In response, Mark Friedenbach explained that addresses are simply a UI convention that does not factor into the bitcoin protocol at all. However, changing address formats would require consensus and a coordinated upgrade of the entire network. Friedenbach suggested that bech32 addresses offer better security guarantees for pay-to-script or lower fees for pay-to-pubkey-hash than 1.. or 3… addresses. Regarding SegWit adoption by exchanges, De Luigi expressed his concern that the total blockweight is increased when a tx has at least ten outputs using SegWit, and how it may discourage adoption among exchanges without cheaper fees. He also argued that SegWit decreases fees only for specific kinds of transactions and increases the total weight of the blockchain. However, Friedenbach noted that SegWit can approximately double the capacity of the blockchain if the entire block is filled with this kind of tx. De Luigi then asked whether bech32 would change anything. Bech32 allows saving about 22% of space compared to legacy tx, but it is non-compatible with the entire ecosystem. De Luigi thought that a hard fork bech32 upgrade + 2x could help in scalability while testing LN, and it might be the only way to effectively promote or enforce SegWit adoption.