ePBS breakout room #2

[potuz]

• June 7 (Fri), 2024 , 14:00 UTC :clock2: ❗
• Duration: 60 minutes
• Zoom link

Agenda

• EIP Status
• Spec discussion
• Implementation complexity discussion

[terencechain]

If time permits, I'd love to chat about the current status of (block, slot) fork choice analysis for ePBS. As analysis done in these documents:

• https://ethresear.ch/t/sandwitch-attacks-on-epbs/19538
• https://hackmd.io/@potuz/rJ9GCnT1C#Forkchoice-considerations

Besides the current post-anti and ex-anti analysis of two slots sandwich attacks, what other analysis/attack is worth studying over. One example is multi-slot liveness attack. As the performance of network conditions becomes worse, chain growth stops completely if block latency is > 3 seconds. Whether a backoff scheme is necessary to facilitate chain progression during periods of extended latency

[hwwhww]

@potuz you should not use 🕙 emoji here 😄

[potuz]

Lol copied from a different issue :)

[potuz]

If there's time (writing here mostly for the next breakout) I'd like to brainstorm how an "off-protocol" solution actually would play out. It seems to me that ePBS makes any out of protocol solution very very hard to be viable.

[terencechain]

If there's time (writing here mostly for the next breakout) I'd like to brainstorm how an "off-protocol" solution actually would play out. It seems to me that ePBS makes any out of protocol solution very very hard to be viable.

I wrote a post today on out-of-protocol timelines under ePBS: https://hackmd.io/@ttsao/comp_epbs_timelines. It's the worst of both worlds.

[terencechain]

Call notes https://hackmd.io/@ttsao/epbs-breakout2

[shyam-patel-kira]

Was the call recorded? Can you share the recording @timbeiko

[poojaranjan]

Recording: https://youtu.be/w7Wa6oprEhQ

Please refer ePBS Playlist for other meetings.

[shyam-patel-kira]

Thanks @poojaranjan!