Skip to content

Latest commit

 

History

History
59 lines (44 loc) · 3.65 KB

16-26September2018.md

File metadata and controls

59 lines (44 loc) · 3.65 KB

dat protocol WG (Meeting #16)

Sept 26, 2018 via IRC (#datprotocol)

Agenda discussion

People

  • pfrazee
  • bnewbold
  • dcrobins
  • mafintosh

Agenda

  • Review last meeting/action items
  • AGENDA ITEM: Discuss new Hyperswarm project https://pfrazee.hashbase.io/blog/hyperswarm
  • AGENDA ITEM: Should we deploy hyperdrive-v2 with a simplified trie structure, or with hyperdb single-writer?
  • AGENDA ITEM: "Follow redirects in DNS?" dat-ecosystem-archive/DEPs#40 TL;DR: should we allow HTTP redirects of .well-known/dat? In particular in handling of www.example.tld vs. example.tld. Alternative is to require direct response on both like Let's Encrypt.
  • AGENDA ITEM: "Added functions for sign/verify to DEP 0002" dat-ecosystem-archive/DEPs#41. TL;DR: include more Javascript+libsodium code in a DEP as a clarification. My dispensation is to merge this PR, and also expand the english language with reference to specific (crypto) RFCs and what is getting hashed signed.

Summary/Review of Previous Meeting

Action Item Review:

  • ACTION ITEM: Submit a DEP for hyperdrive/v2 (the trie-structure update) (@mafintosh, @pfrazee)
    • pfrazee: I ended up punting this for further discussion, which is now an agenda item
  • ACTION ITEM: get hyperdb update ready for vote (@bnewbold)
    • bnewbold: I started digging back in to updating hyperdb DEP this morning, but not ready yet
  • ACTION ITEM: respond to datproject/security-discussion#1 (@mafintosh)

Meeting Notes

  • AGENDA ITEM: Discuss new Hyperswarm project https://pfrazee.hashbase.io/blog/hyperswarm
    • There was great rejoicing
  • AGENDA ITEM: Should we deploy hyperdrive-v2 with a simplified trie structure, or with hyperdb single-writer?
    • mafintosh: I "productionized" the hyperdb trie as a single writer one in the hypertrie module, in an effort to get that part out
    • hypertrie module is currently not compatible with hyperdb spec due to a couple of small differences
    • bnewbold and pfrazee raised concerns about added effort of another implementation (specwork, multiple migrations)
    • decided: we'll integrate hypertrie's work with hyperdb and aim to deploy hyperdb singlewriter
  • AGENDA ITEM: "Follow redirects in DNS?" dat-ecosystem-archive/DEPs#40 TL;DR: should we allow HTTP redirects of .well-known/dat? In particular in handling of www.example.tld vs. example.tld. Alternative is to require direct response on both like Let's Encrypt.
    • overall response was, we should do the simplest thing
    • concern: ties us to behaviors which may only be possible with wellknown DNS lookup, which may not be a longterm solution
    • concern: adds complexity and hard to predict risks
    • decided: "soft no" - we'll leave open to more commentary but keep simple for now
  • AGENDA ITEM: "Added functions for sign/verify to DEP 0002" dat-ecosystem-archive/DEPs#41. TL;DR: include more Javascript+libsodium code in a DEP as a clarification. My dispensation is to merge this PR, and also expand the english language with reference to specific (crypto) RFCs and what is getting hashed signed.
    • decided: accept the PR

ACTION ITEMS

  • respond to the email thread behind datproject/security-discussion#1 (@mafintosh)
  • write up some notes on spec changes for hyperdb (simplifications & single-writer) (@mafintosh)
  • respond to dat-ecosystem-archive/DEPs#40 with "soft no" (@pfrazee)
  • merge dat-ecosystem-archive/DEPs#41 (@bnewbold)