Skip to content

Latest commit

 

History

History
64 lines (48 loc) · 2.69 KB

18-7November2018.md

File metadata and controls

64 lines (48 loc) · 2.69 KB

dat protocol WG (#18)

DATE via IRC (#datprotocol)

Agenda discussion

People

  • mafintosh
  • pfrazee
  • jhand
  • yoshuawyts
  • bnewbold

Agenda

  • Review last meeting/action items

Summary/Review of Previous Meeting

Action Item Review:

  • respond to DEPS#45 to explore the .well-known/dat solution (@pfrazee)
    • pfrazee: "that's done, I basically replied that we'll give this another look. I haven't done any work on it but I expect we'll revisit the spec when the timing is right"
  • respond to the email thread behind datproject/security-discussion#1 (@mafintosh)
    • pfrazee: "mafintosh: you wrote up the majority of a response, didn't you? Is there some way you can just point to that response?"
    • mafintosh: "yep"
  • write up some notes on spec changes for hyperdb (simplifications & single-writer) (@mafintosh)
    • Decided to drop this since the changes are taking a while, will get a status update in the meeting.
  • get hyperdb update ready for vote (@bnewbold)
    • voted, merging
  • prioritize having beaker leave a swarm when you leave the page (@pfrazee)
    • pfrazee: "not done yet"
  • security/privacy docs update (@joehand)
    • jhand: "not done yet"
  • OTF red team audit application (@joehand)
    • Removed item, will handle in the future
  • try using dat over tor (@pfrazee)
    • Removed item
    • pfrazee: "Metadata privacy is a priority but will need dedicated resources that I dont have atm"

Maf's update on the protocol work

<mafintosh> We are integrating hypertrie into hyperdrive directly
<mafintosh> This closes all existing bugs / scaling issues with the data structures
<mafintosh> But keep the api the same which is great
<mafintosh> Then we integrate the hypertrie into hyperdb with a union data structure
<mafintosh> Gives us multiwriter
<mafintosh> Then switch out hypertrie with hyperdb in hyperdrive
<pfrazee> will the hypertrie update in hyperdrive be a major version bump on the hyperdrive structure?
<mafintosh> yes, a needed one
<mafintosh> but the subsequent multiwriter one will be a minor bump on the data structure

ACTION ITEMS

  • Gist/blogpost about scaling hypercore and what did/didnt help (@mafintosh)
  • Review Sleep Headers DEP dat-ecosystem-archive/DEPs#56 (all: @pfrazee, @bnewbold, @mafintosh, @jhand, etc)
  • Finish wire protocol DEP (@bnewbold)