Topic: How a floating blocksize limit inevitably leads towards centralization - page 23. (Read 71590 times)

I agree with this. It's not time yet. Free transactions actually get into the blocks quite well at the moment. They are not even significantly delayed. I'd say that if a free transaction gets stuck for over a day (144 blocks), then it is already time to start planning the change.

How, I'm asking again? The proof of work will naturally favor smaller blocks that can be spread faster to majority of the network (and thus support decentralization). Yes, worst connected nodes will be left behind, but as long as there are plenty of "middle class" nodes that will naturally favor blocks less than certain size and slow propagation of oversized blocks making them into orphans more likely, I don't see a problem.

You seriously insist on making mining or running full node possible through 1200 baud dialup? Please don't be ridiculous. 10mbit-like satellite connection is affordable in most of the world now.

The miners with slower bandwidth will be outcompeted by miners who have faster bandwidth. In the same way as miners with lower electricity costs outcompete miners with higher electricity cost. In the same way as miners who can use the heat of their miners for heating can outcompete miners who can't. In the same way...

I may be fairly privileged to live in a country where a 100mbit connection is fairly standard these days. I believe a 2mbit connection is the legal minimum over here, that must be provided for all people. That will probably change to 10mbit in a few years, which is probably the same time we'll be seeing gigabit home connections.

The same goes for running a full node. Not everyone can do it now, nor can everyone do it in the future. It doesn't mean Bitcoin can only be run by megacorporations in the future. Based on the calculations I've seen in this thread, even if Bitcoin has the amount of transactions PayPal has, running a full node is still well within a decent computer setup with a good connection.

I do think it is kind of black-and-white.

Basically until fees get so insane that the vast majority of stakeholders insist "something must be done" it is probably not time to change the max block size.

Either the fees people have to pay to get reasonable-value transactions into the blockchain are too high, resulting in effective transaction fees people actually have to pay to get a transaction into a block within about six blocks or so skyrocket or they are not.

It is a pity that generally the lower the value of a transaction the faster the person doing it, or person receiving it, tends to want it to be confirmed, but on the other hand the smaller the value of a transaction the less likely it is that the merchant who gets hit with the occassional chargeback by handing out product without waiting overnight for transactions to clear will be bankrupted by not getting "instant" irreversibility.

As long as paying a few cents suffices to get you into a block within an hour or so there probably is no big need to increase block size.

Heck right now people are still able to get free transactions into blocks, until that is no longer possible there seems time yet. Should free transactions ever get in, really? Even after many days? Or would a free transaction getting in within a week or so indicate there is still not full paid utilisation of the already available space?

-MarkM-

At first how would that work? Are they supposed to accumulate transactions and make bitcoin transactions to somewhat reflect the micro-transaction balance at intervals? How do you suppose the compability between them works? You are looking at an ever increasing demand for storage capacity and network bandwidth here. This overhead is at worst O(n!).... with every new chain.

Second: What's the incentive to do it? If you go through the hassle of creating another protocol why tie it to bitcoin? Not only you miss out of the early adopter reward you create other problems in the bitcoin economy by deluding the supply of the transaction market.

The proof-of-work system.

You aren't paying attention to the debate.  This is about miners with slower internet bandwidth.  Try again with a 1200 baud dial-up connection and see how long it takes you to receive the next 1MB block when it is added to the blockchain.

There are most likely methods of adjusting the block limit dynamically in a way that is not easy to game. Even if there is not, I would still advocate for a one time increase of the block limit. Artificially keeping it at a fixed level forever is absolutely ridiculous. The limit can be much higher than 1 MB without any significant setbacks, and thus we could handle a lot more transactions.

Some people in this thread are trying to make this into a black & white issue, which it is most certainly not. My view is probably closest to Pieter Wuille's on this, in that increasing the limit needs to be done, but it needs to be done carefully.

I could easily turn the argument around as well, I mean what are the proponents of keeping the limit fixed afraid of? More centralization I guess. Well, if it was for "pure profit motive" as you call it (which is false), early ASIC companies would have the incentive to simply take 51%. On cost basis it's an absolute no brainer. That's it for Bitcoin. However, the fact that it is very bad for Bitcoin, likely stops them from doing that.

Think merged-mining. Don't think of it as alternate currencies, just alternate chains. They are not competing with bitcoin, they are more chains that are all part of the same family, bitcoin providing high value high security international settlements and the basic security shared by all the merged chains, other chains providing space for smaller value transactions, whatever stuff the primary chain becomes too expensive to be economical to do. The same miners mining all the chains or however many of them they choose to.

It is not competition, it is co-operation.

Floating the blocksize would maybe lead to more centralisation in chains too, detracting from miners' motivation to merge more chains, instead leading to them ignoring more of the merged chains than they otherwise would do when the primary chain starts to be full-to-overflowing with high-fee transactions.

-MarkM-

I understood the argument as "big miners/validators will mine too big blocks, leaving smaller users unable to receive/verify them behind". My objection i don't see answered is, what stops other miners to spontaneously build longer chain with smaller blocks that more easily propagate? (In absence of said cartel.) Everyone starts with strawman "we will inevitably have single primary megaminer/supernode, heeelp!" and argues from there. Why? How? I'm running myself two instances of bitcoind on one 5 years old desktop PC with dedicated 5400rpm HDD on 10mbit internet and it does fine. I estimate this setup would happily accept 10x, and with some updates 100x current volume.

Miners can game the system too simply, such as by lots of them all deciding to only accept transactions with a high fee then spamming high fee transactions in concert ("I will put a high fee transaction if you will, between us we have huge chance one of us will get all those fees").

If it were not for that I would think judging by the fee required per kilobyte to get into a block would be a useful metric, since by relating that to the smallest transaction the primary chain "should" be used for (secondary chains existing for lower value transactions) and the percent the fee "should" be on that smallest value transaction in order to borderline discourage it one could tune the size to achieve the kind of scale of minimum value and percent fee on that minimum value.

So I do not think floating the size is a good idea.

Maybe Satoshidice could serve as an estimation tool, we could watch to see how much fee people making tiny (low-value) wagers turn out to be willing to pay to play, maybe it will turn out that they tend toward making fewer, more value per wager bets when blocks are full enough that Satoshidice has to increase the fees it is paying? If so the "problem" might be self-correcting for a while, and meanwhile many miners are already stocking up on secondary-chain coins as a currently not immediately-profitable but forward-looking natural consequence of merged mining so small-value gamblers and thus maybe small value transactions already have alternatives to continuing to clog up the primary chain with "trivial value" transactions.

-MarkM-

So you are concerned with competition from alt-chains? Is that what this about?
Well that's life.


How is there supposed to be a market if there is unlimited supply? A limit which automatically raises if it is reached is no limit at all, just a bait and switch game.
The block size limit is exactly as artificial as the total coin limit, should we rise this as well?

I have just realized that this is a major crossroad for Bitcoin. Leaving the limit as it is would lead to transaction fees rising to such levels that no one would use Bitcoin for sending money unless it's absolutely necessary and the amount is large enough to justify it. Another cryptocurrency would take Bitcoin's place as the currency for day to day exchange. What would happen to Bitcoin? If the competitive cryptocurrency has similar features... scarcity of the monetary base etc, would Bitcoin not simply wither away? Maybe, maybe not.

The other option is to compromise and decide how crowded the blocks need to become before something is done. This would lead to the risk of in my opinion marginal exploits and marginally more centralization. I don't think the third option is very viable, removing the limit entirely that is. For many reasons.

I'm not afraid of the markets nor do I think it's actually possible to achieve super cheap transactions with Bitcoin regardless of how this is handled. The problem is deeper and has to do with proof of work. The security that is achieved with proof of work does not come without a price. In the very long term proof of work / proof of stake hybrids will most likely be able to handle more transactions with less hashing power and less cost, and still be reliable enough.

I have long since accepted that Bitcoin transactions will cost more in the future than they do now. That is basically inevitable. The question is how much should they cost. The block limit is essentially an entirely artificial limit that's quite unintuitive. In my opinion the only truly valid reason to retain scarcity with the block limit is to keep miners incentivized in general. There needs to be scarcity, but not too much scarcity.

The solution in my mind for now would be to only increase the block limit when the current limit has truly reached its limits. We're not there yet so there is no immediate problem. The ideal solution would be to create a not-easily-rigged method of automatically adjusting the block limit. That can be implemented later, though.

The ability to frivolously use the primary chain being "severely handicapped" need not be a "bad thing", it could actually be a "good thing". Blockchains are not ideal anyway as payment networks, so trying to compete in the payment network space is probably doomed to failure from the start. Storing value securely and moving significant value securely should not be sacrificed for the ability to tell the whole world about every penny-ante gambler's every wager and every child's purchase of a popsicle.

Let the transaction fees on the primary chain go up to whatever the market pushes them to, lower value transactions can use the first secondary chain for its presumably lower transaction fees (possibly also with lower security if less miners bother to merge that chain), people wanting to transact for free can migrate to the least of the secondary chains as long as there remain enough chains in the family for one or more of them to still offer space for free transactions.

Throw in Ripple to make exchanges between the chains transparent, or clients that do atomic cross-chain exchanges, to hide the underlying many-chains "complexity" from users who prefer not to think about the nitty gritty technical details of how money is implemented. Add more chains as needed.

-MarkM-

For the blocksize limit removal proponents: What are you afraid of?

That you cannot make transactions for basically free anymore? Don't you get it?
The limit is there for a reason... some of which were already mentioned in this thread.

Maybe once we reach the limit there will be finally some steady income for miners (read not related to the initial block reward).
If the limit is reached consistently there would finally be a market for transactions which wasn't there before. You preach free markets and shit, but when it comes down to it you are afraid of them.

I just don't like the fact that there is deliberate attempt to steer Bitcoin towards being a gold-like settlement currency. It can retain a high value like that, true, because it still offers significant advantages to gold. However, it would lose all appeal as a day to day money transfer mechanism.

Maybe centralized systems need to handle that, or other blockchains, that is possible. The problem I have is that I see no problem in higher requirements to run a full node. Nor do I see a problem with the mining market, which is bound to become more centralized regardless (thanks to the cost structure of ASIC manufacturing). Based on the posts in this thread there are many possible methods of making it difficult for miners to start seriously screwing it up too much.

I can accept that Bitcoin won't be used for super micro transactions, that can be done elsewhere. With this I mean transactions worth cents. However, paying the equivalent of $1 very cheaply, internationally, should and needs to at least be a goal for Bitcoin. That's what I think.

I think you know well that the original creator of Bitcoin, Satoshi, has clearly stated that he intended for the block limit to be raised or removed eventually. This means that for all intents and purposes, Bitcoin was not complete regarding this issue.

It's pretty much the same thing as the fee structure put place in Bitcoin. Everyone knows it's not the final system, it was just something that was put there to make it work for now.

This shouldn't be compared with changing the inflation schedule. That is absolutely false. The inflation schedule is a part of Bitcoin that is complete, and has been complete, ever since Bitcoin started. I also think that pretty much everyone agrees with that.

Where do you think the value in bitcoins comes from? They have no use beyond their capabilities as a medium of exchange and store of value. What happens to the value of bitcoins if the ability to use bitcoins as a medium of exchange is severely handicapped? Not good things, I'd reckon.

It is true that Bitcoin can be used together with various other means and maybe that is how it will eventually be, but that is not the most efficient goal. Bitcoin has the promise of being universal money, which means it can be more or less, a jack of all trades. There is obviously a large amount of skepticism regarding this. I agree that it may not be possible.

However, I do not agree in not trying. It is a massive step down in efficiency if Bitcoin has to be used together with other systems to be usable for anything else than value storage. That is not what we should be aiming for, far from it.

"Making"? The system already has the scarcity built in, and the suggested removal of it would require an incompatible hardfork no technically different than changing the inflation schedule. Bitcoin has never been changed like that.

If you're going to go for the hyperbole: I'll accept that and raise you—  Violating the contractual agreement embodied in the Bitcoin software in order to remove some of the scarcity which is currently necessary for making the act of honestly expending resources to secure it economically rational. This would be a change which has a universally anticipated consequence of increasing centralization, forcing an unknown number of participants to trust third parties, and removing all reason from the long future security arguments of the system... Peter outlines an argument that the equilibrium of an uncapped blocksize is a single miner. All for the sake of increasing short term transaction capacity when we are no where near utilizing the levels currently available and when that capacity could be provided by other means. This is what we're basically talking about.

... I think this is horribly over-stated, but it's at least more pedantically correct than suggesting that anyone is proposing adding any limitations.

Huh? It's the arguments that the pools (note: miners are not even invoked here, the idea that the authority is handed over to a couple pools is already so easily accepted) will reject "too large" blocks that demands a conspiracy.  Peter Todd's point is that a single miner can push off all the smaller ones, all on their own— unless there is a conspiracy to stop them from including these otherwise valid blocks.  As you argue such a conspiracy "will result in abuse of current rules", and I agree.  Worse, if we as a community can't figure out a criteria to conspire on and put it into the system— where there is no risk of defection— how can we expect a pool-conspiracy to do better when they have less powerful tools to enforce conformance?  Also, while the network is choking on blocks which will be rejected users are left around waiting longer times for confirmation because its not obvious when a chain is going to get orphaned because it violated the invisible cartel rules.  Finally, even if the cartel is successful at preventing single miners from breaking the system, its in each member's individual best interest to mine the largest possible blocks, so we should expect a slow evaporation which has the same conclusion: a single primary miner/validator.

Not quite— only if you assume miners don't directly peer (either intentionally or by connecting to everyone). They do.  Besides— a single megaminer has no orphans. Miners and non-mining users are different, counting orphans only count the impact on miners. If you disenfranchise the non-miners from validating you're handing control to a special interest group with high fixed costs— the kinds of blocks you can get away with mining are very different if you're trying to appease all users broadly vs just miners.

Using orphans may well be a valid way to dynamically control block speed without a fixed time constant, but it doesn't address the concerns around the economic motivations of mining, or the disenfranchisement of the non-mining users whos voice is embodied by the strict rules of the system which constrain miners absolutely no matter how powerful a miner conspiracy they have.

I'm interested in hearing more thoughts about how to safely integrate other blockchain like systems... e.g. what protocol extensions they need to share more security and facilitate cross chain trades. I've generally been a bigger fan of non-chainlike alternatives because I think the diversity has the most potential to make a big marginal improvement— but I also think there is room to add value with chainlike systems too.

Bitcoin is just the primary chain in a whole family of merged-mine-able blockchains; it provides, primarily, the world's most difficult proof of work.

Anything other than providing the world's most difficult proof of work can be delegated out to one or more of the secondary chains, most of which currently seem to in any case be mostly "backed" by bitcoins thus not only have a vested interest in keeping the world's most difficult proof of work difficult but also in keeping the coins/tokens of value of that primary chain, which provide most of the funding for the massive proof of work, valuable.

More transactions per day, faster transactions, anything other than sheer proof of work power and sheer (store-able?) value (or, at least, some means of funding that difficult work) are secondary thus can reasonably be delegated to secondary chains, surely?

-MarkM-