Consensus-layer Call 101

[djrtwo]

Consensus-layer Call 101 Agenda

prev: call 100

Meeting Date/Time: Thursday 2023/1/12 at 14:00 UTC
Meeting Duration: 1.5 hours

livestream

1. Capella
   • pre-interop testnets update
   • CL/EL TX and withdrawals commitment to better support light clients
   • Fork version gossip boundary issues Use Capella fork version for BLSToExecution consensus-specs#3176
   • engine api
     • Engine API: unify failure mode for mismatched structure versions execution-apis#337
     • Engine API: a bunch of cleanups execution-apis#338
     • Engine API: define payload bodies requests execution-apis#352
2. 4844
   • Quick client status updates
   • d-star name Consensus-layer Call 101 #702 (comment)
3. Research, spec, etc
   • Add block/state multiproof request/subscription feature (WIP) beacon-APIs#267
   • Add /eth/v0/beacon/light_client/instant_update beacon-APIs#270
4. Open Discussion/Closing Remark
   • Shanghai/Capella Community Call #1 #708

[djrtwo]

EDIT: we've discussed a lot of this in other contexts. stale comment

---

My preference on how to handle unstubbing Capella functonality for 4844

• CL spec release -- v1.3.0-alpha.3 -- by Jan 5 to include historic accumulators for Capella (essentially the mainnet release candidate)
• Any 4844 testnets pre-interop target v1.3.0-alpha.3 (final capella data structures but capella functionality stubbed)
• Capella testnets pre-interop and during interop target v1.3.0-alpha.3 (full capella functionality)
• 4844 testnets at interop keep stubbed Capella functionality and unstub in a release immediately after interop (first week of feb)

This requires clients to update 4844 data structures to latest capella release but otherwise keeps them unentangled during this intense January sprint

[hwwhww]

I'd like to bring the D-star name discussion to the table again. It would be nice to use the new name in the eip-4844 interop target.

Given that Dubhe is difficult to pronounce, it seems Deneb (proposed by @LukaszRozmej) is the most popular candidate now. I hope we can make it legit in this call if people tend to agree.

[mkalinin]

I would like us to merge this change before Engine API Shanghai spec release:

• Engine API: unify failure mode for mismatched structure versions execution-apis#337

And also consider the following cleanups for inclusion:

• Engine API: a bunch of cleanups execution-apis#338

[yperbasis]

I'd like to discuss the issue of withdrawals root being different between CL & EL (raised at the ACD last week).

[terencechain]

My preference on how to handle unstubbing Capella functonality for 4844

* CL spec release -- `v1.3.0-alpha.3` -- by Jan 5 to include historic accumulators for Capella (essentially the mainnet release candidate)

* Any 4844 testnets **pre-interop** target `v1.3.0-alpha.3` (final capella data structures but capella functionality stubbed)

* Capella testnets **pre-interop** and during **interop** target `v1.3.0-alpha.3`  (full capella functionality)

* 4844 testnets at **interop** keep stubbed Capella functionality and  unstub in a release immediately after interop (first week of feb)

From 4844 implementers call 10, client teams decided to update 4844 on top of capella with everything unstubbed. And I believe this was reflected in the v1.3.0-rc.0 release. Prysm passed spec tests

[zsfelfoldi]

I'd like to join this call and bring up my beacon API proposals:

ethereum/beacon-APIs#267
ethereum/beacon-APIs#270

These would greatly help the Geth light client efforts on the short term and I believe they are also sensible general purpose additions to the beacon APIs. Some work has already been done on these besides writing the specs, @wemeetagain is already implementing the first proposal in Lodestar. I also have a testing tool on a Geth branch that validates the implementation of these proposed endpoints.
I suppose people mostly care about Capella these days but API standardization can be a long process so I'd like to start this as a parallel thread :)

[mkalinin]

I'd like to double check EL client implementation status on the following:

• Engine API: define payload bodies requests execution-apis#352

Considering the lack of time before the interop spec freeze, the proposal is to move this feature to experimental with intention to further include it into Shanghai if it is implemented by all EL clients at the end of the interop.

[lucassaldanha]

I'd like to bring the D-star name discussion to the table

I think Denebola is a good candidate for the name!

Some references:

• https://www.merriam-webster.com/dictionary/Denebola (includes how to pronounce it)
• https://skyandtelescope.org/astronomy-news/meet-denebola-and-catch-a-lion-by-the-tail/

[rolfyone]

I'd like to bring the D-star name discussion to the table

I think Denebola is a good candidate for the name!

Some references:

• https://www.merriam-webster.com/dictionary/Denebola (includes how to pronounce it)
• https://skyandtelescope.org/astronomy-news/meet-denebola-and-catch-a-lion-by-the-tail/

I was hoping for something shorter to type... Deneb, Dabih, Dalim for example...

[zsfelfoldi]

How can I join this call? Maybe I'm not in the relevant channel or just don't know where to look but I couldn't find an invite link anywhere :)

[zsfelfoldi]

Found it :)

[benjaminion]

Notes from the call: https://hackmd.io/@benjaminion/Hkm5x5acj