Topic: Why the anti-blockstream sentiment? (Read 894 times)

This isn't a problem. A fork would only happen when a huge amount of miners is mining with the alternative clients. So an attack would be more likely on the losing chain. But even then. No miner would be stupid enough to make such attack. It is very unlikely that someone could gain from such a theft because who will be stolen from?

And besides that... you would lose a lot of money mining with that hashpower only to scam. You could make more safe money by mining normal blocks.

Obviously larger blocksizes increase the orphan rate, but if you compare apples with apples like 1M@10m and 50KB@30s then you get higher % of waste hashpower with the 50KB@30s. The average block propagation time to miners for 50KB is not 20x faster than 1M, there are lattencies not linearly dependend on blocksize.

And even with average 33 seconds, it is not enought at point of sale locations to wait up to few minutes to get unreversible transaction. You need at point of sale locations confidence for 0-conf transactions anyway. And dont talk about LN because it does not work yet (and there is chance LN will not work in decentralized manner - so LN could to be turned out as just another centralized option which existing already - fund debit card with Bitcoin and use it anywhere with instant transactions)

Well I'm still relatively new to Bitcoin, but my understanding is that larger blocksizes increase the orphan rate, and block generation can be effective and efficient at a rate of one block every 33 seconds. I'll try to find the statistical paper that contains those calculations. This seems to be especially true as transaction volumes increase.

There is no point to make average block creation times faster, you only increase orphan %. For instant transactions you need to improve 0-conf transaction reliability, not doing the opposite with RBF blockstream feature (but who would use blockstream second layer -only works in theory so far- and pay fees to non miners then?).

Can we be clear about one thing. The real demand is not for larger blocks, but for faster transactions. Increasing the blocksize is just one trivial and easy to understand temporary solution. I believe that true growth in the mainstream will require far more than that. I also believe that the blocksize should be halved to speed up block generatiion and confirmation.

With crippling Bitcoin I mean artifically restrict blocskize capacity even though there is higher demand. To get the idea lets imagine regulator sets the maximum blocskize at 10 KB. The advantage might be rasberry PIs and other obsolete computers might be full nodes and even solo miners for many decades, but you really hurt any further Bitcoin adoption and even current userbase has to shink because the onchain blocksize define how many people can use Bitcoin - and why protect obsolete computers to be as possible full nodes?

I dont mind if every miner set his fee policy and maximum limit of transactions he adds to the block regardless of number of unconfirmed transactions, I just dont like someone else trying to set fee policy with artifical blocskize limits instead of free market finding economically sound fees + blocksizes. Current government systems are already over regulated, so why Bitcoin should join this regulatory path and not let free market find the best values - no human regulator can beat the free market.

Yes, but I think of on-line wallets as a form of sidechain.

Think of security issues.

If the side chain is not equal secure than the main chain but lower, you not really want that, since than your BTC can be stolen from a side chain attak.

Of course there are blocks that are not full still. Though there are times on the day where all the blocks are full. It doesn't help when you tell someone that he could get his transaction confirmed somewhere in the night because then there might be half full blocks. That is effectively killing one of the advantages bitcoin has.

What things of the foundation have changed?

I doubt that you would find nearly enough consensus to change the 10 minute blocks. Besides that, orphaned blocks rise in occurence the smaller that timeframe. Which makes it more effective to deliver zero transaction blocks.

Bitmixers will always stay on the main chain. They need it for their service to work. And their customers use the service for a reason so they would pay more than the average guy for it. Bitmixers will survive on main chain for sure.

verbose transactions are things like pet obituaries, or long text messages.
Under utilisation of blocks means leaving lots of potential unused. It's not just empty blocks, but blocks under 500k for example. If a block is under 500k, then presumably the miner would have added extra transactions for the additional revenue. The fact that he hasn't probably means that those transactions were not available.
I believe that a number of things that were in the "foundation" have been changed, why not average block interval? 10 minutes is too long, and it could be 20 or 30 minutes in some circumstances. Increased fees or blocksizes won't alter that. I think I read somewhere that Bitcoin can work efficiently with a block interval of 33 seconds. As transaction volumes increase, we will need to work towards that to provide an efficient and speedy service - reducing the blocksize would help to achieve that goal. A three minute block interval with a one sixth of current miner's reward seems a good short term solution instead of halving.

I've also been looking at some stats, and it seems that long chain transactions are distorting the figures. Maybe we should penalise bitmixers, or move them onto sidechains.

The long block generation delay is a real setting from the foundation of bitcoin. The blocksize limit is not. It was set arbitrarely to protect against spam. Which it never did until the blocks became full and it became affordable to spam.
It wouldn't be so bad to wait 10 minutes for a confirmation IF you could predict the outcome somehow. If you can't know if your transaction will go through at all and you have to wait maybe alot of hours then this is way way worse.

I googled what you might mean with verbose transactions but I did not see what you might mean.

Under-utilisation was never a problem when blocks only were full a couple of percent. Why should it be when blocks were full a third again? Or do you mean the zero transaction blocks? Yes, that is surely not helpful. There should not really be a an advantage anymore doing that. There are ways to propagate full blocks in a matter of seconds already. Including >50% acceptance.
But this behaviour can't be stopped except with a hard fork. Which surely won't happen.

SPV Mining should have become a smaller problem now. Those doing it promised to verify the transactions of the previous block though while they mine on that block already. Verifying takes time so it is understandable. It will happen again probably but it should not be a big problem as long as the company mining that blocks doesn't have a huge share in total hashrate.

Interference with node sync? You mean ddosing?

Invalid transactions look like a problem to me too when spammed. Though it should not be a huge one except delaying nodes with the calculation. I did not hear of an actuall attack through this.

The future of bitcoin inevitably needs a raise in the blocksize limit. Otherwise you will be stuck with what we have. Segwit can only do his part until there is no way to improve. And not using bitcoins but sidechains instead... well, that sounds like giving up on solving problems. Surely I will not have my bitcoins being bound in a payment channel... and so on...

I don't think blocksize is the most important limiting factor. A few other factors slowing the network are -

Long block generatiion delays
Verbose transactions
Under-utilisation of blocks
SPV mining
Interference with node synchronisation.
Invalid transactiuons

and a number of other factors that more knowledgeable Bitcoiners have mentioned

Increasing blocksize is just a trivial easily understood attempt to solve a fairly complex problem, without considering the future of Bitcoin imho.

Sidechains are great in fact. The problem many people have with blockstream is that bitcoin is working worse and worse because it is restricted by blocksize limit. The developers don't want to fix this but instead suggest using the lightning network, which comes from blockstream. So alot of bitcoiners feel like they get extorted by the bitcoin core wallet developers to use the system they create besides bitcoin.

That is the problem many have with it.

How does it "cripple Bitcoin"? Its not equal to the argument for removing the limit altogether; youre just ignoring all the points people made here

This idea has been floated a few times over the years and I'm not sure why it never seemed to gain any traction. But I don't know much about it specifically.


Well empty blocks are empty blocks. They don't take up much room. The bigger issue there is that the miner isn't validating, so he could be mining on top of an invalid chain (and maybe taking some SPV nodes with him). The other thing is that miners like Antpool mine 10-15% empty blocks, while arguing for bigger blocks. Of course, with their hash rate, those empty blocks could handle a big chunk of network congestion at any given time. Personally I think they need to fix their own mining code and/or upgrade hardware so they can cut out SPV mining and empty blocks before trying to push anyone into a hard fork for the sake of capacity.

Thank you for that post exstasie. I think I agree with you. Doubling the blocksize seems to be a trivial uneducted act, that provides no benefit to Bitcoin on its own. I believe that it may be better to halve the blocksize, and to increase the block generation frequency. Couple this with the removal of long text strings onto privately owned sidechains, and you could have a fast system that could cope with a very significant rise in traffic. Doing something to provide rewards for mining empty blocks could also help. Why should I as an unpaid node provide permanent storage for an empty 2Mb block that some selfish miner received a large chunk of money for cluttering the Blockchain.

The fact is blockstream want artifically limit onchain blocksize so fee market has to develop sonner than would happen if you removed the limit and free market finds the best blocksizes. So yes, they try to act as another regulator who knows best and have no faith in free market to find ideal economically sound blocksizes (and fees) thats why I canot agree with them.

And trying to cripple Bitcoin in order to speed up creation of level 2 solutions is as good argument as removing limit altogether and speed up creation of solutions to prevent attack vectors on bigger blocks (BIP109 is already progress in this area which limiting sig operations thus not possible long block validation times anymore)

The trouble with these people is they think that bitcoin has no limits. That's silly. Does Visa or Paypal have to process data so inefficiently as Bitcoin? No way! But that's how we make it trustless--every node validates every transaction.

But that has costs. So allowing endless freeriders who want free or at least extremely cheap (1-5cent) transactions is one option--and we can achieve that for a (perhaps very) short time by simply increasing the block size. But free riders will continue free riding and anyone that knows statistics knows that the 80/20 rule will not provide users much relief if they want free transactions. We can endlessly increase the block size in this way, but data validation and relay have costs, as stated above. You jack up those costs and nodes will drop off the network like flies like they already have the last couple years as block size grew above and beyond retail bandwidth and processing improvements over the same time.

What this boils down to: Bitcoin is a broadcast network. Broadcast networks are generally considered to be not scalable. On its face, that seems problematic. But as with IPv4, we don't need to limit ourselves to the Internet layer. We can build additional layers on top, to deal with its shortcomings (e.g. incomplete, non-ordered, badly-error-checked packets)... So we build TCP and UDP to maximize reliability of data streams and reduce latency. Above that, we have the Application layer--DNS, FTP, HTTP, SSH, SSL, Telnet.....

Similarly, we can build bitcoin in layers. BUT not if the bottom layer is broken. Bitcoin was built and intended as a P2P system--that means retaining a decentralized bottom layer is of utmost importance. P2P loses its meaning if nodes on the network are centralized--and that is inevitable if we do not control the bandwidth nodes are subjected to--because without decentralized validating nodes, users (who are probably running SPV clients or online wallets) have to trust an ever-smaller group of nodes to validate. That means transaction censorship becomes more likely. It means double spends via sybil attack become more likely. And this last point is especially relevant as block halving approaches: the cost (lost reward) of mounting double spending attacks drops in half. Every halving adds incentive to double spend, and the only thing that will ever address that is an active fee market.

I think the main anti-blockstream sentiment comes from people who believe that the protocol are deliberately crippled to favor the creation of these side chains. You would in theory not need them, if they simply bumped the block size as suggested by the Classic proposal. I think the other side is saying, bigger block sizes will increase the security risk for attacks and that their suggestion solves both problems. I can see merit in doing both, without compromising the security and also adding extra functionality and scalability. < SegWit>

Private or public ownership of sidechains doesn't really matter, as long as they are helping to keep the main blockchain public and decentralised. In fact private ownership (of sidechains ) should be good for the Bitcoin ecosystem.