70 sats \ 7 replies \ @Zepasta 6 Feb 2023 \ on: Blockchain Bloat Doesn't Matter bitcoin
deleted by author
yeah, kinda sucks, a 1tb drive will probably be full in less than a year from now.
I think there will definitely be strong arguments in favour of putting some limits on Witness and Taproot transaction sizes, it won't immediately solve the problem but as more nodes upgrade to versions that soft-fork to sane size limits on transactions and especially witness data will slowly strangle the excessive growth of data size. Or there is another option where such transactions get the garbage data left out of the on disk storage, since it isn't important to the security of the transaction anyway.
There's multiple approaches, I suspect we may see changes in protocol implementations that push back the upgrade necessity.
reply
Na, even if every block was 4MB from now to 1 year later would only be around 220,000MB or 220GB. That wouldnt put use at or over 1TB. This is 10 min average block time.
reply
I dont want that though lol. Id rather have 1.5MB blocks but if they are paying for the space, then so be it.
reply
I think the point is that they are paying less for space than transactions would have
reply
Yep, this is the open wound in the protocol from the blocksize wars. Weight of witnesses is under-counted compared to normal data. The first two shots broke btcd's witness size limit parameter, which was not in protocol but there because of the likely resource exhaustion attack.
I told y'all that's what they were doing. But everyone wanted to hate on btcd instead of think about why btcd had that limit. Most people probably even think it was a "bug" instead overzealous prudence.
reply
Yeah, I suppose if the end result is NGU then the price of storage has to go up faster to matter.
reply
The rewards "curve" also has an impact. Anti-tail-emissions people don't consider the fact that the "increase of supply" is a sawtooth wave with halvenings.
reply