What optimizer runs should we have?

[andreivladbrg]

As I started working on package tethering issue, a question came to mind, as we have different deployment configurations between v2-core and v2-periphery.

Core:

v2-core/foundry.toml

Line 18 in 9926384

optimizer_runs = 1000

Periphery:

https://github.com/sablier-labs/v2-periphery/blob/18d08400fa75194ac48c016e7913064eb63c32b1/foundry.toml#L20

We have two options:

1. Increase the optimizer_runs and use the same configuration for both deployments (we can increase it to 4000)
2. Add two foundry profiles (core/periphery) and deploy the contracts with different optimizer_runs

I personally prefer the first option, as I don’t find the added complexity worth it. Also, regarding option 2, if we want to use different optimizer runs for different contracts, we can use a different number of runs for NFTDescriptor (which has the biggest size), LL, LD, and LT. (I don’t find it worth it)

@sablier-labs/solidity wdyt?

[smol-ninja]

I will also vote for option (1). But why 4000? Doesn't it also exceed the limit for NFTDescriptor.

In that case, since higher the better, we can safely use 8,000 given LD has the highest size. And then we can use 1000 for NFTDescriptor deployment.

[andreivladbrg]

We can't increase it to 4000, I was wrong.
We have to keep it to 1000

[PaulRBerg]

That's fine

[PaulRBerg]

Let's go with option 1 and set the highest number of optimizer runs that doesn't result into the NFT Descriptor going above the bytecode size limit.