asheesh changed the topic of #sandstorm to: Welcome to #sandstorm: home of all things sandstorm.io. Say hi! | Channel glossary: "i,i" means "I have no point, I just want to say". b == thumbs up. | Public logs at https://botbot.me/freenode/sandstorm/ & http://logbot.g0v.tw/channel/sandstorm/today
<zarvox> It sends the grain owner an email, indicating who requested access and which grain they asked you to share
<tannercollin> ahhh, I haven't set up SMTP yet, unfortunately. I'll probably do that once my midterms are over.
frigginglorious has joined #sandstorm
frigginglorious has quit [Client Quit]
adas has quit [Ping timeout: 250 seconds]
frigginglorious has joined #sandstorm
frigginglorious has quit [Client Quit]
frigginglorious has joined #sandstorm
joshbuddy has quit [Quit: joshbuddy]
_keturn has joined #sandstorm
joshbuddy has joined #sandstorm
_keturn has quit [Ping timeout: 246 seconds]
_keturn has joined #sandstorm
KooBaa has joined #sandstorm
aldeka has quit [Remote host closed the connection]
ocdtrekkie has quit [Remote host closed the connection]
KooBaa has quit [Ping timeout: 276 seconds]
jacksingleton has quit [Ping timeout: 248 seconds]
joshbuddy has quit [Quit: joshbuddy]
ocdtrekkie has joined #sandstorm
aldeka has joined #sandstorm
joshbuddy has joined #sandstorm
joshbuddy has quit [Read error: Connection reset by peer]
joshbuddy has joined #sandstorm
joshbuddy has quit [Client Quit]
|jemc| has joined #sandstorm
frigginglorious has quit [Quit: frigginglorious]
|jemc| has quit [Quit: WeeChat 1.4]
|jemc| has joined #sandstorm
KooBaa has joined #sandstorm
jadewang has joined #sandstorm
KooBaa has quit [Ping timeout: 276 seconds]
kecors has quit [Quit: Leaving]
jacksingleton has joined #sandstorm
jacksingleton has quit [Ping timeout: 250 seconds]
mnutt__ has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
NwS has joined #sandstorm
KooBaa has joined #sandstorm
home has joined #sandstorm
|jemc|_ has joined #sandstorm
|jemc| has quit [Ping timeout: 248 seconds]
KooBaa has quit [Ping timeout: 276 seconds]
|jemc|_ is now known as |jemc|
|jemc| is now known as jemc
jemc is now known as |jemc|
|jemc| is now known as jemc
jemc has quit [Quit: WeeChat 1.4]
jemc has joined #sandstorm
jemc is now known as jemc_
jemc_ is now known as jemc
jadewang has quit [Remote host closed the connection]
jadewang has joined #sandstorm
jemc has quit [Ping timeout: 246 seconds]
tannercollin has quit [Ping timeout: 276 seconds]
jadewang has quit [Remote host closed the connection]
joshbuddy has joined #sandstorm
joshbuddy has quit [Client Quit]
KooBaa has joined #sandstorm
joshbuddy has joined #sandstorm
joshbuddy has quit [Client Quit]
tobald has joined #sandstorm
KooBaa has quit [Ping timeout: 276 seconds]
tobald has quit [Remote host closed the connection]
tobald has joined #sandstorm
ecloud has joined #sandstorm
digitalcircuit_ has joined #sandstorm
digitalcircuit_ is now known as digitalcircuit
synchrone has joined #sandstorm
KooBaa has joined #sandstorm
joshbuddy has joined #sandstorm
mcpherrin has joined #sandstorm
fonfon has joined #sandstorm
home has quit [Remote host closed the connection]
joshbuddy has quit [Quit: joshbuddy]
KooBaa has quit [Ping timeout: 276 seconds]
fonfon has quit [Remote host closed the connection]
myguidingstar has joined #sandstorm
asmyers has joined #sandstorm
dvn has joined #sandstorm
asmyers has quit [Remote host closed the connection]
amyers has joined #sandstorm
KooBaa has joined #sandstorm
synchrone has quit [Ping timeout: 252 seconds]
TwoJames2 has joined #sandstorm
TwoJames has quit [Ping timeout: 276 seconds]
amyers has quit [Remote host closed the connection]
amyers has joined #sandstorm
amyers has quit [Read error: Connection reset by peer]
amyers has joined #sandstorm
KooBaa has quit [Ping timeout: 244 seconds]
jon1012 has joined #sandstorm
synchrone has joined #sandstorm
dwrensha has quit [Remote host closed the connection]
dwrensha has joined #sandstorm
synchrone has quit [Read error: Connection reset by peer]
frigginglorious has joined #sandstorm
synchrone has joined #sandstorm
synchrone has quit [Ping timeout: 244 seconds]
KooBaa has joined #sandstorm
fonfon has joined #sandstorm
synchrone has joined #sandstorm
jemc has joined #sandstorm
KooBaa has quit [Ping timeout: 240 seconds]
notevil has joined #sandstorm
fonfon has quit [Ping timeout: 264 seconds]
wolcen has quit [Quit: Leaving]
wolcen has joined #sandstorm
joshbuddy has joined #sandstorm
tobald has quit [Quit: Ex-Chat]
_whitelogger has quit [Ping timeout: 264 seconds]
_whitelogger_ has joined #sandstorm
jemc_ is now known as jemc
notevil has quit [Quit: Leaving]
sknebel_ is now known as sknebel
wolcen has quit [Ping timeout: 276 seconds]
tannercollin has joined #sandstorm
joshbuddy has quit [Quit: joshbuddy]
notevil has joined #sandstorm
amyers has quit [Read error: Connection reset by peer]
amyers has joined #sandstorm
_whitelogger_ has quit [Excess Flood]
_whitelogger has joined #sandstorm
jon1012 has quit [Ping timeout: 260 seconds]
amyers has quit [Read error: Connection reset by peer]
amyers has joined #sandstorm
joshbuddy has joined #sandstorm
KooBaa has joined #sandstorm
synchrone has quit [Ping timeout: 260 seconds]
mnutt__ has joined #sandstorm
joshbuddy has quit [Quit: joshbuddy]
KooBaa has quit [Remote host closed the connection]
KooBaa has joined #sandstorm
indiebio has joined #sandstorm
joshbuddy has joined #sandstorm
_whitelogger has quit [Excess Flood]
_whitelogger has joined #sandstorm
<ocdtrekkie> Wow, so, the government is giving Eric Schmidt a job now too. Really amazed at how corrupt the relationship between Google and the Obama administration is.
<pdurbin> ocdtrekkie: oh? how so?
<ocdtrekkie> asheesh: That looks cool and I would use that on Sandstorm.
<ocdtrekkie> pdurbin: Eric Schmidt is heading up an 'innovation board' at the Pentagon.
<ocdtrekkie> Also, asheesh, the pre-Android 2.0 screenshots. Love it.
<pdurbin> not sure what corrupt about that but ok
frigginglorious has quit [Quit: frigginglorious]
<ocdtrekkie> pdurbin : Almost every tech position the government has hired is a Googler. Including having made their FTC commissioner someone paid by Google to fabricate 'academic studies' that made them look good. Schmidt is their evangelist they send to convince governments to do what Google wants. It's impossible to suggest he'd place our country's interests above his employers.
<pdurbin> I wonder if it's because a Googler saved healthcare.gov.
<ocdtrekkie> Perhaps. An incredibly good return on investment for Google if so.
<pdurbin> there was a nice article in Time about all this
frigginglorious has joined #sandstorm
wolcen has joined #sandstorm
wolcen has quit [Client Quit]
wolcen has joined #sandstorm
synchrone has joined #sandstorm
<mrdomino> i,i the Googler who saved healthcare.gov was my boss while i was there
<dwrensha> nice: "WARNING: `vagrant-spk up` is deprecated and will be removed on August 1, 2016."
<jemc> probably has more to do with campaign contributions: https://www.opensecrets.org/pres12/contrib.php?id=N00009638
<jemc> and another way to interpret "saved healthcare.gov" is "got a nice big government contract"
<synchrone> dwrensha: vagrant-spk? why? where's that coming from?
<dwrensha> synchrone: the rest of the warning reads: "Prefer `vagrant-spk vm up`."
<synchrone> ah, I see
<synchrone> speaking of which
<synchrone> how could I proxy my ENV inside vagrant, inside `dev` command, to check for in my launcher.sh ?
<synchrone> I kind of want to handle `DEBUG=true vagrant-spk dev` scenario
<synchrone> to enable e.g debug-mode FE resource pipeline, or python pydevd debugger module startup
<synchrone> currently I have to patch the scripts\uncomment debugger in the sources
<zarvox> hmmm, good questions!
<synchrone> the way I see it, vagrant inherits from ssh a need for an explicit list of ENV var names
<zarvox> at the moment, I think the easiest thing would be "edit build.sh and set your boolean in a variable at the top"
<zarvox> and then we can try to figure out how to cause that variable to be set in the guest based on command line args or host environment
<zarvox> I'm sympathetic to the need for "dev mode" faster iteration, but there's also the unfortunate fact that spk dev traces file accesses to figure out what should be included in your package
<zarvox> so if done naively, you'd wind up including your entire build/hot-reload toolchain in your SPK
<synchrone> yeah, I'm already including vagrant-spk destroy + fresh dev run in production mode in my `release` process
<dwrensha> zarvox: so delete the .sandstorm-files list after playing around in dev mode
<synchrone> so I guess this will progress into git a separate subtree for production-mode builds
<zarvox> dwrensha: and then communicate to devs that they need to play with every feature again to make sure the files get picked up in the trace
<zarvox> synchrone: +1 for making sure you can build the package cleanly :)
<synchrone> +1 for making a list of currently pending apps available under https://apps.sandstorm.io/?experimental=true :D
<zarvox> But yeah, there's this friction between "something that works in dev works exactly the same in production" and "dev mode has fast iteration" and "packages don't carry a bunch of stuff they don't need", and I'm not completely sure what the right direction to go is to balance these things
<zarvox> I do think that in the fullness of time, spk dev should do less magic tracing file access and take more explicit manifests or something
<synchrone> wasn't a super-duper-easy packaging your prio?
<dwrensha> zarvox: you can use it that way now, with alwaysInclude
<synchrone> I mean for marketing purposes, etc
<synchrone> low entry barrier
<zarvox> that's also a design goal :)
<zarvox> (meeting time, be back in a bit)
jadewang has joined #sandstorm
eglimi has joined #sandstorm
<pdurbin> mrdomino: nice! he sounded great
<mrdomino> he was. also, google was my first job out of college -- he was my first boss. tough act to follow as it turns out.
<ocdtrekkie> jemc: That is true. Google's funding a truly amazing number of Congresscritters on a regular basis now.
fonfon has joined #sandstorm
<pdurbin> mrdomino: I can imagine.
joshbuddy has quit [Quit: joshbuddy]
mnutt__ has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
joshbuddy has joined #sandstorm
joshbuddy has quit [Client Quit]
synchrone has quit [Ping timeout: 264 seconds]
eglimi has quit [Ping timeout: 264 seconds]
mnutt__ has joined #sandstorm
<chilts> zarvox: just going back a bit, you mentioned if spk dev needs to trace file access to figure out what needs to go in the package
<chilts> however, as the developer, I know already know exactly what should go in and what doesn't ... is there a way to do it manually and speed everything up?
<chilts> e.g. my server.js and everything in lib/ and node_modules/ ... easy :)
<chilts> actually, not even node_modules since that's installed later
<chilts> so server.js and lib/ (as an example)
<zarvox> chilts: sounds like you also need /usr/bin/node and all the things that it references
<chilts> I'm installing node in setup.sh
<chilts> and would /bin/bash be automatically installed
<zarvox> and probably also /bin/bash for launcher.sh
<chilts> ah wait, am I getting confused, this is just locally
<chilts> ok
<chilts> sorry
<chilts> I get it now :D
<dwrensha> there are lots of layers!
<zarvox> FWIW, you can specify paths in https://github.com/sandstorm-io/sandstorm/blob/master/src/sandstorm/package.capnp#L55 to always include in your package
<chilts> yeah, I'm definitely going through the docs with a fine toothcomb finding nuggets all the time
<zarvox> :)
<chilts> is that for the local dev package or for final distribution? I think that's where my confusion comes from
<zarvox> "yes" in the sense that we want the latter to behave precisely like the former, or else people will mistakenly publish packages that worked fine in dev mode but break when released, if they're not vigilantly testing every codepath after packaging
<chilts> so in my local dev, I want it (perhaps automatically) to install /bin/bash, but in prod that'd be in the base vagrant file (such as sandstorm/debian-jessie64)
<zarvox> sorry, lemme try again :)
<chilts> yes, colour me confused
<chilts> :)
<chilts> hmm, so wouldn't it already be there anyway even in dev
<zarvox> In dev mode, all files on the host can be found. They are mapped according to the sourceMap from the sandstorm-pkgdef.capnp .
<zarvox> The ones which are open()ed while spk dev is running get added to a list (fileList).
<zarvox> when you say "spk pack", sandstorm builds a package including all files in fileList and all files/directories in alwaysInclude.
<chilts> ah ok, hmm, I should go re-read those docs again
<chilts> funnilly enough, that's almost exactly where I am at the moment with sandstorm-pkgdef.capnp anyway :)
<zarvox> That is how things work today. The upshot is that anything your app opened while running in dev mode gets added to the packed app. (Even this is a heuristic - some code will do things based on the presence/absence/timestamp of a file without opening it <_<)
<chilts> just filling in various author and license fields
<zarvox> I kinda wish there were a way to say "all the files from these Debian packages" but then you have the transitive dependency problem, and wind up either pulling in the entire base userspace, or missing a dependency
<chilts> I think what I'm missing is : "in dev we have a base image called sandstorm/debian-jessie64, but in prod we don't have a base image as such, but just the contents of what was copied into the pkg?
<chilts> right, yeah
<chilts> hmm, so I don't actually need to install node.js in setup.sh?
<chilts> since it'll copy that (and any libs it opens) for me
<dwrensha> node.js is not on the base debian install
<chilts> yeah, but it is installed on my machine
<chilts> and in setup.sh I added the repo from deb.nodesource.com
<chilts> ah, I see
<chilts> I mis-intonated your last message :)
<chilts> anyway, I'm hoping to get a first package created in the next few days
<chilts> perhaps a week or two then!
jadewang has quit [Remote host closed the connection]
jadewang has joined #sandstorm
jadewang has quit [Remote host closed the connection]
nicolagreco has joined #sandstorm
<dwrensha> heh, apparently /bin/sh does not have a "source" command on Debian, but it does on Arch
xet7 has quit [Quit: Leaving]
<maurer> dwrensha: /bin/sh on debian is "dash"
<chilts> interesting, it's usually a symlink to bash or dash ... perhaps Debian is different to my Ubuntu :)
<maurer> dwrensha: dash is a minimalistic sh-comaptible shell
<chilts> yeah, that
<maurer> dwrensha: On arch, and most non-debian derived distros, it is usually bash
<maurer> dwrensha: I would reccomend you either make your shell sh-compatible (not worth your time) or specify that you want bash if you want bash
<maurer> s/shell/script/g
<dwrensha> ha, yeah "dash" does not mean "david's shell"
<maurer> yeah, the tl;dr is that debian is basically just being pedantic about what sh is
notevil has quit [Quit: ...]
synchrone has joined #sandstorm
<dwrensha> darn, even with gems installed in a non-synced folder, gitlab is really slow to start up in vagrant-spk
joshbuddy has joined #sandstorm
<zarvox> still has to go through fuse for the O(n^2) existence checks
amyers has quit [Ping timeout: 250 seconds]
<zarvox> also if you're using Ruby 1.9, you're probably hitting http://accidentallyquadratic.tumblr.com/post/113407169892/ruby-application-startup
<zarvox> dunno which ruby version you have
<dwrensha> 2.1.6
jadewang has joined #sandstorm
fonfon has quit [Remote host closed the connection]
jadewang has quit [Ping timeout: 244 seconds]
<dwrensha> also, debian's busybox has fewer functions defined than Arch's
<maurer> What busybox does is _wildly_ variable
<maurer> iirc some versions even include gcc?
<kentonv> wat
<maurer> I think I saw that on a router once
<maurer> a symlink to busybox called gcc
<maurer> that ran like, a local gcc-2.x series
<kentonv> zarvox: ruby startup is still quadratic despite the specific issue you mentioned having been fixed
<maurer> In any case, busybox gets the hell customized out of it
<zarvox> y'know, maybe this is the time where I ping Dave Thomas and ask him about that thing he mentioned for speeding up Ruby startup
<asheesh> Flattened bundle maybe?
<asheesh> Unrelated: "Bundling From RubyGems is Really Slow
<asheesh> " "Add the --full-index flag when bundling from the rubygems server" -- http://bundler.io/v1.11/faq.html
<asheesh> Maybe we should do that, though it's unrelated this particular problem.
xet7 has joined #sandstorm
<dwrensha> whatt "SQLite3::BusyException: database is locked"
<dwrensha> could putting the sqlite database in a snyced directory possibly have an effect like this ^?
<dwrensha> the initial database goes is built in setup.sh and then gets copied over to /var on first app boot
rustyrazorblade has joined #sandstorm
<asheesh> dwrensha: I wish I had advice for you!
<asheesh> Did you stop the database before copying it?
<asheesh> I mean, stop the Ruby processes before copying the database out from under it.
<asheesh> I presume you are, in which case, I have no idea for you yet
<dwrensha> I kind of feel like `vagrant-spk ssh` should drop me directly in /opt/app
<asheesh> dwrensha: Seems reasonable to me. /home/vagrant is a weird place to end up.
<zarvox> `vagrant-spk ssh` is deprecated and `vagrant-spk vm ssh` is the private property of `vagrant`
<dwrensha> but still
<zarvox> I guess you could set the vagrant user's homedir to /opt/app or something, but then other stuff will be weird
<zarvox> or you could put "cd /opt/app" at the end of /home/vagrant/.bashrc
<zarvox> which gives you the affordance you seek
xet7 has quit [Quit: Leaving]
neynah has joined #sandstorm
xet7 has joined #sandstorm
wolcen has quit [Ping timeout: 268 seconds]
mnutt__ has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
frigginglorious has quit [Ping timeout: 268 seconds]
jadewang has joined #sandstorm
synchrone has quit [Ping timeout: 248 seconds]
sydney_untangle has quit [Ping timeout: 276 seconds]
<asheesh> elimisteve: I sent you a bunch of emails and filed a bunch of bugs. Filed & sent in the hopes of being helpful! Happy to discuss here, or by email, etc.
<asheesh> Playing with cryptag is fun; thanks for making a package.