I wrote in reply to k00b about my issues with sidechains specifically as it applies to thinking of it as a scaling solution: #226879

Not mentioned in that comment are things like "Hey, all the limitations of our decentralized timestamp server are going to apply to these side decentralized timestamp servers too"

For example "Hey, the reason we have a small blocksize is going to apply to sidechains too"

"Hey, the reason we have a high block time is going to apply to these sidechains too"

And remember, this is constrained to only talk about sidechains for scaling and nothing else. Fuck all the nonmonetary use cases. Take them out of the conversation, they're a distraction and off topic to the resolution. What's the resolution?

Resolved: "How do we scale Bitcoin?"

That is the topic. I spend so much time dedicated to talking about how I want to constrain the discussion to just that and only that, because I know people are stubborn and I kinda feel scaling isn't the real reason behind this suggestion, so lets constrain the topic to only be about scaling to really put to the fire how much about scaling this push for sidechains really is.

The question is why not?

A thing I haven't seen much discussed but that seems important: fragmentation of the utxo set when there are 100x more users. Utreexo is one way of approaching this, but I'd be interested to be pointed to discussions on the topic, especially ones that do computational modeling / simulations.

Imagining high on-chain fees + utxo fragmentation into quasi-dust is scary.

"Bitcoin will die without this!" is how you know it's absolutely useless.

Although this is true, it's very easy to become a custodian, you just need a BTCPayServer, I'm already doing it. This solution is not only for uncle Jim, many small custodians may sprout to serve local markets. This is an important issue, as we're never dealing with trustless services, but trust minimized. It could be possible that you personally know the person in charge of your custodial wallet, and this person wouldn't be incentivized to rug you, because for one it's his business to serve as custodian, and two it's just not worthy for a few million sats. I understand that custodial solutions are not ideal, but are they actually that bad? Sometimes I wonder why people coming to bitcoin, go full paranoia, cypher punk, private maximalism... like bruh, a month ago you were using only fiat on a KYC bank, just take it easy and let's go step by step.

On the other hand, one good reason to go forward with drivechains could be to bring together many crypto enthusiasts that are now working in other projects where they can develop things not available now on bitcoin, but as Jimmy Song said, most of the crypto world is there mostly for gambling and degeneracy, not for the tech, drivechains won't bring them back.