00:00:01  * ircretaryquit (Remote host closed the connection)
00:00:08  * ircretaryjoined
00:02:41  * Altreusquit (Read error: Connection reset by peer)
00:06:11  * ins0mniaquit (Ping timeout: 260 seconds)
00:09:00  * ins0mniajoined
00:12:20  * lyaunzbequit (Remote host closed the connection)
00:17:29  * mikealjoined
00:29:08  * lyaunzbejoined
00:33:44  * ins0mniaquit (Remote host closed the connection)
00:39:03  * dominictarrquit (Quit: dominictarr)
00:44:16  <gwenbell>dominictarr: i spent last night watching all your lovely talks
00:48:32  * ins0mniajoined
00:59:04  <gwenbell>Real-time cats, check: http://facesofnyfw.com/
01:02:48  <mbalho>gwenbell: :D
01:03:04  <gwenbell>mbalho: hello!
01:03:39  <mbalho>gwenbell: did you make the facesofnyfw?
01:04:30  <gwenbell>mbalho: negatron - this is the 2nd year of this site, won a bunch of awards last year - uses isotopes
01:05:08  * mikealquit (Quit: Leaving.)
01:05:53  <gwenbell>mbalho: these peeps made it - http://bentrova.to/
01:06:17  <mbalho>gwenbell: ah cool
01:06:19  * CryptoQuick_quit (Quit: CryptoQuick_)
01:08:20  * tilgoviquit (Remote host closed the connection)
01:19:30  * dominictarrjoined
01:47:51  * alunnyquit (Quit: Lost terminal)
02:05:02  <Raynos>evbogue: it depends
02:05:16  <Raynos>for p2p social
02:05:21  <Raynos>you need users to be able to store their own shit
02:07:23  <evbogue>Raynos: exactly. I've been having trouble sleeping trying to figure out how to liberate the entire world from living in siloed centralized social platforms. I figured you guys might have an idea of how to do that.
02:07:35  <Raynos>First go look for p2p social protocols
02:08:00  <evbogue>Like Activitystreams?
02:08:23  * gwenbellquit (Remote host closed the connection)
02:08:48  <Raynos>evbogue: http://news.ycombinator.com/item?id=4418904
02:09:14  <evbogue>Raynos: I used tent, it doesn't work as advertised.
02:09:45  * shuaibjoined
02:09:50  <evbogue>They have a centralized closed-sourced version that works, the open sourced servers don't communicate with each other.
02:10:04  <Raynos>lol
02:10:26  <evbogue>So, problem not-solved.
02:10:31  <evbogue>But I still have hope!
02:11:10  <evbogue>Is this the kind of thing that has to wait to be solved until the new node stream APIs get finalized?
02:11:47  <evbogue>Tent is also Ruby/rails, so gets really slow once you load it up with users.
02:12:11  <evbogue>Hence why I rest my hopes on Node.
02:14:34  * shuaibquit (Client Quit)
02:14:55  <evbogue>Is Scuttlebutt + Append-only the kind of thing + a front-end UI that could make distributed social work? Or am I not right about that?
02:17:42  <Raynos>the important thing is that it has to be p2p
02:17:46  <Raynos>which means no centralized database
02:17:51  <Raynos>scuttlebutt can work but thats a tradeoff
02:18:31  <evbogue>what's the tradeoff?
02:20:50  <Raynos>its eventually consistent
02:20:54  <Raynos>so you have no consistency
02:21:10  <Raynos>also the current implementation is memory heavy
02:23:20  <Raynos>evbogue: but you have to be eventually consistent
02:23:31  <Raynos>because a distributed p2p system is under a lot of partition faults
02:23:38  <Raynos>and you have to have high availability
02:23:46  <Raynos>although I guess thats optional
02:25:55  <evbogue>so if I had 150 nodes up micropublishing 140 character messages to each other, using scuttlebutt and leveldb as the backend, it might never reach eventual consistency across all nodes?
02:26:20  <evbogue>(this is in theory, because I have not coded this yet, because I do not know how)
02:26:50  <Raynos>it will reach eventual consistency eventually
02:26:58  <Raynos>assuming a) no network splits
02:27:04  <Raynos>b) no network saturation
02:30:38  * gwenbelljoined
02:33:13  <evbogue>I guess I'm trying to figure out if these projects, like Append-only/Scuttlebutt are intented to lead towards my distributed peer 2 peer social utopian dream, or if they are intented for another purpose.
02:33:55  <evbogue>and if I should be learning the coding chops to build on them, or if I should be learning to build on something else.
02:34:11  <evbogue>like Go or something.
02:42:46  <Raynos>well you can learn from scuttlebutt
02:42:50  <Raynos>and reimplement something similar
02:42:57  <Raynos>scuttlebutt feels to wasteful on memory
02:43:11  <Raynos>so Im highly tempted to write something similar but more based on storage
02:44:02  * gwenbellquit (Remote host closed the connection)
02:44:46  <Raynos>evbogue: the next things you need for distributed & p2p is Webrtc
02:44:56  <Raynos>Look at https://github.com/Raynos/painter/tree/master/examples/chat
02:45:06  <Raynos>Provide feedback if it needs improvement
02:45:29  <evbogue>Raynos: I like WebRTC. I used the apprtc demo to do an interview the other day, and it worked well.
02:45:51  <evbogue>For a video call, that is.
02:47:25  * lyaunzbe_joined
02:47:25  * lyaunzbequit (Read error: Connection reset by peer)
02:48:35  * ins0mniaquit (Ping timeout: 248 seconds)
02:51:11  <evbogue>Raynos: is WebRTC something that would be useful for a peer to peer distributed micropublishing app. Where every peer is a client/server?
02:51:29  <Raynos>yes
02:51:35  <Raynos>because its the only p2p protocol in browsers
02:51:40  <Raynos>assuming you want to do p2p in browsers
02:51:46  <evbogue>I like browsers.
02:51:49  <evbogue>Everyone has them.
02:56:15  <evbogue>Raynos: I've only seen video calls, chat apps, and your example of a paint app using webRTC. have you seen anything with multi-node publishing?
02:56:29  <Raynos>the paint app does multi-node publishing
02:56:32  <Raynos>open multiple clients
02:56:35  <Raynos>draw on multiple clients
02:56:51  <evbogue>Raynos: ah ha.
02:58:12  <Raynos>you should help build out the p2p foundations
02:58:18  <Raynos>and then build your social thing on top :D
03:00:13  <evbogue>Raynos: I guess I'm trying to figure out what needs to be built out, what's here already, and where to focus my attention.
03:00:38  <Raynos>your going to have to build your protocol and storage mechanism,
03:00:43  <Raynos>and then the app
03:00:47  <Raynos>you could use existing ones
03:00:54  <Raynos>but they probably aren't right for your use case
03:04:50  <evbogue>So you're saying what I'm looking for probably doesn't exist and I need to build from the ground up?
03:06:54  <Raynos>if you want to target node, browsers & webrtc yes
03:07:06  <Raynos>if you want to target desktop apps over UDP then there's probably a solution
03:09:20  <evbogue>Raynos: thanks for chatting. Signing off for the night. I appreciate your input!
03:09:30  * evboguepart
03:30:07  * defunctzombie_zzchanged nick to defunctzombie
03:37:09  * st_lukejoined
03:50:01  <defunctzombie>substack: I need an async resolve
03:50:33  <defunctzombie>and i have a branch of required for you to try with a sample function you can plug into it
04:02:17  <defunctzombie>substack: https://github.com/shtylman/node-required/tree/resolve
04:03:11  <defunctzombie>and the basic thing I hacked on while testing it: https://gist.github.com/shtylman/4760150
04:03:23  <defunctzombie>which provides the browser field support (but is kinda ghetto still)
04:04:02  * substackat SFO
04:04:18  <substack>defunctzombie: excellent
04:04:38  <substack>defunctzombie: I have a synchronous require implementation in http://github.com/substack/node-resolve
04:04:46  <substack>could be made to be async much more easily than any core apis
04:04:57  <defunctzombie>substack: yea, that is what I used in the gist
04:05:44  <defunctzombie>substack: basically, extending required at the resolve point will do everything we talked about wrt replacing code/requires
04:05:53  <defunctzombie>it was actually quite painless a change and is pretty cool
04:11:17  * fotoveritejoined
04:12:42  <substack>sounds pretty good
04:12:57  <substack>oh another thing I realized on the train
04:13:30  <substack>required should have an option to send the package.json contents in the results
04:13:47  <substack>I need that data for browserify plugins
04:14:07  <defunctzombie>in what sense?
04:14:25  <defunctzombie>I've only ever gotten that data when actually doing a require('package.json');
04:15:48  <substack>I'd like a way for modules to define source filters
04:16:19  <defunctzombie>source filters?
04:16:41  <defunctzombie>why would a module specify that?
04:17:02  <substack>so modules can do things like asset packing
04:17:21  <substack>without compiling to intermediary js files
04:17:28  <substack>I want to write an fs.readFileSync() plugin for instance
04:17:41  <substack>that will inline source files
04:18:18  <defunctzombie>my current experience with asset packing css is to just use npm-css
04:18:28  <defunctzombie>css is better loaded upfront anyway for rendering reasons
04:18:35  <defunctzombie>html templates is a different story
04:19:41  <defunctzombie>fs.readFileSync() is tricker since you have to find it in the code, and start replacing their code
04:20:02  <defunctzombie>versus a require, which could be used to load arbitrary text files
04:20:08  <substack>not necessarily
04:20:10  <defunctzombie>but I think that gets much more involved
04:20:25  <substack>using require to load text is a bad idea
04:20:29  <defunctzombie>I agree
04:20:43  <substack>require() is for loading javascript and json, a subset of javascript
04:21:16  <defunctzombie>I am not sold on the asset packing case :/
04:21:22  <defunctzombie>people will want to pack images
04:21:24  <defunctzombie>or other crap
04:21:33  <defunctzombie>I think those must be solved in different ways
04:21:45  <substack>possibly
04:22:17  <defunctzombie>you should add node 0.8 and 0.9 to travis for resolve module
04:22:26  <substack>I'd like to figure out how to build modules that you can just require() that handle a bunch of shadow dom / scoped css automatically
04:22:50  <defunctzombie>substack: I currently do it with a combination of bundling js for js
04:22:57  <defunctzombie>and npm-css for a single widgets.css file
04:23:00  <defunctzombie>which scopes the css
04:23:23  <defunctzombie>it also has be benefit that you can override css if you want or selectively use any widget css or not
04:23:27  <defunctzombie>and it can be deployed today
04:23:38  <defunctzombie>I am not sold on shadow dom yet
04:23:45  <dominictarr>Raynos: scuttlebutt is only memory heavy if you abuse it's assumption that your application is split into documents
04:24:15  <substack>defunctzombie: I think the <x-tag> stuff is silly but scoped css is really great
04:24:15  <dominictarr>like in regular document database
04:24:30  <substack>there are apis to control whether the css inherits from the document or not
04:24:42  <defunctzombie>substack: you can do scoped css now just by putting it in a class
04:24:58  <defunctzombie>the problem is that I want css to inherit from document
04:25:04  <defunctzombie>because I want to style things
04:25:12  <substack>not quite cleanly though
04:25:24  <substack>you can't turn off outside influence from the rest of the dom
04:25:26  <dominictarr>if you have an potentially 'infinite' document (like a news feed), you'd have to split that into many documents in mongo - same with scuttlebutt.
04:25:39  <dominictarr>level-scuttlebutt allows you to handle that.
04:25:44  <defunctzombie>substack: correct, and I think that is a feature more than a bug (at least currently)
04:25:59  * CoverSlidequit (Ping timeout: 256 seconds)
04:26:02  <defunctzombie>I want outside influence to leak in, but inside css to not leak out
04:26:30  <defunctzombie>having a widget that you can't outside style is kinda boring
04:26:38  <substack>another issue is that query selectors could be picking off your internal elements
04:26:48  <substack>that's where the shadow dom stuff is really powerful
04:27:01  * substack&
04:27:50  <defunctzombie>substack: maybe use more specific selectors?
04:28:12  <defunctzombie>I guess I don't have 100% answers :) just methods I have found to be "good enough" that I can deploy today
04:28:23  <defunctzombie>without making everything crazier
04:32:28  * CoverSlidejoined
04:46:01  * substackfg
04:47:00  <defunctzombie>bg
04:50:01  <st_luke>jobs
04:51:30  * Number9quit (Ping timeout: 256 seconds)
04:57:00  * defunctzombiechanged nick to defunctzombie_zz
05:04:03  * CryptoQuickjoined
05:05:54  * devaholicquit (Ping timeout: 251 seconds)
05:09:43  * CryptoQuickquit (Ping timeout: 245 seconds)
05:21:35  <Raynos>Dominictarr: all history is forever kept.
05:22:01  <dominictarr>Raynos: huh?
05:22:13  <dominictarr>you mean it's disk memory heavy?
05:27:13  * AvianFluquit (Remote host closed the connection)
05:56:23  * st_lukequit (Remote host closed the connection)
06:35:36  * shamaquit (Quit: shama)
06:45:31  * mikealjoined
07:49:15  * jden_joined
07:49:39  * jden_changed nick to Guest22241
08:04:08  * thatguydanquit (Ping timeout: 252 seconds)
08:08:33  * slaskisjoined
08:11:03  <Raynos>dominictarr: the fact that scuttlebutt generally keeps the entire transaction history in memory
08:11:11  <Raynos>means it's not good for something like heartbeats
08:34:21  * Altreusjoined
08:35:12  * thatguydanjoined
08:50:20  * thatguydanquit (Quit: thatguydan)
08:51:55  * thatguydanjoined
09:11:55  * No9joined
10:00:18  * thatguydanquit (Quit: thatguydan)
10:28:13  * Num9joined
10:31:08  * No9quit (Ping timeout: 255 seconds)
11:11:49  * Guest22241quit (Ping timeout: 240 seconds)
11:24:07  <dominictarr>Raynos: that isn't true
11:31:55  * fotoverite_joined
11:33:44  * fotoveritequit (Ping timeout: 256 seconds)
11:33:44  * fotoverite_changed nick to fotoverite
11:36:34  * thatguydanjoined
11:38:06  * jden_joined
11:38:30  * jden_changed nick to Guest75189
11:47:23  * slaskis_joined
11:49:36  * Guest75189quit (Ping timeout: 245 seconds)
11:49:57  * slaskisquit (Ping timeout: 240 seconds)
12:05:00  <dominictarr>every scuttlebutt has some sort of garbage collection stragety
12:15:39  * jden_joined
12:16:02  * jden_changed nick to Guest29884
12:28:18  * Guest29884quit (Ping timeout: 252 seconds)
12:52:57  * dominictarrquit (Quit: dominictarr)
12:54:07  * jden_joined
12:54:31  * jden_changed nick to Guest65299
13:02:16  * yorickjoined
13:02:16  * yorickquit (Changing host)
13:02:16  * yorickjoined
13:03:11  * ins0mniajoined
13:06:48  * Guest65299quit (Ping timeout: 252 seconds)
13:27:34  * AvianFlujoined
13:32:32  * jden__joined
13:35:46  * yorickquit (Remote host closed the connection)
13:45:12  * jden__quit (Ping timeout: 256 seconds)
13:46:53  * yorickjoined
14:11:12  * jden_joined
14:11:35  * jden_changed nick to Guest16164
14:21:00  * defunctzombie_zzchanged nick to defunctzombie
14:23:44  * Guest16164quit (Ping timeout: 256 seconds)
14:33:51  * heathquit (Remote host closed the connection)
14:38:41  * pikpik_changed nick to pikpik
14:39:41  * slaskisjoined
14:40:10  * mikealquit (Ping timeout: 256 seconds)
14:41:40  * slaskis_quit (Ping timeout: 248 seconds)
14:42:46  * mikealjoined
14:43:21  * heathjoined
14:45:29  * heathquit (Client Quit)
14:46:50  * heathjoined
14:50:19  * jden__joined
14:52:14  * defunctzombiechanged nick to defunctzombie_zz
14:58:22  * evboguejoined
15:02:46  * jden__quit (Ping timeout: 252 seconds)
15:17:47  * st_lukejoined
15:20:56  <jesusabdullah>substack: https://gist.github.com/jesusabdullah/034c5352b2a0a89059c3
15:21:46  <jesusabdullah>people are dicks
15:29:01  * jden_joined
15:29:25  * jden_changed nick to Guest54162
15:41:38  * Guest54162quit (Ping timeout: 252 seconds)
15:45:50  * defunctzombie_zzchanged nick to defunctzombie
15:46:28  <jesusabdullah>https://github.com/jesusabdullah/hoarders#for-those-of-you-with-no-sense-of-humor having to do this makes me ;_;
15:50:15  * devaholicjoined
15:51:08  * thatguydanquit (Quit: thatguydan)
16:05:29  * CoverSlidequit (Read error: Operation timed out)
16:06:18  <defunctzombie>jesusabdullah: I don't think that is really bad, they just want real modules depending on their stuff I guess.
16:06:41  <defunctzombie>I mean, a registry isn't meant to be some sort of joke imho
16:06:49  <defunctzombie>and a place for people to publish their work for others to use
16:07:29  <defunctzombie>jesusabdullah: I know I will sometimes look at what module depends on to see what could be interesting
16:07:34  <defunctzombie>and having fake modules just adds noise
16:07:40  * jden_joined
16:07:43  <defunctzombie>to an already pretty noisy system :)
16:08:04  * jden_changed nick to Guest289
16:09:21  <jesusabdullah>I mean I understand that people may not want to play and that's fine
16:09:28  <jesusabdullah>just don't be a dick about it y'know?
16:09:57  <jesusabdullah>https://github.com/jesusabdullah/hoarders/pull/8 nice enough dude
16:10:11  <jesusabdullah>https://github.com/jesusabdullah/hoarders/pull/9 passive-aggressive butt
16:12:10  <defunctzombie>yea, I know people are mean
16:12:12  <defunctzombie>:(
16:12:22  <defunctzombie>isaacs: wow, that partial tree install issue is worse than I thought actually
16:12:48  * CoverSlidejoined
16:12:56  <defunctzombie>isaacs: basically, I had a dependency specified at my top level of some version N.N.N
16:13:05  <defunctzombie>and some other modules in the tree depended on that same version
16:13:10  <defunctzombie>so npm installed it once at the top level
16:13:27  <defunctzombie>when I updated the top level dep to N.N+1.N
16:13:32  <defunctzombie>and ran npm install
16:13:38  <defunctzombie>those other deps are broken
16:13:44  <defunctzombie>they will load the wrong version of the module
16:15:04  <defunctzombie>and no amount of npm install will fix it, you have to blow away node_modules and start over
16:15:32  <defunctzombie>not only that, but it leaves people with installs which are completely busted
16:19:15  * Guest289quit (Ping timeout: 248 seconds)
16:31:24  * st_lukequit (Remote host closed the connection)
16:45:54  * jden__joined
16:50:16  * mikealquit (Quit: Leaving.)
16:53:39  * mikeal1joined
16:57:36  * jden__quit (Ping timeout: 256 seconds)
16:59:06  * No9joined
16:59:19  * Num9quit (Ping timeout: 260 seconds)
17:06:11  * shamajoined
17:06:26  * shamaquit (Client Quit)
17:06:35  * shamajoined
17:06:53  * shamaquit (Client Quit)
17:09:34  * shamajoined
17:10:41  * shamaquit (Client Quit)
17:10:53  * shamajoined
17:10:58  * shamaquit (Client Quit)
17:11:12  * shamajoined
17:12:18  * devaholicquit (Ping timeout: 240 seconds)
17:12:46  * fotoverite_joined
17:12:46  * fotoveritequit (Read error: Connection reset by peer)
17:12:46  * fotoverite_changed nick to fotoverite
17:13:23  * shamaquit (Remote host closed the connection)
17:13:34  * shamajoined
17:13:53  * fotoveritequit (Read error: Connection reset by peer)
17:14:47  * fotoveritejoined
17:22:57  * mikolalysenkojoined
17:25:35  * CryptoQuickjoined
17:28:21  <isaacs>defunctzombie: you could also install the other dep again.
17:28:40  <isaacs>defunctzombie: or: npm explore other-dep -- npm install first-dep
17:28:54  <defunctzombie>isaacs: these are all excuses :/
17:29:00  <defunctzombie>npm install should have done the right thing
17:29:06  * jden_joined
17:29:30  * jden_changed nick to Guest72096
17:30:21  <defunctzombie>isaacs: to a user (and not the creator of npm) it would be completely non obvious wtf happened
17:30:24  <defunctzombie>and why an npm install
17:30:30  <defunctzombie>whcih in typical cases works as expected
17:30:41  <isaacs>jesusabdullah: how is it "inflating his downloads"??
17:30:42  <defunctzombie>fails to install what the package.json says for the dependencies
17:30:58  <isaacs>jesusabdullah: does he think that people are actually installing hoarders?
17:31:13  <isaacs>defunctzombie: agreed. patch welcome, please.
17:31:19  <defunctzombie>has anyone actually tried to npm install hoarders?
17:31:24  <defunctzombie>that would be a nightmare I imagine haha
17:31:49  <defunctzombie>isaacs: I will try to dig into it, some of the code is a bit confusing to me atm :)
17:32:27  * mikeal1quit (Quit: Leaving.)
17:36:18  * Guest72096quit (Ping timeout: 276 seconds)
17:36:35  <Raynos>dominictarr: ive not seen a scuttlebutt with GC yet
17:51:32  * mikolalysenkoquit (Ping timeout: 252 seconds)
17:52:58  * nk109joined
18:01:34  * jden__joined
18:02:10  * mikolalysenkojoined
18:05:12  <chrisdickinson>isaacs: is anyone directly involved with node.js on TC39?
18:06:36  <isaacs>chrisdickinson: i don't think so.
18:06:41  <chrisdickinson>hmm
18:06:45  <isaacs>we all have jobs, and also node.
18:06:57  * chrisdickinsonunderstands
18:07:31  <isaacs>domenic seems to follow es-discuss pretty closely
18:07:41  <isaacs>there are discussing participants that use node, for sure
18:08:45  <chrisdickinson>i was just wondering, since (and i might be totally mistaken) there seems to be an attitude from TC39 that node.js is a "vocal minority"
18:09:07  <chrisdickinson>or at least, that it's okay to brush off comments from node.js devs because we're perceived as being the minority.
18:11:18  <chrisdickinson>(this is at odds with the fact that frontend devs will have to wait for universal browser support to really start using ES6; so for a long while the only folks that might be able to use the new features are extension developers and the node.js community.)
18:12:49  * jden__quit (Ping timeout: 248 seconds)
18:13:07  <mikolalysenko>In npm, is it possible to have cyclic devDependencies?
18:13:19  <mikolalysenko>For example, I have this library rle-core and another library rle-sample
18:13:35  <mikolalysenko>I want to use rle-sample to generate test data for rle-core, but rle-sample depends on rle-core
18:14:04  <mikolalysenko>I could fuse the libraries, but rle-sample is really separate from rle-core
18:16:32  <CoverSlide>run-length encoding?
18:23:59  <defunctzombie>mikolalysenko: sure, but I wouldn't do that
18:24:11  <defunctzombie>I would make rle-core not depend on anything it doesn't need to
18:24:21  <defunctzombie>or maybe depend on rle-sample via devDependencies
18:24:31  <defunctzombie>if it isn't needed at regular runtime, don't put it in dependencies
18:25:26  * defunctzombiechanged nick to defunctzombie_zz
18:25:35  * nk109quit (Quit: laters...)
18:31:33  * devaholicjoined
18:34:24  <niftylettuce>hey folks
18:35:05  <CoverSlide>hello niftylettuce
18:35:30  <niftylettuce>CoverSlide: you from SoCal?
18:35:34  <CoverSlide>yes
18:36:23  * mikolalysenkoquit (Ping timeout: 260 seconds)
18:39:13  * jden_joined
18:39:37  * jden_changed nick to Guest67446
18:44:10  * st_lukejoined
18:50:48  * Guest67446quit (Ping timeout: 252 seconds)
18:56:09  * st_lukequit (Remote host closed the connection)
19:10:52  * dguttmanjoined
19:15:48  * dguttman_joined
19:19:22  * blobaumjoined
19:20:34  * yorick_joined
19:20:35  * yorick_quit (Changing host)
19:20:35  * yorick_joined
19:24:54  * dguttmanquit (*.net *.split)
19:24:54  * devaholicquit (*.net *.split)
19:24:54  * yorickquit (*.net *.split)
19:24:55  * dguttman_changed nick to dguttman
19:25:45  * defunctzombie_zzchanged nick to defunctzombie
19:29:15  * mikolalysenkojoined
19:33:50  * mikolalysenkoquit (Ping timeout: 252 seconds)
20:01:20  * tphummeljoined
20:01:45  * mikolalysenkojoined
20:01:45  <mikolalysenko>defunctZombie: Ok, I can put it in as a devDependency, I don't need it except to generate test data
20:02:34  <Raynos>chrisdickinson: its not a vocal minority, its just a vocal dickish community
20:03:00  <Raynos>miko: I have cyclic deps all the time
20:04:20  <mikolalysenko>Raynos: fair enough. I've always figured cyclic dep => packages should be merged
20:04:31  <mikolalysenko>though in this case the things really should be split...
20:06:40  * No9quit (Quit: Leaving)
20:07:48  * evboguequit (Ping timeout: 240 seconds)
20:10:41  * yorick_changed nick to yorick
20:17:37  <Raynos>well
20:17:45  <Raynos>I have the same problem with topology & signal-channel
20:17:53  <Raynos>its a modular version of webrtc but the examples are coupled
20:24:44  * mikolalysenkoquit (Ping timeout: 252 seconds)
20:28:23  * blobaumquit (Ping timeout: 255 seconds)
20:30:57  * mikolalysenkojoined
20:31:31  <defunctzombie>substack: you around?
20:35:18  * CoverSlidequit (Ping timeout: 240 seconds)
20:35:52  * slaskisquit (Remote host closed the connection)
20:47:54  * mikolalysenkoquit (Ping timeout: 260 seconds)
20:53:03  * slaskisjoined
21:12:31  * mikolalysenkojoined
21:18:05  * ryan_stevensjoined
21:28:21  * mikolalysenkoquit (Ping timeout: 276 seconds)
21:30:12  * mikolalysenkojoined
21:35:04  * mikealjoined
21:39:50  * slaskisquit (Ping timeout: 260 seconds)
21:44:35  <defunctzombie>isaacs: npm is too complex :(
21:44:41  <isaacs>defunctzombie: agreed.
21:45:05  <defunctzombie>I am not sure I even understand what the install code is trying to do :|
21:45:47  <defunctzombie>I believe what I want to replace is the "readDependencies" call with readInstalled
21:46:11  <defunctzombie>but then read-installed returns some sort of 'data' which I don't really see documented what form that takes
21:46:27  <defunctzombie>or which of the installMany* calls I would use
21:48:46  <defunctzombie>isaacs: alternatively, can I just recurse on each dependency and run install on it?
21:49:00  <defunctzombie>maybe like what install does during the first time you run it with an empty tree
21:49:21  <isaacs>defunctzombie: definitely read the code and understand it before making changes.
21:49:25  <isaacs>it's not trivial.
21:49:48  <isaacs>it's doing some fancy pants recursive tree walking stuff, and making decisions based on the info it has at each level.
21:49:57  <defunctzombie>I feel like it evolved out of control and has taken on some sort of form
21:50:19  <isaacs>defunctzombie: it's not that complicated, really
21:50:27  <defunctzombie>the install algo itself shouldn't be complex, just a recursive walk
21:50:34  <isaacs>defunctzombie: "just" a recursive walk.
21:50:36  <isaacs>that's cute.
21:50:43  <isaacs>it's a recursive walk over a tree that it's building while it walks.
21:50:49  <isaacs>so... yeah. it's going to get a *bit* hairy
21:51:05  <defunctzombie>I don't see why, the tree builds but children can reference parents
21:51:06  <isaacs>also, we want to default to being relatively deduped (though not entirely, since we can't plan)
21:51:15  <defunctzombie>it isn't like it builds in some unpredictable way
21:51:15  <isaacs>children can reference parents or uncles
21:51:19  <isaacs>or siblings
21:51:34  <isaacs>but not cousin
21:51:35  <isaacs>s
21:51:45  <isaacs>lunch time. be back in a few hours.
21:57:14  * st_lukejoined
21:57:42  * mikealquit (Quit: Leaving.)
22:02:53  * mikolalysenkoquit (Ping timeout: 255 seconds)
22:14:43  * evboguejoined
22:30:17  * blobaumjoined
22:46:08  * CryptoQuickquit (Ping timeout: 272 seconds)
22:47:54  <st_luke>mbalho: when you were doing geo stuff for cities did you ever do like neighborhood boundary lat/long data?
22:53:46  * thatguydanjoined
22:57:28  <isaacs>substack: where you flying to?
22:57:41  <isaacs>substack: er, where have you flown to?
22:58:46  * defunctzombiechanged nick to defunctzombie_zz
23:02:49  <mbalho>st_luke: yea but its hard to get nation wide data, twitter actually bought a company once that just did that
23:02:57  * Madarsquit (Quit: reboot)
23:06:27  * No9joined
23:07:16  * evboguequit (Read error: Connection reset by peer)
23:10:51  <st_luke>mbalho: ah not surprising i guess
23:11:15  <st_luke>anyone in nyc want a giant box of brand new free clothing that amazon mistakenly sent me?
23:11:34  <fotoverite>What size?
23:11:49  <st_luke>fotoverite: large
23:11:59  <fotoverite>Nope won't fit
23:12:12  <st_luke>it seems like ok stuff, i haven't opened it, it's just been sitting in the box in my apartment for a couple weeks now
23:12:20  * mikolalysenkojoined
23:13:45  <fotoverite>http://www.designinghypermediaapis.com/
23:13:54  <fotoverite>Sorry ignore
23:14:09  <st_luke>HYPERMEDIA
23:14:41  <fotoverite>THE FUTURE!
23:16:38  <st_luke>LOUDBOT: twitlast
23:16:38  <LOUDBOT>http://twitter.com/LOUDBOT/status/301469893054844928 (Tittiephish/#perl)
23:18:27  * ralphtheninjajoined
23:22:17  * tilgovijoined
23:33:51  * ryan_stevensquit (Quit: Leaving.)
23:35:31  * defunctzombie_zzchanged nick to defunctzombie
23:50:14  * thatguydanquit (Quit: thatguydan)
23:55:05  * st_lukequit (Remote host closed the connection)
23:56:25  * dominictarrjoined
23:56:42  * ryan_stevensjoined
23:58:02  * No9quit (Read error: Operation timed out)