Topic: How a floating blocksize limit inevitably leads towards centralization - page 6. (Read 71590 times)

THIS RIGHT HERE

Every person that converts his USD to BTC to play SatoshiDice (at minimal fees) accelerates its growth and therefore its exchange rate against fiat.

Transactions shouldn't be free, they should cost what the market dictates they should could. A market that doesn't limit them based on an arbitrary technical restriction that fails to scale forever with computing resources.

The exchange rate doesn't earn the network any money, what you are describing is an externality.  And it's a bad policy to get new users, normally accustomed to a three day wait for cheques or a 30 day wait for CC's, accustomed to an hour wait for free transactions.  Yes, we should have free transactions, for many reasons besides your's.  It is not necessary for those free transactions to finalize at a nominally equal rate to fee paying transactions.  If you are donating to a charity, you can wait for three hours or three days; after all, it's a gift.  If you are buying something from Wal-Mart, WM can suck it up and sponsor their own mining company if the wait times are problematic for them.  But if you want your transaction to be processed quickly, you should get used to paying for that.  Users should get used to paying for speed now, before we are all forced into it.

Yea it would take time.

IIRC, that 7 tpm metric is a calculated max rate based upon the current average transaction size and the 250KB soft limit.  The hard limit is currently one MB, so we can presume that 28 tps or so is the realistic limit.  This doesn't sound like much, and it's not really compared to payment processors such as Paypal.  However, as a payment processing backend network, comparable to ACH, it's not really bad and it's a couple orders of magnitude cheaper.

The real question is, does the main bitcoin network really need to support such transaction rates?  I would argue that it does not, and that I'd give even odds that a one MB hard limit wouldn't result in catastrophy.  I would also argue that a 10 MB hard limit, which would put us up close to paypal, isn't a bad plan either.  Keep in mind though, as we raise the limits, the theoretical resource usage by the network goes up exponetially.  We need out-of-band networks in order to fulfil Bitcoin's original promises.  They have always been expected to develop around the main network as the economy grew, we just can't really predict how it's all going to play out.  Free markets are like that, efficient because they are unpredictable.

And if you can run a full node on your wristwatch I'll eat my hat.

Actually, free transactions do earn money. They encourage new users. More people will try bitcoin, accelerating its growth. By implication this contributes a proportion to its exchange rate against fiat. So, miners that include free transactions are indirectly benefiting from them as their block reward is higher in fiat terms.

We know that the equilibrium when not restricted by block size is minimal fees per transaction (even free transactions) because that is the reality now.

An imposed sub optimal block size does not create a free market, it restricts one.

Bitcoin definitely doesn't need free transactions to be successful. But it needs transaction costs on the order of around a cent  (to prevent spam), not a $15 wire fee. What is the actual cost to store a single transaction forever? Some very small (and decreasing) number.

The market doesn't work like that.  The free market involves parties reaching an equilibrium.  "too expensive, therefore collapse" is not a realistic scenario depicting an open market adjusting to costs.

Something just occured to me.

Part of the problem here is that, while we want to keep free transactions available within the system, users are accustomed to sending free transactions and getting them processed as fast or nearly as fast as fee paying transactions.  On some level, this is a psycological problem, and one that we can address fairly easily.

What needs to be done, is alter the transaction priority scores so that, not just spammy transactions, but any free transaction will not realisticly be included into any block for a time period.  If we were to alter said priority scores so that all simple transactions that were free (or with a fee lower than the minimum) had to wait two or three hours before they could be included into any block, then the bitcoin user base would be better conditioned to expect and adapt to such delays if the hard limit were ever to force the issue.  We might actually see some changes in behavior, and perhaps better fee paying compliance as opposed to simply complaints from the userbase accustomed to free transactions.

After all, processing these transactions may be very low cost, but they are not free.  If there is no downside to sending mass numbers of free transactions, there is no rational reason for most people to ever start paying them.  However, if we were to tell everyone that this is going to start happening as a matter of course in, say, three months; Satoshi Dice and others will have the time to adapt their models and behaviors, and we might actually start seeing more people paying the minimum fee out of time preferences before we get to the point that we are bumping up against the hard limit.

Free transactions, as they are currently performed, are charity.  The last thing we really want to do is get an entire economy accustomed to such a charity, no matter how nominal that cost might be.

And thats why, even if Bitcoin isn't the theoretically the best system for micropayments, it WILL be the system for micropayments. Betamax was technically superior to VHS, didn't matter, it lost. Alt currencies might be fine if you are a crypto nerd, but the public won't accept them. We are having enough of a challenge convincing ordinary people that Bitcoins are money. But when we do, they want them and they use them and THEY BECOME MORE VALUABLE.

Bitcoin's value comes from its function and trust. If it becomes expensive to use Bitcoins, the game is over.

I'm definitely not hostile towards out-of-band transaction systems. They're GREAT, if you trust the centralized resources that provide them. I'm hostile towards a hard permanent limit of 7 transactions per second / 50GB per year transaction growth in Bitcoin regardless of the growth of storage and bandwidth capabilities because some people believe that fewer total transactions of a less used currency (with then decreasing utility and value) will somehow make themselves more money, perhaps, in the very short term until the system dies.

Buy hey, at least you can run a full node on your wristwatch!

Anon136, the problem is that while competing alt-currencies are perfectly acceptable within the cryptographic community, it would be a major pain to the public at large.

The public just want The Bitcoin. They have no stomach for multiple variations, it would be like the VHS/Betamax or Blu-Ray/HD-DVD debacles all over again, worse, bitcoin 1,2,3,4 would be the same kind of problem on steroids!

If it can all be done in the background with multiple blockchains seamlessly integrated, invisible to the public so that there is only bitcoin - then great. But how soon can that kind of technical solution emerge? Not as quickly as saturated 1Mb blocks will happen.

I hardly think so you're not talking to me.

And this is exactly why some of you guys shouldn't be so hostile towards out-of-band transaction systems, even if that results in some degree of localized centralization.  For that matter, member2member transactions on MtGox is localized centralization, but does not imply that MtGox suddenly has some special power over the Bitcoin community.

10a Use cases for Bitcoin decrease as more transactions types become economically not feasible
10b New investment in Bitcoin decreases as potential use cases decrease
10c Additional integration into Bitcoin decreases
10d Network effects decrease value in Bitcoin
10e BTC / USD drops



Ask Friendster what happens when you can't f'in scale. Anybody remember Friendster?

Some of you guys are really overthinking this.

More guessing.

While BTC/USD > 0
(15. Alternate chains become more and more popular.
16. People sell bitcoin and buy altcoins
17. BTC/USD exchange rate decreases. Profit of bitcoin mining decreases
18. Miners switch to SHA256 altcoins
19. Goto 15)
20. End of bitcoin

Have you encountered the term "merged mining" yet?

If you are not familiar with it, take a look, as you seem to be trying to reach out and find it...

-MarkM-

Not entirely, the main/root chain effectively secures all the value for all the chains in the hierarchy.

You could also have software that can handle the promote/demote system efficiently.


You might even be able to merge the signatures in some way.  The transaction would have to be valid in both chains.