Couple of thoughts about bitcoin scalability future

[piotrgal]

I was thinking recently about that issue with bitcoin doing only few transactions (or so) per second and long confirmation times. Last times the community is “desperate” about the need to increase scalability / speed in order to compete with traditional payment systems in future.

I just realized that this is simply wrong direction of thinking (in my opinion). Essentially we get all that goodness related to distributed / decentralized model, but that comes with the price of speed. It’s simply impossible to achieve high efficiency in such heavily democratized structure. On the other hand, really high speeds are possible clearly due to centralization of systems.

I would say that instead of trying to “catch the rabbit”, maybe the solution is to accept that unpopular concept of centralization and do something similar what Green Address does (let’s call it “pragmatic trust”), and simply leave the current decentralized approach as the “super-feature” to carve in stone all the transactions without rush, or serve as the arbiter in case of problems ?

[knightkon]

I am reading what you are putting down, but what about these issues in the future.  If the confirmation times are so long right now, would I be correct in saying that over time the more difficult the mining level, the longer we will be waiting for confirmations?  Also the more Bitcoin is used, will the confirmations come faster since there will be more transactions, making the blocks fill faster?  I personal think that the speed will be the killer of this all and will prevent it from moving forward later in the future, the far far future.

[piotrgal]

I am reading what you are putting down, but what about these issues in the future.  If the confirmation times are so long right now, would I be correct in saying that over time the more difficult the mining level, the longer we will be waiting for confirmations?  Also the more Bitcoin is used, will the confirmations come faster since there will be more transactions, making the blocks fill faster?  I personal think that the speed will be the killer of this all and will prevent it from moving forward later in the future, the far far future.

Not sure which times exactly you mean, but AFAIK the "standard" confirmation times are kept on more / less constant level by auto-correcting the mining difficulty factor. So if the load will increase, but the mining power will not, the difficulty will be lowered to keep ~ 10 minutes per block.
The problem with filling up the blocks with transactions - I think that's one of the current debates now - to increase block size, so I guess that (have to) will be solved soon...

[piotrgal]

I think I need to elaborate a bit more about my proposal. What Green Address does is providing "safe" transactions, so they check and guarantee that any transaction sent from their wallet will not be fake (doubled).

That means If you check that your payment comes from address hosted by them, you can ignore network confirmations and accept that transaction instantly. And here is the place when "traditional" trust comes back to the game.
But that's exactly what I'm trying to say - if we will be a bit more pragmatic and accept, that we can trust "somebody", this will solve the problem of waiting for confirmations (and of course all transactions will still be mined, so we can always fall-back to that in case of doubts).

Then we can have more such wallet services where we trust (anyway you can't do business completely without trust - it's an utopia...) and still whoever wants, can use private wallet soft, where transactions will be processed by network and full confirmations.
So no problem with dictating a "trustful" approach or lack of choice.