#dat

[dat] mafintosh pushed 1 new commit to beta: http://git.io/A9Sb

dat/beta 64a0b88 Mathias Buus: update the beta cli api

ogd: we might wanna add a batch api as well to the js api

mafintosh: oh yea

batches should replicate atomically

[dat] maxogden pushed 1 new commit to master: http://git.io/AHa8

dat/master 42f482e Max Ogden: Update getting-started.md

ogd: we forgot `dat init`

i'll add it

mafintosh: lol oh yea. details

as well as replication hah

oh yea haha

so we're going with dat-wide checkout?

(as opposed to dataset-wide)

karissa: yea

i think that makes sense

will it prevent `dat checkout <dataset> <hash>`?

or is it an either-or scenario

karissa: either-or

hmm

ok

karissa: because the idea behind 'datasets' was to have multiple related datasets in the same dat

yeah

karissa: so IMO the whole thing should update together

karissa: otherwise you can just make another dat

yeah

karissa: also for e.g. publication you only have to publish one hash

right

[dat] mafintosh pushed 1 new commit to beta: http://git.io/AHo0

dat/beta 2cf672c Mathias Buus: add init, push, pull and sync

i think we need to make this clear so that people don't build the incorrect mental model

ogd, karissa you can create 2 dat instances and use different checkouts on each of if you *really* want to checkout datasets at different commits

in JS

in CLI you can only have one checkout at a time

you could execute two different, dat rows get --checkout my-hash

oh yea but thats the stateless checkout so its cheating

heheh

I think it makes sense. For help in the UI we could find which datasets were modified on that particular hash

karissa: currently we only modify one dataset per commit

mafintosh: perfect

it might modify multiple branches of that dataset though if it is a merge

aye

but commits below that commit might be different datasets though

right..

hmm

i see

just sketching the PR process and thinking about the kind of diff people will need to make accurate decisions

mafintosh: theres a really cheap flight ($80) from SJC to PDX at 7am on apr 13th, otherwise its 2x the price for later in the day

mafintosh: (also non-direct later)

mafintosh: otherwise we could do 10pm flight on sunday night for ~$140

karissa: me and mafintosh will be in the bay from the 8th - 13th now, 11th-12th will be hacking w/ the IPFS folks

karissa: also im setting up a talk at 18F in SF hopefully on the 9th

ogd: nice!

ogd: what date is the sunday flight?

mafintosh: apr 12

cph -> oak is 250 usd now, wow

mafintosh: which airline?

ogd: either is fine but 80$ is really cheap though

ogd: norwegian

one way

mafintosh: oh i didnt know norwegian did low far out of cph, nice

ogd: you need to transfer in stockholm.

ogd: but its not that bad

mafintosh: but even w/ the transfer its still only $250?

ogd: yea!

dayum

whoa! if you fly directly from stockholm its 167 usd!!

lol

ogd: its actually cheaper for me in total to fly to portland first, then fly pdx->oak, oak->pdx and then finally pdx->cph

because roundtrips are cheaper than direct flights

mafintosh: oh yea

mafintosh: you can stay at our house as long as you want in pdx

mafintosh: we could also go up to seattle/vancouver

mafintosh: the train journeys are good for hacking. and there are fjords

ogd: never been up there so am definitely up for that