Suddenly the promise of "trust-less" makes no sense, and we are just left with poor performance and immutability of human errors.
(This isn’t to say that there are no bad actors, but that a wolf in sheep’s clothing would be forced to act like the sheep it presented itself as.)
Using your example, you could write a contract such that: you buy an item for X dollar amount, the contract holds it in escrow, waiting for FedEx to confirm receipt of the package and a deposit of Y% of X into a rolling insurance fund from both FedEX and the seller. At this point, if FedEx fails to deliver you’re fully refunded out of the insurance pool and the seller gets their money back. If they deliver, FedEx gets 90% of the insurance payment back, the shipper gets your payment, and they split the seller’s premium as a shipping fee. If the seller never drops off the package you get your money back.
You could even throw in a return period where a third party can confirm whether a bad package is eligible for refund.
There are obviously holes in this haphazard scenario but the big picture is that there’s a lot of stress that can be offloaded to a truly unbiased 3rd party, even if you can’t achieve full trustlessness in real life.
Who is in charge of confirming to the smart contract that FedEx delivered? Whoever you pick, they have a strong financial incentive to lie. If the recipient confirms, they can lie in order to keep the item and get their money refunded. If Fedex confirms, they can lie in order to keep the item and they only lose the shipping costs.
Trust is a lower energy state - it allows specialization, and conservation of energy that might need to be spent defending against threats from different directions.
But people who hold anti-social ideas seem to indulge in thinking that they can create a novel system where they have no obligations, just privileges. Crypto prices going up causes these anti-social delusions to flourish because the increasing prices seem to validate the libertarian futurists fantasy of a world with no trade-offs, where we can avoid having responsibility to other humans.
I genuinely have no idea what you mean by this. Unless you manage a 51% attack there’s no discretion to be had. You can for sure push nonsense to the chain, but it would either be ignored or punished [1][2] in some cases.
Blockchains are just distributed computing, it’s all computer code. The first two sections here (https://paulx.dev/blog/2021/01/14/programming-on-solana-an-i...) explain it much better than I can.
[1]https://support.polkadot.network/support/solutions/articles/... [2]https://docs.solana.com/proposals/slashing