Segwit transactions are not "the same set of transactions" as non-Segwit transactions.
Let's back up the bus a sec. Seems we have a terminology overload. When I said "the same set of transactions", I'm speaking of economic transactions, not their encoding upon the wire.
If, in regards to Segwit's [flex] block size increase, you're asking whether more data = more data, then yes. More data = more data. To which I would ask, what point are you trying to make?
For the same set of _user_transactions_ -- as in user X transfers value Q to address Z' -- more data must must be stored by a fully validating node using SegWit than without. It ain't a matter of the centrally-planned block size increase, its a matter of adding the ability to correlate the correct data in the 'main' block chain with the correct data in the 'witness' block chain. This is not free - there is overhead added.
in standard transactions, due to the size of signatures vs. preimages, Txouts are about 25% of the size of TxIns (hence the 75% discount for Segwit transactions).
'About', yes. Today. On average. Kind of. Tomorrow? Who knows the ratio? Nobody - that's who. The direction BU seems to be headed to deal with this is via emergent consensus - let the free market decide, rather than being centrally planned 75%, which will be simply wrong for certain transaction mixes. Which, incidentally, is what BU does today for the maxblocksize -- let the market decide.
Rather than rely on the centralization of the ever-so-well-meaning dev oracles (who are bright guys, but I think are headed down a blind path on economic matters), I would rather put my trust in the decentralized market.
Um, what did you think "Peer-to-Peer" referred to in "Peer-to-Peer Electronic Cash System?" Feel free to google "peer-to-peer":
https://www.google.com/search?q=Peer-to-Peer&ie=utf-8&oe=utf-8denoting or relating to computer networks in which each computer can act as a server for the others, allowing shared access to files and peripherals without the need for a central server_.
[emphasis added] I note your quoted definition has exactly no correlation to the type of 'centralization' we are discussing. Note that 'a'? Note 'server' in the singular?
It looks foolish on you to argue that decentralization doesn't matter in regards to bitcoin
It looks foolish on you to attribute an argument that I do not make. There are many ways that a complex system can be decentralized. True,
one is in the number of non-mining fully-validating nodes. I would argue that we could lose at least an order of magnitude here, and it would affect the security of bitcoin not one whit. After all, other than the fact that non-mining, fully-validating nodes tend to be run by bitcoin owners, such nodes are essentially powerless. The node's only option is whether or not to forward a particular transaction. Sure, the user operating the node can abandon the chain created by the mining majority. But that is not the node acting, it is the holder of the coins.
But
again, for the same number of economic transactions, The Omnibus SegWit Changeset actually
increases the bandwidth and storage burden upon fully-validating nodes.
Another dimension the system might be centralized across is in mining power.
Or implementations.
Or decision making.
Or even total number of users of the monetary unit. But we don't care about these, do we? (Well, I do, but it seems you may not).
Cheese? Who's in the Free Shit ArmyTM now?
What are you babbling about? I'm contributing all the bandwidth I can
Want some cheese to go with that whine?
If you're contributing all the bandwidth you can, you're gonna be kicked out of the club of contributing nodes as soon as SegWit activates.
Gee, thanks Core!Incidentally, I operate a fully-validating node as well. Several, in fact. So sorry you live in a backwater. Must be analogous to trying to mine in Sweden, rather than China.
Big blockers suggest that we increase capacity simply to keep transactions cheap or free for users.
Perhaps some. Not the ones I speak with. We suggest we increase capacity in order to increase capacity. Making room for more transactions. Which would make room for (all else equal) more bitcoin users. More users, more momentum, more value, more hashpower, more security, more headstart before the vampire squid awakens to the significance of this radical new money.
We trust that the miners are capable, as a group, to provide the proper value for maxblocksize. As an emergent property of the market. Your solution is a centrally-planned fixed unit of size for this crucial economic variable. Yet you claim to desire decentralization? *psh!*
transaction relaying and validation is not free
Yet nodes do not get paid for this service - only miners do. Ergo, miners are the ones with the incentive to arrive at the proper value for this crucial economic variable.
You seem to think I should pay for users to transact for free or extremely cheap, forever, with no regard for the actual cost of validating/relaying transactions
You make no sense. You are already working for free.
Bitcoin isn't a public service for users to transact for free -- it is an incentive-based economy.
The only possible
incentives for operating a non-mining, fully-validating node are the ones you mentioned earlier - security and altruism. Well, maybe the ability to handle your transactions without needing to rely on others. Nevertheless, you don't get a share of the monetary incentives under any scenario. (kinda sad nodes are not renumerated, maybe someday the next satoshi will figure that problem out).