Content
@
0 reply
0 recast
0 reaction
Barnabé Monnot
@barnabe
Capacity increases may have mixed effects on prices, if network effects/induced demand overwhelm the new capacity. What they do not have a mixed effect on: The total realised demand, which increases strictly. This means more users on-chain, more interactions, more useful possibilities to transact. Unequivocally good!
1 reply
1 recast
10 reactions
eric siu đ
@randomishwalk
have you ever estimated price elasticity of demand for blockspace?
2 replies
0 recast
1 reaction
Barnabé Monnot
@barnabe
I haven't no, one grantee was planning to work on this, but turned out to be not that simple, I don't remember the reason exactly
1 reply
0 recast
1 reaction
Thomas
@aviationdoctor.eth
I guess weâll find out empirically very soon, and this will inform future decisions
1 reply
0 recast
2 reactions
Barnabé Monnot
@barnabe
Which decision would it inform according to you? One that I am seeing is the decision to allocate future gas limit increases to L1 gas vs blob gas. But otherwise I feel we should just have as much blobspace as possible, even if the elasticity is low
1 reply
0 recast
1 reaction
Thomas
@aviationdoctor.eth
Future expansions of blobspace for sure (full danksharding?), & more broadly any design decision that relies at least partly on some induced demand assumption (Braess? Jevons? Canât remember which one). In my line of work these empirical measures are super helpful because they are hard to derive from first principles
1 reply
0 recast
2 reactions
Barnabé Monnot
@barnabe
More Jevons than Braess in this case yes, but I believe the difference is we have a strict preference for more capacity, the negative externality/social cost of providing more blobspace is bigger nodes, so we have a constraint on the provision, but within that constraint, we should max out capacity imo
1 reply
0 recast
1 reaction
Thomas
@aviationdoctor.eth
And for that constraint, we are lucky to have both Mooreâs and Nielsenâs laws đ
1 reply
0 recast
2 reactions
Barnabé Monnot
@barnabe
And PeerDAS and Danksharding designs!
1 reply
0 recast
1 reaction
eric siu đ
@randomishwalk
good framing â yes the âbudget constraintâ (to force a basic model on this) here is the cost of syncing a full node + ongoing costs to keep them running at reasonable performance levels
0 reply
0 recast
0 reaction
