if it used stenography
18:24:40  <juliangruber>defunctzombie_zz: https://github.com/juliangruber/tape-run/
18:25:40  <juliangruber>dominictarr: and generative art!
18:25:52  <dominictarr>that is a good plan
18:26:45  * mcollina_quit (Ping timeout: 248 seconds)
18:27:58  * thl0quit (Ping timeout: 246 seconds)
18:29:25  * shuaibquit (Ping timeout: 256 seconds)
18:31:04  * shuaibjoined
18:35:54  * thl0joined
18:53:44  <juliangruber>substack: do you have something link travisify for testling?
18:59:46  <juliangruber>pkrumins: ^
19:01:07  <thl0>juliangruber: nothing there yet - was thinking of doing it if no one else does soon ^^
19:01:40  <juliangruber>thl0: go go forit
19:03:13  <thl0>ok, when I find some bandwidth, also trying to get a quick prototype of https://github.com/thlorenz/modurater off the ground (I think we need something like this)
19:03:57  <thl0>juliangruber: so depending on how well that goes I'll try to get to it this weekend unless substack beats me to it (wink wink)
19:05:25  <juliangruber>thl0: modurater <3
19:05:36  <juliangruber>thl0: that's a problem many ppl tried to solve
19:05:45  <juliangruber>thl0: that best I found still is nipster
19:06:02  <thl0>I've seen nipster as well
19:06:15  <thl0>juliangruber: but we need a way to downvote as well (ala stackoverflow)
19:06:42  <juliangruber>thl0: re component: I'd use that a lot, it's such a shame that we have 2 ecosystems...I've had sleepless nights because I had to choose between component and browserify for projects
19:06:50  <thl0>juliangruber: so at the worst modurater will be a sandbox to learn what works and possibly pull some of it into the npm org
19:06:56  <juliangruber>thl0: yes, downvotes with clearly stated reasons
19:07:20  <thl0>juliangruber: exactly how I wanted it - i.e. link to issue required in cases
19:07:36  <juliangruber>sweet
19:07:38  <thl0>juliangruber: https://github.com/thlorenz/modurater#downvotes
19:08:13  <thl0>no way to downvote without reason - make sure to throw some ideas around on the related gist
19:08:19  <juliangruber>thl0: There are good papers on rating and flock intelegence out there
19:08:25  <juliangruber>yes
19:08:44  <juliangruber>thl0: will it require authentication?
19:08:51  <thl0>juliangruber: if you don't mind could you link the papers you know on the gist?
19:09:07  <thl0>juliangruber: yes - it may even autostar on upvote
19:09:16  <juliangruber>thl0: I've seen then and read a bit but can't remember. google ftw
19:09:17  <thl0>https://github.com/thlorenz/modurater#github-interaction
19:09:35  <juliangruber>so github oauth?
19:09:51  <thl0>there'll be no way to upvote without starring (i.e. it will sync)
19:09:54  <thl0>juliangruber: yes
19:10:29  <thl0>juliangruber: it basically adds features on top of github features in order to not create two separate voting systems which would be stupid
19:10:52  <thl0>juliangruber: it'll have to use github oauth
19:11:17  <juliangruber>thl0: could it work without authentication too? give authenticated users more power / credibility but count the anonymous masses too
19:11:57  <thl0>juliangruber: it could, but the moment you want to vote I'd have to know your credibility and downvote credits (derrived via github)
19:12:16  <juliangruber>mh true
19:12:19  <thl0>juliangruber: so to query modules no login required, but for everything else it is
19:12:44  <thl0>juliangruber: I mean to star on github you also need to login ;)
19:13:51  <juliangruber>yes
19:13:53  <juliangruber>sounds good
19:13:56  <juliangruber>i am excite
19:27:44  <thl0>juliangruber: just saw that you commented on component as well - yeah that's a shame
19:27:48  * thl0sighs
19:28:37  <juliangruber>thl0: I was one of the early adopters and even wrote a 2 months website at work using components
19:29:02  <juliangruber>I was even using components before the components binary or libraries existed :D
19:29:23  <thl0>juliangruber: wow - I understand the concept, not sure we need it though - just inline the css
19:29:45  <juliangruber>thl0: http://tjholowaychuk.com/post/27984551477/components
19:29:56  <juliangruber>you could use insert-css too
19:29:58  <juliangruber>BUT
19:30:11  <juliangruber>component has way faster downloads than npm
19:30:29  <juliangruber>and starts with a new ecosystem, e.g. component's eventemitter is way smaller than node's
19:30:45  <juliangruber>and tj writes all the stuff everyone needs which is awesome too
19:30:56  <thl0>juliangruber: that's all great, but npm integration is way more important than anything else
19:31:18  <juliangruber>thl0: yeah sure
19:31:39  <juliangruber>still there are already more useful browser-only packages in component than in npm
19:31:49  <thl0>juliangruber: it also allows to reuse modules server and client side (don't think a component makes sense anywhere else but the browser)
19:32:03  <juliangruber>components can work in node too
19:32:12  <thl0>juliangruber: and that's the sad part - duplication of efforts
19:32:24  <chrisdickinson>browser directive in package.json seals the deal for me
19:32:43  <juliangruber>I think one of the basic differences between component and browserify is that component is dead simple and browserify does all this static analysis magic
19:32:46  <thl0>people from the differnt md systems will have to solve same thing each time
19:32:50  <juliangruber>browserify's .browser is great, yes!
19:33:10  <chrisdickinson>the big problem with each is the lack of integration with other language environments
19:33:13  <chrisdickinson>imho
19:33:19  <thl0>juliangruber: final verdict - if it doesn't integrate with npm we don't need it
19:33:23  <juliangruber>chrisdickinson: like?
19:33:30  <chrisdickinson>neither will win until there's good bindings/versions in ruby+rails / python+django
19:33:46  <juliangruber>thl0: modules can be written in a way that they work with npm and component -> possibly more usage
19:33:54  <chrisdickinson>there are far more folks from those communities working on frontend js than there are node.js folk working on frontend js
19:34:02  <juliangruber>hmm
19:34:05  <chrisdickinson>juliangruber: there's a fatigue problem with that
19:34:09  <juliangruber>haven't seen it that way
19:34:31  <thl0>juliangruber: don't want to include the module dance for each of my modules
19:34:37  <juliangruber>and those ppl use grunt and require.js and yeoman and stuff...oh and bower
19:34:42  <chrisdickinson>in that i already have to package my code with standalone for require.js users, and add a component.json, and add whatever bower uses, and ender, and ..
19:34:43  <juliangruber>thl0: true that
19:34:59  <chrisdickinson>right, we (I?) want the node solutions to win
19:35:17  <thl0>go npm \0/
19:35:18  <chrisdickinson>require.js is an inglorious box into which have been placed all of mankind's sins
19:35:25  <chrisdickinson>*has
19:35:27  <juliangruber>I don't really care who wins, I want ONE thing
19:35:30  <chrisdickinson>right
19:35:42  <dominictarr>npm is the best pattern
19:36:00  <juliangruber>but there are problems with browserify and production browser development
19:36:05  <chrisdickinson>?
19:36:06  * thl0is happy that people in the know know
19:36:09  <juliangruber>some ppl don't want crossbrowser fixes in their modules
19:36:17  <juliangruber>so you have to fork
19:36:23  <juliangruber>and use git urls
19:36:28  <juliangruber>and update manually
19:36:29  <juliangruber>blablabla
19:36:34  <chrisdickinson>or write a new module that uses the old one
19:36:41  <juliangruber>whereas components are for the browser and you know that's what you do
19:36:47  <chrisdickinson>i.e., write something that picks between "ever" and "ie-ever" depending on availability
19:37:01  <juliangruber>chrisdickinson: but that's that glue code again
19:37:08  <chrisdickinson>yes
19:37:11  <chrisdickinson>but
19:37:14  <chrisdickinson>it's on npm so it's solved forever
19:37:32  <chrisdickinson>for some values of "forever"
19:37:40  <mikolalysenko>the component is "for-the-browser" argument is pretty ridiculous imo
19:37:47  <juliangruber>chrisdickinson: it's concepts are forever but not the implementation and interfaces
19:38:42  <juliangruber>mikolalysenko: I've had to spend days making npm modules work in browsers, with component modules already available but in the wrong module system
19:38:57  <chrisdickinson>i haven't really found a convincing argument for component yet, aside from "npm has things that aren't for the browser"
19:39:09  <chrisdickinson>which, while true, npm has things that aren't for recent versions of node, either
19:39:21  <juliangruber>haha yeah
19:39:30  <chrisdickinson>hell, npm has things that aren't for anyone running node without --enable_harmony or whatever the flag is
19:39:41  <juliangruber>so, to come back to the initial topic, thl0, a bridge would be uber
19:39:52  <chrisdickinson>and things that don't work unless you slap a `require('coffee-script')` somewhere in your code
19:39:52  <substack>the problem with using not-npm is that npm is the biggest so if you need something that has only been published to npm, you've got to go and patch it manually to work in your custom thing
19:40:04  <substack>and this exists to a lesser degree in the opposite direction
19:40:07  <substack>but npm is the biggest
19:40:31  <thl0>juliangruber: is uber like this UMD thing that tries to bridge all module systems?
19:40:41  <substack>and when you build an app or a library, it's really nice to have an expansive dependency graph to draw upon
19:40:42  <dominictarr>juliangruber: npm's strength is things that work in both the browser and the server
19:40:45  <juliangruber>substack: if you need everything to run in old ies and can't utilize streams so often I find that there can be more component modules
19:40:50  <dominictarr>things like leveldb
19:40:57  <juliangruber>thl0: uber == super cool awesome
19:40:59  <dominictarr>scuttlebutt, streams
19:41:04  <substack>juliangruber: TONS of stream stuff works all the way down to ie6
19:41:12  <substack>testling badges prove as much
19:41:21  <juliangruber>substack: I know. But often it makes 0 sense to use them in browsers
19:41:25  <mikolalysenko>honestly, why would you be programming in js unless you were planning on reusing code between the server and browser? the argument that npm is not for browsers is absurd and defeats the whole point of using node
19:41:34  <chrisdickinson>juliangruber: https://github.com/PDXNode/pdxnode/tree/master/2013-mar/modules/part-3
19:41:42  <juliangruber>mikolalysenko: jisasnif
19:41:42  <chrisdickinson>i feel like streams are a great fit for the browser
19:41:44  * st_lukejoined
19:42:12  <chrisdickinson>and not just for things like leveljs/websockets/etc
19:42:22  <juliangruber>re streams: you have to educate people. I'm working in a php shop at the moment and held a talk on streams but they won't get it unless they spent enough time with it
19:42:45  <thl0>juliangruber: I know what uber means (bin auch deutsch), just wasn't sure if that was a thing somewhere?
19:42:46  <juliangruber>they even enfore jquery and I'm the only one not using it
19:42:48  <chrisdickinson>right, but that's the same with JS in general -- evented programming is not a concept that "happens" much in other major web languages
19:42:57  <juliangruber>thl0: oh, hallo :D
19:43:10  <chrisdickinson>we use streams with jquery, too :)
19:43:12  <thl0>juliangruber: ;)
19:43:20  <substack>I'm going to tie together some of this streaming html templating stuff together that I've been working on for the testling-ci project page refactor
19:43:28  <chrisdickinson>cool
19:43:53  <substack>it's a really good fit for repurposing server-side rendering logic for client-side realtime updates
19:44:10  <substack>it's just really hard and the primitives aren't quite where they need to be to make it easy yet
19:44:11  <juliangruber>chrisdickinson: I'd be using streams all the time if it was for me, but other ppl need to maintain my code and when they see through().pipe(duplex()).pipe(throughout) they just fall
19:45:22  <juliangruber>so I'm completely on your side guys, but day to day work at my company has left its mark on me too
19:45:31  <chrisdickinson>juliangruber: i've noted that, but on the other hand, they won't have a super great time with other evented primitives either
19:45:41  <juliangruber>(which is why I'll have a new job starting june :))
19:46:27  <thl0>juliangruber: cool, but in your case you could help it by adding a comment or wrapping it in a well named function
19:47:00  <thl0>juliangruber: if code becomes too elegant to be easily understood then that's not good either though
19:47:44  <juliangruber>it's not just that though...as substack said it's not easy enough yet
19:47:44  <chrisdickinson>actually, the only issue i've run into with streams and co-workers is that occasionally one of 'em will go far too streams-happy :)
19:47:57  <thl0>juliangruber: "easily understood" is relative though, if someone doesn't know streams, then that's not your fault ;)
19:48:01  <juliangruber>if i'd be doing realtime collaborative things that'd a no-brainer
19:48:11  <juliangruber>but with simple crud things it's harder
19:48:34  <chrisdickinson>which was an important lesson on both sides -- for my part, i found out that i need to write down that "it's usually not a good idea to have an event emitter instance that listens to itself" (which i knew prior, but sort of neglected to communicate)
19:48:42  <juliangruber>chrisdickinson: someone wanted to use only streams, no callbacks...took me 2 days to fix that
19:48:57  <thl0>juliangruber: chrisdickinson we ran into similar problems using RX in .Net (which are basically streams on steroids)
19:49:12  <chrisdickinson>it's a learning experience -- i mean, i have made the same mistakes while getting a handle on streams at first
19:49:27  <chrisdickinson>so, a rocky experience for people new to the concept is to be expected
19:49:35  <chrisdickinson>but, it gets better :D
19:50:17  <juliangruber>yeah
19:50:28  <juliangruber>I think things will become way better for me with the new job too
19:52:16  * shuaibquit (Ping timeout: 252 seconds)
19:53:20  <juliangruber>tbh one motivation for writing components was to get retweeted by tj to get known more...which worked :D
19:53:27  <chrisdickinson>haha
19:53:30  <chrisdickinson>:)
19:53:35  <mikolalysenko>the truth comes out!
19:53:49  <chrisdickinson>yeah, i suppose some portion of my reticence about component is that i generally don't share TJ's aesthetics
19:53:58  * kevino80quit (Remote host closed the connection)
19:54:56  <mikolalysenko>I just think the whole component project seems really unnecessary, and looks like a great way to further fragment an already highly fractitious and confused ecosystem
19:56:15  <substack>something that does asset building on top of npm would be good though
19:56:23  <juliangruber>the learnboost guys just do their own thing...iirc they're running a modified version of npm, because they need multiple package.json's inside a package, like app/login/package.json app/blog/package.json but also app/package.json and vanilla npm doesn't seem to do that well
19:56:35  <chrisdickinson>i was playing with that with http://npm.im/static-assets
19:56:40  <chrisdickinson>for voxel.js
19:56:48  <substack>I use multiple package.jsons all the time
19:56:57  <substack>hasn't been a problem at all with vanilla npm
19:57:08  <substack>you just (cd somedir; npm whatever) in a bash script
19:57:14  <juliangruber>oh, good to know!
19:57:18  <thl0>chrisdickinson: that's kind of what's annyoing a bit - TJ seems to star things just bc they are components
19:57:31  <juliangruber>and as soon as we get a distributed npm (dominictarr) it'll have won anyways
19:57:34  <thl0>sorry meant juliangruber
19:57:48  <juliangruber>thl0: I've been retweeted A LOT :D
19:57:57  <thl0>that's what I mean
19:58:04  <juliangruber>he's like a catalyst, but also a super nice guy
19:58:08  <substack>on top of tacodb!
19:58:27  <juliangruber>substack: * on top of tacodb!
19:58:31  <substack>with selective, prioritized replication
19:58:36  <chrisdickinson>haha, i can whip up a checkout-tree module on top of what's there with js-git now
19:58:36  <thl0>should be the module's content not the attempt to push an alternative module system on people by making it look like everyone is using it
19:58:41  <dominictarr>stars should be weighted by how much you star things
19:58:48  <chrisdickinson>so distributed npm with git urls should be possible without relying on git-core
19:59:07  <mikolalysenko>damn. I star every damn thing on github like a chump
19:59:16  <thl0>dominictarr: plz contribute related thoughts to https://gist.github.com/thlorenz/5610508
19:59:18  <dominictarr>git is awkward, because you would have to poll each repo
19:59:26  <chrisdickinson>thl0: it does strike me as a bit of framework empire building at times
19:59:27  <juliangruber>mikolalysenko: ppl have unfollowed me on github because I star too much :S
19:59:34  <thl0>I'm trying to get a community driven rating system going there
20:00:03  <jjjjohnn1y>component is full of garbage
20:00:26  <substack>that's not necessarily bad
20:00:29  <mikolalysenko>jjjjohnn1y: I wouldn't say that, there are some cool things in component
20:00:39  <substack>I would even say that component isn't full ENOUGH with garbage
20:00:41  <chrisdickinson>it is sort of hilarious that github has largely the same discovery problem as npm
20:01:01  <substack>npm is so full of garbage but that's the only way to get enough raw materials for the good stuff to float to the top
20:01:29  <dominictarr>thl0: so, your vote power is how many pull requests of yours are accepted?
20:01:30  <substack>discovery is worse on github because it's not limited to a language/platform/module system
20:01:47  <thl0>dominictarr: whatever best measure we come up with
20:02:02  * kevino80joined
20:02:15  <thl0>dominictarr: trying to figure out how credible a rater is by looking at contributions to his own or other modules
20:02:17  <juliangruber>substack: this distributed npm thing could replace github too, at least for public projects
20:02:50  <dominictarr>thl0: also, it's a great deal of work to fake that
20:02:54  <thl0>dominictarr: that way people in the know will be able to influence a rating more than others
20:02:54  <jjjjohnn1y>juliangruber: i want npm to be a git server
20:02:59  <thl0>dominictarr: exactly ;)
20:03:07  <chrisdickinson>npm just needs a better mechanism for wading through what's there to find what you want. whether that's modurator or just more vocal folk working on packages, or a search interface that works by submitting tests...
20:03:07  <thl0>dominictarr: keeps the trolls out
20:03:13  <dominictarr>a nice approach to the sybil attack
20:03:16  <jjjjohnn1y>so i dont have to fork into github to contribute to a module
20:03:20  * shuaibjoined
20:03:25  <dominictarr>you could script it
20:03:27  <thl0>dominictarr: you'll not be able to vote at all if you have no contributions
20:03:37  <dominictarr>what about issues?
20:03:42  <dominictarr>issues are contributions too
20:03:50  <chrisdickinson>i'd honestly love to be able to filter by keyword + send a module that exports a test function that can require each package and run a given simple test against it
20:03:50  <jjjjohnn1y>dont favor the creaters over the users
20:04:01  <juliangruber>issues will be failed tests commited to a branch, that's the only way to post one
20:04:03  <chrisdickinson>(if that makes sense)
20:04:06  <mikolalysenko>maybe there could be like an npm style thing for demos. right now npm gives you packages, but what if you could put applications or examples somewhere?
20:04:16  <chrisdickinson>"i need colors, i want it to support this api"
20:04:16  <mikolalysenko>(other than github obviously)
20:04:32  <chrisdickinson>then have some sort of community site where people can save their search functions or something
20:05:33  <substack>store search functions as repos in the distributed network itself
20:06:02  <juliangruber>haha
20:06:15  <juliangruber>predestined ddos
20:06:21  <thl0>dominictarr: yes, but they'll way less - actually when you down vote in most cases you need to link an issue
20:06:43  <thl0>dominictarr: PRs are much higher weighted
20:06:51  <chrisdickinson>search for search functions with a search function that evaluates the results of the tested search function
20:07:24  <thl0>dominictarr: I very much welcome comments on that gist to come up with the best way to have a simple enhancement of what github already gives us and integrate with it seamlessly
20:07:39  <chrisdickinson>i can't decide whether that sounds uncomfortable or powerful.
20:08:46  <chrisdickinson>i suppose all it would take is for someone to add tons of popular tags to a module that removes a user's home directory contents before that goes unwell
20:09:31  * dguttmanjoined
20:09:32  <chrisdickinson>(solution: use browserify + add-module-deps to run captured modules in a node.js "jail"?)
20:10:16  * st_lukequit (Read error: Connection reset by peer)
20:10:24  * st_lukejoined
20:10:24  <dominictarr>yeah, I've been working on that
20:10:43  <dominictarr>https://github.com/dominictarr/securify
20:12:26  <chrisdickinson>awesome
20:13:48  <Raynos>Is annoying else finding streams hard to do in production?
20:14:01  <Raynos>they feel like an easy way to leak memory / bottle CPU
20:14:11  <thl0>dominictarr: btw are you in NYC currently?
20:14:37  <dominictarr>thl0: no, I'm in sweden
20:14:59  <thl0>wow Sweden is nice!
20:15:32  <dominictarr>gotheburg with ralphtheninja
20:15:46  <dominictarr>tomorrow, olso for webrebels
20:17:02  * AvianFluquit (Remote host closed the connection)
20:19:36  * jolissjoined
20:22:56  * st_lukequit (Remote host closed the connection)
20:28:22  * mcollinajoined
20:54:45  <mikolalysenko>I can't believe that there isn't a single library to do bsplines on npm
21:00:20  * mikealjoined
21:19:32  * mikealquit (Quit: Leaving.)
21:22:07  * st_lukejoined
21:27:32  * mcollinaquit (Read error: Connection reset by peer)
21:27:38  * mcollina_joined
21:29:22  <mikolalysenko>convolutions and cross correlations: https://github.com/mikolalysenko/ndarray-convolve
21:29:23  * mcollina_quit (Read error: Connection reset by peer)
21:30:06  * mcollinajoined
21:30:06  * mcollinaquit (Read error: Connection reset by peer)
21:31:45  <chrisdickinson>mikolalysenko: do you have anything for longest common subsequence / shortest edit distance for ndarrays (or even 1d arrays)?
21:32:34  <mikolalysenko>chrisdickinson: no, mostly focusing on image processing/volume graphics stuff right now
21:32:40  <chrisdickinson>cool
21:32:53  <mikolalysenko>if you want to take a crack at it, go for it
21:33:11  <chrisdickinson>i've possibly gotta do one of the two of those for git
21:33:26  <chrisdickinson>to produce binary copy/insert deltas
21:34:04  <mikolalysenko>yeah, I know the standard dynamic programming algorithm to solve that problem
21:34:06  * tilgovi_joined
21:34:13  * mcollinajoined
21:34:23  <mikolalysenko>not sure if it is efficient enough for big text files though, since it is quadratic
21:34:33  * nicholasfjoined
21:39:06  * mcollinaquit (Ping timeout: 264 seconds)
21:39:32  * mcollinajoined
21:44:42  * stagasquit (Read error: Connection reset by peer)
21:52:14  * kevino80quit (Remote host closed the connection)
21:52:20  * spionjoined
21:57:21  * tmcwjoined
22:06:43  * djcoinquit (Quit: WeeChat 0.4.0)
22:07:24  * st_lukequit (Remote host closed the connection)
22:13:16  <dominictarr>substack: you there?
22:18:09  <jesusabdullah>hello Mr. Tarr
22:18:29  * thl0quit (Remote host closed the connection)
22:34:44  * nk109quit (Quit: Computer has gone to sleep.)
22:38:47  <juliangruber>substack: since browser-resolve removed all the node core things in 1.0.0 we need a new way to put them into browserify. do you have a plan already?
22:39:01  <juliangruber>defunctzombie_zz: ^
22:39:39  * shuaibquit (Ping timeout: 245 seconds)
22:43:39  * substackhere
22:45:12  * fallsemoquit (Quit: Leaving.)
22:49:17  <substack>just happily depending on an old version
22:51:49  * mikolalysenkoquit (Ping timeout: 248 seconds)
22:52:42  * no9joined
22:55:03  * nicholasfquit (Remote host closed the connection)
23:01:29  <substack>dominictarr: that tshirt was so doomed
23:01:33  <substack>the limit was set at 300
23:01:41  <substack>I don't think they understood how tspring works
23:02:18  <dominictarr>too high?
23:02:52  <substack>by 10 times
23:03:54  * shuaibjoined
23:05:48  <jesusabdullah>yeah definitely
23:05:57  <jesusabdullah>cool shirt though
23:09:01  * spionquit (Ping timeout: 246 seconds)
23:10:59  <dominictarr>substack: can you merge this https://github.com/substack/browser-pack/pull/13 ?
23:20:58  <substack>oh that's really minimal
23:21:23  * jaz303_quit (Ping timeout: 252 seconds)
23:21:38  * jaz303joined
23:23:12  * timoxleyjoined
23:23:32  <substack>dominictarr: merged, published
23:23:38  <dominictarr>sweet!
23:23:51  <substack>does browserify core need to update to 0.9.0?
23:24:04  * no9quit (Ping timeout: 246 seconds)
23:24:33  <dominictarr>browserify doesn't depend on it, but there is another pull request on node-detective that i just put it
23:24:34  <dominictarr>in
23:24:45  <dominictarr>that fixes an browserify issue
23:24:49  * mcollinaquit (Ping timeout: 248 seconds)
23:28:34  * ins0mniaquit (Ping timeout: 276 seconds)
23:31:10  <substack>dominictarr: your browser-pack patch breaks the test/export.js in browserify
23:31:15  <substack>when I upgraded it
23:31:29  <dominictarr>oh, damn
23:32:23  <substack>it's just the empty case
23:32:27  <substack>where there are no exports
23:33:04  <dominictarr>fixing...
23:34:43  <substack>thanks!
23:35:03  <substack>it seems that there is some unmatched logic at the end
23:35:29  <substack>the trailing "},{},[])\n;" has no counterpart earlier in the output
23:36:17  <dominictarr>oh, the bit right after the prelude
23:36:29  <dominictarr>aha, because it's written on the first dep!
23:37:17  * no9joined
23:38:59  <dominictarr>substack: fixed
23:39:57  <substack>merged your detective patch
23:42:55  <substack>having so many tests is the only way I can iterate on this stuff at all
23:43:27  <substack>but the best part is that if the tests pass and the patches are within scope, I don't even really need to look at the changes very closely
23:58:43  * _ddgbotjoined