Live community hangout #1 - Dec 1st, 10AM PST

[max-mapper]

CALL IS OVER NOW! - The full video + audio of the call is here: http://www.youtube.com/watch?v=4sk6kcF7viM

I've scheduled our first live community hangout for December 1st at 10AM PST (everytimezone link). We want to invite anyone to ask us about what we're working on.

Leave your questions in the thread below, in our gitter web chatroom or tweet at @dat_project

Agenda:

• New stuff in Dat core (@finnp @maxogden, 15mins)
  • https://github.com/maxogden/dat/pulse/monthly
  • https://github.com/maxogden/get-dat
  • https://www.youtube.com/watch?v=9n8sGD4Qncs (dat intro talk video)
  • https://www.youtube.com/watch?v=Ef17lkx7s0U (dat and bionode talk at Sanger)
  • https://github.com/maxogden/dat-server (free dat on openshift)
  • https://github.com/maxogden/dat/issues/195
  • https://github.com/maxogden/dat/pull/220
  • https://github.com/datproject/discussions/issues/5
• Discussion on merging data, branches, and checkout (@mafintosh and @maxogden, 15mins)
• Sneak peek of the Dat Python module (@pkafei, 5 mins)
• Q/A from the community (20 mins)

We'll be using Hangouts On Air.

Note that only Dat team members and others on the agenda will be invited to join the Hangout so they can broadcast voice and video. Hangouts only supports about 10 people max, but the youtube stream can support an unlimited amount of viewers.

[taterbase]

This may be covered by the "New stuff in Dat core" portion but I've heard talk about the synchronization api changing and would love to hear more about that.

[L1fescape]

There's a lot of background noise. Hard to hear.

[max-mapper]

@akenn sorry about that!

[mafintosh]

nextgen multi-master replication

consider this graph:

c1  b2
|   |
 \ /
  |
  b1
  |
  a2
  |
  a1

it has these users: a, b, c

look at the graph as multiple logs:

a1  b1  c1
a2  b2

a replication example where you pull from the remote looks like this:

1. remote sends c:1, b:2 (heads of the graph)
2. you send c:0, b:0 (since you do not know c or b)
3. remote sends c1, b1, b2
4. you send a:0 (since b1 references a2 and you do not know a)
5. remote sends a1, a2

if you pull again it looks like this

1. remote sends c:1, b:2 (heads of the graph)
2. you have c1 and b2 so you know you're in sync

follow development here, https://github.com/mafintosh/vector-clock-causal-crawling

references: https://github.com/mafintosh/multi-master-merge https://github.com/dominictarr/scuttlebutt https://github.com/mafintosh/merkle-dag