we're on the same team.  we all want more decentralization.  we're simply arguing about the best way to achieve that.  multiple ppl here have argued how increasing the block size will facilitate that.  i happen to think those arguments are most sound.

i'm not sure why you say Gavin is a USG lap dog.  over Bitcoin's entire history since his involvement, he has been shown to provide sound leadership.  my poll and the other github poll show this.  the general tenor of comments show this.  yes, there are haters but on balance he is the right one to have been chosen to be lead.  

* edited to read more clearly.
I agree with the notion that miners are for the most part unaffected by block size and are empowered not to care, this is also why I dismiss those developer arguments that want to solve the block size problem by manipulating mining fees, or some variant of this idea. The incentive is not in the TX fee to reduce block size - that's paying miners not nodes if anything you want the incentive to be supply and demand based on node size.

Ironically it is only the competition for the fee between miners that will force writing blocks to the marginal cost and force the block size to the smallest size capable of sustaining a profit, this could be neatly modeled by the Nash equilibrium.

As block rewards diminishes the Nash equilibrium is introduced and miners become marginalized with little to no power in the system.

I participated in discussing the idea in 2012  of financially incentivizing notes in a market driven way to regulate miners and block size. But after pondering the idea over time it seemed it was not  necessary.

People invested in the idea that money is memory who store that memory on the blockchain - They have a lot of "memory",ie. an invested interest in the blockchain will want to preserve the blockchain, some call it altruistic but I prefer to think they will use greed for the grater good.

The conclusion I draw is as long as wealth is distributed and people are in competition with one another the blockchain will remain distributed. by the nature of the design of Bitcoin, the wealth that is still to move in to Bitcoin cannot be transferred into Bitcoin with out redistributing it to the participants who grow the network.  

 

There's a peculiar kind of incoherence about people who can argue both for decentralization and also argue that users of the system can not be relied up to decide their own best interests.

It's not possible to build a currency on misanthropy.

http://nakamotoinstitute.org/reciprocal-altruism-in-the-theory-of-money/


If you want your money to be valuable you need the people who produce the products and services you want to consume to use that money.

There is no other way to imbue currency with value.

If bringing sound money into existence requires an mass education project to overcome many generations of propaganda-induced fallacies, then that's what it going to take.

There is no shortcut.

the Blockstream devs have said they would like to see & study what happens upon the repeated filling of blocks.  they'd like to study what happens to frustrated users, hoodwinked merchants, exchange price volatility, confused full nodes, etc.  to what end?  to satisfy their own curiosity?  and only to then raise the limit as they've admitted to?  what a bunch of misplaced pseudo-academia.

they're like a wife who begs her husband to beat her just so she can experience what it is like.  so he beats her repeatedly.  she finally decides she doesn't like it.   but it's too late; he's already lost all respect for her and leaves for another woman.  that's what will happen to Bitcoin if users and merchants have bad experiences at this early stage of the game.  they'll just leave and may not come back for a 100 yrs.

i think thats more or less how its currently works when there's a backlog of unconfirmed TX

this is fine for now, but at one point if theres alot of TX the mem pool will just grow and grow, and TX will confirm slower and slower.

You're saying miners currently sometimes only put the hash of all the tx in a block, instead of the tx themselves?

the mempool is rarely uniform across all nodes.  it would be impossible to reconstruct which unconf tx's a node would be missing.

your idea is a variation on IBLT.  but in that case, nodes can reconstruct their missing tx's due to the math of the IBLT.

and your idea would totally render SPV clients unusable as they rely on retrieving the Merkle tree path with it's block header to their specific tx history they are interested in.

mayeb the last one

yes, gold is useless in the new digital age.