#dat

mafintosh: ogd: ready to push out https://usopendata.org/2015/07/29/dat-beta/ on twitter?

mafintosh: ogd oh waldo already did! K

lol

ogd: didn't see it on @dat_project, but it was on @opendata ;)

ogd: gonna RT

ogd: can you dat clone http://npm.dathub.org ?

karissa: yea

ogd: mine is stalling

ogd: glargs

ogd: oh or its just taking a while?

ogd: hmm ok its working but the progress looks like its not going anywhere.. 2800... 5-10 seconds later, 6000

hmm it seems ok for me

ogd: k i'll tell the guy then in that issue

ogd: did you see this? https://github.com/maxogden/dat/issues/361

oh yea ill try and respond

[dat] karissa closed issue #353: "dat clone http://npm.dathub.org --skim" no longer works -- "Error: Invalid remote" http://git.io/vqhZx

(arfon) Congratulations folks on the beta release!

:o

mafintosh: reading the dat announcement, I had another idea for where a hyperlog signed application feed might be very useful: reproducible science

if the computation on a dataset is in javascript, the calculations and graphs can be put in a merkle dag along with the data

substack: throw the whole linux VM in dat preloaded with your science

good but heavy

substack, ogd: but lazy load the vm using something like torrent-docker

:P

substack: yea signed feeds are gonna be a big deal

signed feeds will be huge for deployment and auditing

because they can be public and a bot can consume them to trigger redeploys

substack: i'm factoring out the signing stuff now from the bigger update so we can get it out faster

the other update also changes the replication protocol but that can wait until after the signing stuff is stable

cool thanks!

idea for using dat-npm:

we make a npm install cli tool that just uses that dat-npm data to offline figure out exactly which cached deps are out of date

that'll be insanely faster than waiting for a million 302s from the npm servers

also dat-npm is only 0.5 gb on disk!

brb - biking home

does anyone know off hand what number `dat clone http://npm.dathub.org` counts up to? It's at 150k right now, but I don't know what the denominator is :-)

oh, answer: 168k

boogs: heh

[dat] stefanw opened pull request #362: Fix usage filename to fit status command (master...patch-1) http://git.io/vYAMF

[dat] maxogden pushed 2 new commits to master: http://git.io/vYAD8

dat/master d75dde3 Stefan Wehrmeyer: Fix usage filename to fit status command

dat/master f4828b1 =^._.^=: Merge pull request #362 from stefanw/patch-1...

[dat] maxogden pushed 1 new commit to master: http://git.io/vYAyn

dat/master 09dc79c Max Ogden: try and fix travis yml

maxogden/dat - try and fix travis yml - passed - https://travis-ci.org/maxogden/dat/builds/73272232

whoa cool

Yay the beta is out \o/

Hello. Congratulations on the beta release! I'm working on an open data project to develop a pipeline to download, transform, clean and republish a big (ish) government database and I'm curious if dat might work for your project. Anyone have a minute? You can learn more about our gig here. http://www.californiacivicdata.org/2015/07/22/k...

er. I'm curious if dat might work for *our* project.

palewire: HEY i have been working on a reply to you lol

palewire: now that we shipped :)

excellent!

no hurry really just saw your announcement today and thought I'd check in

though while I'm here I should say we're hiring a developer

palewire: i was hanging out with @aboutaaron recently at srccon

and if anybody in here is interested they should email me at ben.welsh@gmail.com

oh great!

so he probably filled you in

palewire: actually only super briefly, we should definitely sync up

basically, we want to use this project as a chance to try and push then envelope on how these government data clean up efforts go

and get beyond closed shop kludging to make a point and click search

and instead make a high-quality data product people can grip and rip

without the gigantic hassle of untangling this thing

palewire: haha awesome!

so my facile thought is that maybe dat is a way for us to do thi

or one of the things we should try

beyond just bulk CSV downloads, docs, API, etc

palewire: karissa from our team lives near aboutaaron in the east bay i think

oh great

he is in Emoryville

yea perfect

palewire: well im happy to jump on a google hangout to chat in high-bandwidth some time if that works for you

sure. what's a good time for you?

palewire: any time tomorrow is good, im down in the bay area for a conf friday - sunday, available monday too

Heroku is restricting their free dynos in August. They have to be inactive for 6 hours a day

bad news for my hubot

finnp: ohhh i didnt realize they hadn't implemented that yet.

finnp: maybe we should replace our 'deploy to heroku' example with a CLI tool that deploys a dat to digital ocean for you

finnp: then at least its a real hard drive

ogd: yeah we should do that.

mafintosh: hey check this out https://github.com/PeerioTechnologies/peerio-cl...

finnp: i wonder if digital ocean has a cli oauth workflow (e.g. copy paste a pin to login)

ogd: name is a bit confusing since it doesnt seem to be p2p but seems sleek

mafintosh: haha yea good point

mafintosh: kinda reminds me of how dch earlier was saying 'peer to peer' got co-opted by VOIP calls hence the need for his term 'multipeer'

peer-to-a-lot-of-peers

ogd: they use minilock! is that how you found it?

mafintosh: bengl on twitter linked me

ogd: yeah great idea re: digital ocean

what if we called it like `dat-deploy` and it let you either install an empty dat on an existing server, create a new droplet and install an empty dat on it, or push an existing dat up to a server