Notes for Editorial Meeting 27 July

[bumblefudge]

CASA meeting 27 July

Proposed Agenda (Juan)

• Quickies/Last Calls

  • finalize namespaces migration
  • typo fixes 1 and 2 - totally noncontroversial

• To discuss

  • SIWX and CACAO update to match
    • any input or updates needed on namespaces/sol/caip122 PR or can this be merged the second the other two are?
    • All 3 ready to merge upon 2nd approval
  • Namespace for Stacks - OK to merge without a regex or complete examples? Or ping the PR opener and demand those nice-to-haves?
    • ready to merge upon 2nd approval! @ukstv to review
  • Block number redux/new Caip19 PR
    • Query marking: # stripped in HTTP architecture --> @ might work
  • Broader block number question (older PR):
    • Tim: Mental model question: Do CAIP-19's identify an asset (which exists in time) or an asset abstraction (time relegated to query)?
    • Ligi: merge --> Safe (new state abstraction, 3rd option between latest and historical)
    • Use cases? Tim: ERC721s are too mutable for some use-cases-- if resources they link to are mutable, can be dangerous to assume (or sign over/notarize) anything about them... ;
    • permalinks and CIDs (by analogy to github, for ex.)
    • Sidenote: State assumptions: assuming "latest" (or "safe") is very eth focused, other chains/VMs might have diff state assumptions; explicit always more chainAgnostic
    • Ligi: If we're changing the CAIP-19s mental model and introducing the concept of time/state, I'd rather a new CAIP that extends the class
      • Juan: Do other CAIPs also need a query language, and/or the concept of time/state?
        • Sergey: Start a new CAIP and work it out on github? time/state (and ownership) are both out of scope; I'd rather one CAIP for time/state and a second CAIP for ownership/asset-internal-state
        • Tim: Cross-chain research would uncover requirements for a generalized concept of resolution and/or querying; Ligi: GATHERING TOPIC BAYBEE - Solana, Filecoin, and Polkadot people will be represented...
          • Tim: Already did some research and found all except Hedera and Iota have some form of block ticker/++ system...
          • Ligi: Reorgs in EVM? Sergey: Reorg'd blocks change but keep their number, don't they?
            • Tim: Reorg could just break blocknumber references; i'd keep it out of scope; the only safer version would include a hash
          • Tim: Time != block#; the latter is more explicit, even if not sequential or consensual...
      • Tim: PR opener is also working on the web3:// protocol handler at IANA to handle heirarchical URIs for contract + function name + parameters... I'm convinced of the utility of this explicit addressing scheme
  • Org
    • Berlin CASA gathering - thurs 15 (eth berlin = fri-sun)
      • room confirmed @ C-Raum - 100ish capacity, no food, cash bar
        • timing - 11-17? long lunch to accomodate restaurant/catering problem?
        • lunch 1-2.30? maybe book big room in Ming Dynasty?
      • Keep an eye on tickets - 100 max, 7 already claimed by Fission, 2 from Spruce, 2 from Centre
      • topics: Query abstraction? Temporal common-ground? Recruitment to get more namespaces and namespace-caips defined?
        • fission and oliver from Spruce have some topics?

• Stretch-Goal Agenda

  • we could revive this Cosmos/caip19 thread with some pinging and ideally email/discord/etc pinging as well?]
  • whatever happened to the CAIP-25 handshake stuff?
  • whatever happened to the CAIP50 multiformat stuff?
  • are their polkadot wallets and dapps using CAIP-10, or trying to use the cross-network form of addresses? my draft PR is still awaiting commentary