Replies (12)

jb55's avatar
jb55 _@jb55.com 2 months ago
even with full archival nodes, if you don’t have txindex on you won’t be able to easily extract the op_return from a txid.
You’re right. Just all mine need to be full nodes. lol It might raise the hardware barrier, but you still have to use IBD and prune after the download finish’s anyway, it’s the same requirement up front. Less IBD nodes increases centralization and relies on more trust. Saying “we’re making this change, you can just prune” is the worst for decentralization of the nodes. It’s not theoretical. We’ve already seen it! Wallets default to public Electrum servers. People syncing through Blockstream.info. Light clients like wallet-as-a-service models. Some of these users don’t even know that they’re using a trusted service. These are fine options but become points of failure when only option for a node is “just prune it dude”
One of the things keeping node runners interested and leveling up their skills is running services adjacent to the node, like block explorers and lightning. Without indexing, pretty much anything interesting you want to do with a node is off the table.
Man.. you've lost the plot, honestly. Alright, Will, I'll definitely respect your argumentation. Keep going at it for sure. Its the way things should be. My only wish/advice is, you also take an honest look at the reaction from node runners and take that into account as you support a change on core. If you ask me, looking at the reaction from an HONEST point of view, it tells me a substantial amount noderunners disagree with whatever logic you and core devs are presenting. And despite whatever you might think, "a substantial amount of noderunners" disagreeing on a changr should raise serious concerns about that change.
several related applications (etectrum, lnd, etc) require non-pruned node. If your node is pruned you need to trust on a third-party non-pruned node or service ! do you think this is good for decentralization ?