Bitcoin Forum

....

re: average block size - wouldn't it make sense to look at the growth rate as well? Zoom out a bit on the chart and see how it looks:
https://blockchain.info/charts/avg-block-size?timespan=all&showDataPoints=false&daysAverageString=1&show_header=true&scale=0&address=

i think the point is that we cannot wait that it become necessary to increase the blocksize, to actually increase it, we must forecast this, before it happen

if it happen next year what you will do, proceed to fork anyway and built even a bigger mess than what we are facing right now?

if it must be done at some point, it's better to do it early, it's always like that for every thing, not only bitcoin

Sure. Disregarding the spike, even in the chart you posted, you can see the average block size has nearly doubled in one year. If that continues, we'll hit 0.8mb per block in a year or so, and that's assuming a steady growth rate. A year or so to start hitting the 1 mb ceiling seems like a safe estimate if you want to figure out what to do about it before you find out experimentally, like it or not.

but it doesn't mean it would double every year, unless the number active wallet grew explicitly fast, it would be impossible to double the block size per year. plus, i am tired to see most of the junk transactions in the blockchain.

Not every year, no. Next year, though? I don't think you can rely on that not happening.

i mean, even though it would happen, i still would not let the blockchain to increase the size which is unneeded. that would harm newbies to use the full node wallet. doubling the block size limit would ever encourage the spammers to spam in the blockchain network if the active users did not increase. that's shit-awful with those junk transactions.
we need a sustainable currency to help bitcoin mingle with the world , just like a currency joining to the IMF.

Honestly, I think Joe and Jill Average running full nodes is a pipe dream, whatever we do with the block size:

-If you increase the block size indefinitely, the storage and bandwidth requirements will become high enough to deter casual users.

-If you don't increase the block size but increase fees instead, the fees will become high enough to push casual users off the blockchain, to use payment processors etc. instead. This leaves them with no reason to run a full node even if they technically could.

-either way, I don't see many apart from a hard core of enthusiasts wanting to run a node.

I agree. It will become a problem at one point and it obviously needs to be raised, it's just to what and and when that people can't seem to agree on. I don't think it needs raising it 8 though.

Thanks for this.  I appreciated your summary as I have not seen a good one elsewhere.  (If someone can point me to another one I'd appreciate it.)

I don't think increasing fees is a very attractive selling point for bitcoin adoption.  The block rewards becoming very small is very far out in the future.  As even your chart points out, the blocksize is much more of an immediate issue.

I'm afraid you can't really force your #4, moving people off chain, though I agree there is a lot of dust transactions, not sure why though, it is all satoshi dice, faucets?  I fear we are close to having #5 already.

Totally agree on #6.

Thanks again.

After 5 pages in your last thread, which is basically a repeat of the same flawed premise, I can't help but notice you still haven't answered my question, so I'll pose it again here:


We've considered that just fine, thanks.  What you're not considering is the following:

Assume Network 'A' can process a maximum of 2000 transactions per block and the average fee is .0001

Assume Network 'B' can process a maximum of 10000 transactions per block and the average fee is .00005

Which network can potentially generate a higher amount in fees?  That's right, the one that can support more transactions.  Raising fees is not the most efficient way to increase mining revenue.



And also this.  There's a moral argument to go along with the one where small-blockians aren't very good as simple math (more transactions equals more fees, not less, it isn't rocket science).  The more people who are able to use Bitcoin, the stronger it becomes.  What some people think of as a "lesser" transaction could mean a not-insignificant amount of money for those in other parts of the world.




Still waiting for an answer on this one.  Be honest.

So what you're telling me is you're going to keep repeating the same thing over and over because you don't have an actual reply to a question I'm starting to think you haven't understood?

Try working under the assumption you're wrong about "decades away" (because I'm pretty sure you are).  If all the transactions you (wrongly) call spam were already forced off-chain and we frequently have enough "legitimate" transactions to fill a 1MB block, how large would the backlog have to be in megabytes before you would support an increase in blocksize?  

If you are incapable of answering this simple question after numerous attempts, I have to assume your brain isn't functioning correctly.  Please give me a number.

When miners will need funds to secure the network, they won't include too cheap transactions. Who wants their transaction to be included in a block, will need to increase the fee. Competition between miners will allow to discover what is the right fee. If now fees are low, it's just because the new generated coins cover happily all the costs. Miners are not stupid, let them do their work. If you think you can do better, start your own pool.


https://blockexplorer.com/blocks-date/2015-08-16

It seems that in some periods of the day we are already approachin the limit. When would you wait to increase the block size? Your reasoning is like closing the fence door after all the pigs are already escaped. I remind you that the white paper didn't have a block size limit and it has been introduced only for testing purposes, with the intention to remove it already in mind.


When other off-chain technologies will be able to offer instant payments, you will not pay for your coffee with the blockchain, even if the fee is 0. And at the same time, if the block size is large, legitimate transaction will always have theri place.


I don't see anything useful on the rise of the fees. What makes bitcoin special is the adoption. Small blocks means smaller adoption. Small blocks are the best way to lead bitcoin to failure.


Just remove the limit and let the miner decide.


The hardfork happens when the largest part of the network agrees on the new version. At that point there is no confusion.


If you don't want to centralize bitcoin under a couple of developers, allow for its extension.

Please stop the fud.


People: if you love bitcoin, install XT. I already did.

I agree all 6 reasons here. We should get rid of that dust transactions, so we won't need any larger blocksize for a long time. Transaction fees are already low, just double it and we don't see half of the dust.

The question remained identical.  And you still haven't given me a number.  If you honestly believe a global populace can be contained by two-thirds of a floppy disk for another decade, I'm sorry, but there's no hope for what little credibility you have left.  I can only assume you'll happily sit back and watch while the user experience degrades to a point where the forums are packed with complaints about Bitcoin being so slow and unreliable that it's no longer of any use.  Your precious chart shows what's happening now, not what will happen in future.  You cannot predict how large the userbase will become, or how rapidly it might grow.  If Bitcoin sees a surge in adoption and the network can't support those transactions, your words here will come back to haunt you.

Great summary OP.

The next thing is:
Increasing blocksize will also make BTC inaccessible to half of the world - especially the unbanked in developing countries (all of africa, most of south america, most of asia, most of pacific region) due to bandwidth.

... and bigger blocks lead to more centralisation... not even talking about the risks of a meltdown in marketcap in case of a controversial hardfork ...

It's really all not even rational... and i tell you right now: it's not about bigger blocks; it's about a hostile takeover

This whole blocksize-increase-debate is irrational at so many levels and in so many points. Why is bloating the chain ad infinitum even an option for longterm growth? The chain is already so large most people don't download it and downloading the chain is a major hurdle for people to enter the btc ecosystem. Why even raise the bar for people for adoption ahead of demand like you point out?
All this is not rational and because it's not rational we know it's not about the tps or blocksize, it's about getting people to accept their alternative xt client to take over bitcoin entirely. Basically the blocksize drama is an attack on bitcoin by Gavin and Mike.

They thought people were stupid and would go along but their populism campaign in social media failed (though they are still trying hard). It's politics, not technics.

"We need transaction fees to go up, and the only thing that will force them up is the blocksize limit."

We don't need fees to go up, cheap transactions is one of the major points of using bitcoin.

This is an argument for increasing the block size.

"Once we hit the 1 mb limit free and nearly no fee transactions will be forced off chain. There is TONS of dust, spam, and gambling that can be done off chain. Over half of all Bitcoin transactions right now are basically dusty junk. This will make plenty of room for legitimate transactions."

Decentralization is one of the major points of using bitcoin, off-chain transactions are centralized, again, this is an argument for increasing block size.

Don't know what a "non legitimate transaction" is, as far as know, if it follows the network rules it's "legitimate".

Tradeoff with accessibility issue ... would you rather let everyone pay 5 cent for a tx or would you rather exclude half the planet from using it?

And who is mining for free? Raising max txs ahead of demand will weaken the network in the future aswell as exclude billions of users and push centralisation.

Arguing for a contentious hardfork, hostile takeover of btc by Gavin/Mike and all that is really a lost cause...

Then we need to increase transactions on-chain to generate more fee revenue, not the other way around, off-chain transactions are the centralized ones, unless you have a real proposal for decentralized off-chain transactions...

If users choose XT, XT it is!

And because of bitcoin network nature changes need to be planed way ahead.

YOU don't CREATE transactions. They occure or they don't. Bloating the chain won't make more txs magically appear - well, maybe spam and dust-junk increases. But when you want a bloated chain full of junk, why would you want to tarnish bitcoin for that?  

Increased spam and junk is not genuine growth.

I meant with user adoption, in the future, eventually there will be much more transactions on the network, they should occur on chain, at least for people who prefer that way, people should not be forced to use some centralized service because it becomes too expensive to transact on-chain.

Well, RIGHT NOW it's FREE to send coins around and blocks aren't even half full so i have NO IDEA WTF you're even on about.
The future is the future. How do you know for certain these txs will occure? How can you be so sure about it? Do you have magical abilities to know this? What happens in the future in case your predictions of it are wrong? Please elaborate on that scenario.

http://alojaimagens.com/images/9td0dub6eaoccxlwn03e.png

https://blockchain.info/en/charts/n-transactions?timespan=all&showDataPoints=false&daysAverageString=1&show_header=true&scale=0&address=

I have no reason to believe bitcoin has achieved its maximum user base, I'm betting on it, otherwise I would be putting my money in something else...

Accepting and producing bigger blocks will lead to more centralization than using off chain transactions. The rate at which the chain can grow and be spam bloated with larger blocks is a serious issue and almost nobody will run their own node.

Lets not forget there are only around 6000 full nodes as it is right now.

However, Satoshi did foresee a future where the chain was so large that people would only use 'client only mode' (spv wallets).

Stress testing at least proved one thing: Even the blocks are all full, majority of users will not be affected. Blocksize should only be changed when there is a REAL problem, not by fear and imagination

Human are all adaptive, when banks closed during weekends, they don't change bank, but wait for it to open on Monday, and reduce their transaction frequency

That's all fine when you control all the software being changed, bitcoin network is a bit different and changes have to be planed way ahead, not when there's a problem, for everyone to upgrade, or at least a majority, it takes probably a year.

False! Bigger blocks = more transactions = more fees!

That means on average, it also means that at some times during the day, blocks are constantly full and the queue gets long, exactly when people need it the most. If you want the blocks to be full, then you can lengthen the queue to weeks (just like regular banks). You don't want to build a highway that needs to be 100% filled with stationary cars before you decide to put more lanes in it.

False! Free transactions have a pre-allocated spot in each block. They will still be allowed under normal settings, even with high fees.

False! If the blocks are full, fees grow exponentially to obscene levels, such as "I just paid 1 BTC fee to send my other BTC to an exchange to dump this shit, because everyone else paid 1 BTC fee!".

Too many chefs, not enough cooks. They each think that their idea is the best and other ideas are shit. What if they never reach consensus? This shit is being discussed for over 2 years, the actual block usage increased steadily, and now needs to magically be capped at some random number, with no plan to plan for an increase (yes, there is no plan for a plan, they just push the decision for later).

It also amazes me that NOBODY is capable of understanding the social, economical, technological, networking and computational aspects of this change, as a whole and everyone conveniently ignores some aspects to push his propaganda, such as this example list of false reasons to not prepare for an increase in the block size.

1) bigger blocks don't create transactions in and of itself. Larger blockchain probably prevents further adoption or at least slows it. Most txs you'll get will be spam which goes away the very second you charge fees. (lol)

about 2) complete bogus. Fees rise and huge backlogs won't happen even with smaller blocks because fees will drive junk txs off chain. So you got a huge whole in your argumentation there. The doom and gloom backlog-story is a lie. Won't happen. No evidence that supports this idea.

4) No, idiotic idea of exponential fee rise. Do you think btc is the only way to do txs in this universe? No it isn't. Fees will be found in supply/demand  freemarket-fashion (this mechanism never fails and no other mechanism is more efficient, free market, ya know bro? Instead of central planning, got it?)  and txs will cost exactly what people are willing to pay to use bitcoin instead of a substitute. So exponentially rising fees sounds like something right out of Hearns' ass. Again: no evidence to support this idea, niet, nada, poof, zero

all made up crap ...

use da brain, folks, use it!

+1

when blocks were full i had instant confirmation at all time with standard fee included of course. I had no delay during all the spamming cuz i way paying appropriate fee to get into blocks before all the broke asses that couldn't afford 3 cent tx fee and were crying about it on the forum.  ;)


The possibility of tx volume dropping like a rock due to scenario a, b, c or  d is almost never discussed in this forum. What you gonna do if txs drop off? Core could deal easily with a sharp drop in txs and could recover later. XT on the other hand could get into trouble if txs were to drop sharply.

False! Bigger blocks = more transactions with very low or no fees = less fees!

If you pay a decent fee, you won't have issues even when blocks are constantly full. Only dust and spam transactions are hindered. You want to quickly build more lanes on the highway even if there is a significant risk that the bridge over the big river collapses under the weight.

False! That's because the miners are happy with their current incentives, but that will change. Anyway, many stupid dust transactions actually pay a very small fee, and they will only make room for legitimate transactions after fees rise a bit.

False! A fee market will form and noone will be stupid enough to pay 1 BTC fee. This market will reach a price equilibrium at slightly higher fees than now (an equilibrium for which at least the stupid dust transactions are infeasible). And for the long term, I don't think people would want to run full nodes that will verify and store the transactions for all the guys paying for cofee directly with bitcoin.

That list contains completely true reasons that many people, including me, agree with. I don't care if they never reach consensus. We need consensus to perform this kind of hard forks, I see everything else as an attack on the network.

The way I see things, the gullible masses (mainly pissed-off reddit people) lead by Gavin and Mike are preparing for a massive attack on the bitcoin network. Save it by destroying it!

I completely agree with you.

The joke in all this is: we would actually now be good for a bull market if Gavincoin-camp wouldn't create massive FUD for the coin.

Some of these people say scalability (it's actual pseudo scalbility) would bring in new money. But hey, folks, if you would just leave it alone and not threaten to break it price would probably be higher already. These people seem not to understand that they are spooking the investors at the periphery massively and drive away the money with their Gavincoin-circus.

reasonably, what about the people with the XT node? leave them away, and let them die for the insufficient storage?
within 20 years, not even 30 years would be full of blockchain data, and people would have no choice but either buy more hard disk, or stop using the XT node and start using the third party wallet. with or without XT, result would be exactly the same. instead, we should have the flexible block size, to prevent the network spammer. apparently, the blockchain, even stress test would not affect the full node users a lot, however, as the block size increase, the more burden of XT node users would be. no matter how fast the technology would be, the growth of technology is never reaching how fast that block grow. when the day, the block size went to the maximum, meaning 8 gb per block, unless we got the solution for the ultimately large size of hard disk, which is affordable and sustainable enough, otherwise, i am sure people would leave their XT node wallet.

Do you see how you contradict yourself? There will be no larger blocks, then you say there will be larger blocks. This is why your argument is invalid, it's self-contradicting.

The lie is your comment. It just happened this month. A backlog of 2 days for people that paid small transactions. I was affected by this because payments to me were delayed. There is no doom and gloom, it's just a fact, it can be proven ANYTIME, but you seem to be either blind or have a short memory term.

Again, contradiction. Fees will not rise, because when they do people will use something else? Are you really oblivious?

Why not?!

THIS! Cant be said better than this. The spam test showed us even blocks of 100MB is doing nothing to stop real transactions with a bit of fee to be included in blocks. Doing a hard-fork, and allowing ourselves to have 2 people running the show of Bitcoin is madness.

Ok, almost nobody. If they would send 1000 BTC then maybe. What I mean is that if someone asks you 500$ for a cofee, you probably won't buy it, but some will. There is a demand for bitcoin transactions and a supply limited by the block size. Right now there is more supply than demand, and you want to keep it this way disregarding the risks. Even if it means a hard fork without consensus and using software controlled by 2 guys with dubious intentions. This is not about the block size anymore, it's about the controversial hard fork. I want the block size to increase, but with a proposal like sipa's with the block size increase following technological growth.

I strongly disagree with TO.

You have been propagandized to believe that a block increase is possible using only XT. There are 4 alternative proposals. Don't demonize a progressive improvement in the functionality of Bitcoin because of two clowns and the mob fighting them.

When Bitcoin is dumped to 50 cents, someone will pay 1 btc to save what remains of his btc value. See above comment about brainwashing in regards to the "2 guys".

Where can I read about and download those proposals?

http://gtf.org/garzik/bitcoin/BIP100-blocksizechangeproposal.pdf
https://github.com/bitcoin/bips/blob/master/bip-0101.mediawiki
https://github.com/bitcoin/bitcoin/pull/6451
https://gist.github.com/sipa/c65665fc360ca7a176a6

increasing block size is not the solution. even at 1 TB you cannot fit all of the day to day transactions for billions of people + IoT + all of the financial stuff build ton top + micro transactions + unknown unknowns.  Lightning network is the best solution that i see so far.

More here - https://bitcointalk.org/index.php?topic=1154536.0

Then we should do the hard fork when 0.8mb is reached, not now. Don't fix it if it ain't broken.  8)

Fixing it via a hard fork to say, 2MB is easily and quickly done; consensus will be there if the limit is almost reached. I agree with turtlehurricane about waiting, though I am a bit more pro hardfork.

 It's not a "number of wallets" issue. It's a "how many transactions happen in a given (shortish) timeframe" that's going to become the "chokeup limit".

 Bumping transactions fees will NOT address this issue.


 It's a lot better to fix an issue EARLY than wait and wait then try to rush to fix it at the last minute. Some of us REMEMBER the "year 2K bug" issues, and that they were being worked on for 2+ YEARS before they actually happened - and STILL had some stuff not fixed in time, as well as had a few "fixes" turn out to be bugged due to inadaquate testing.

 It's also impossible to predict when sale spikes will happen, other than the known ones around Christmas (and probably Black Friday for Bitcoin, given the high "tech gear" level of Black Friday sales), and TOTALLY impossible to predict for sure how big spikes will be.

1) You are talking about an event that will take place over a hundred years from now .  Its just silly to talk about it.

2) Long time? it will be less than 18 months before we hit a wall. 


3) This may buy you some time but if bitcoin starts to get really adopted removing "dust" transactions won't be enough.  It also takes away a usecase of bitcoin which is micro transactions.

4) If blocks fill then people will bid up fees and compete for the right to be processed.  This will bid up tx costs and weaken another use case for bitcoin. Bitcoin needs to remain cheap to send.  Miners can make money via volume if t/s goes up.


5) Bitcoin is designed to fork and miners have a strong incentive to resolve the fork quickly.

6) The core team is only a few guys , will alot of centralized control of bitcoin.  The core team is the one real weakness of bitcoin.  We need to diversify the development process.


If bitcoin can't process several hundred transactions per second its going to be a expensive hobby and not the dominate global payments systems that we all want it to become.  ACH processes about 600 t/s vis and mastercard process thousands a second.  Nobody is going to take bitcoin seriously if it can only do 7 t/s.  The block limit is arbitrary and ulitimately unnecessary now that we have tthin clients.  When satoshi set the limit it was durign a time where everyone was running core  and the blockchain needed to be small.  We have moved beyond that now.

So how long do you think we can support an ever decreasing number of transactions? And at what number of transactions do you think we will have full equilibrium? 1? 10?

In this scenario, everyday transactions just move to sidechains and bitcoin becomes simply a unit of account for sidechains with no value outside it. Fees become a settlement channel between sidechain operators and the miners directly.

https://bitcoin.org/bitcoin.pdf

I agree with Nick Szabo when he stated that the block size debate needs more computer science and less noise. I want actual numbers and tests, not opinions.

THIS.

And so far the tests show 0.4 when the limit is at 1.0, and getting close to that limit will not mean failure or breakage, but just queuing

---> therefore delaying the dangerous and risky upgrade makes sense

most people cant anticipate things. they wait and fail.

we need bigger blocks to be ready for the next millions of users.

your statement would only be correct if blocksize was a bottleneck, and increasing it would lead to more usage/adoption

it is not

regulation and adoption are the bottlenecks, and therefore we could safely wait until we get closer to the limit...

it is like a dangerous surgery, that can be postponed without harm  ;D

ftfy

That's possible, but Fractalcoin just got mined out and the price shot up.  We'll see if people think it's worth mining for just transaction fees as time goes on, it'll be a good look into the far future of Bitcoin.  Theoretically, the lower rewards will still be worth it if the price goes up enough.

That list is stupid

block increase is needed and necessary in bitcoin enviroment

You realize that you will destroy bitcoin when you try to make up block halving losses with higher fees? I wonder how you guys think that could somehow work.

Besides that... you know that we have way way too many miners? Mining is way too rewarding at the moment, the profits are way too high. We need less miners, not higher fees.

At the moment we have a so high hashrate that we could handle 100 times more transactions and the network would still be safe. So we have 100 times more hashrate then we need.

And after this relatively new study, each bitcoin transaction is using the electricy that 1.75 average us households use in one day. For ONE single bitcoin transaction.

No, really, we don't need higher fees.


Though still we will hit it. And the plans for 8MB are longterm. It will happen in months.

By the way... with 1 MB we have no space for a sudden bitcoin adoption rise. If big players accept bitcoins and the transactions rise up then bitcoin would brutally fail. And it would get a hit that would be really hefty. Very bad user experience. Deadly.


Only because you live in a rich country doesn't mean that people with less money than you do "junk" transactions.

And when you take this important feature of bitcoin, what will you have then? Another paypal? Great. Why should someone switch to use bitcoin again then?


See again my answer to point 1. There is simply no need for rising fees now. And "simply rising" fees are a problem. Might be it is unimportant for you but free and cheap transactions are a key plus feature of bitcoin.


Then why again do we need 100 times more mining power and so costly transactions, power wise, that we look like a danger to ecology?


Yes, the confusion is there. FUD spread from all sides, way too emotional. Though that is something that happens on both sides.

I wonder if there will be a problem. I think the fear of the fork is already the worst thing. The actual thing will go through without problems i think.


I felt free too add. :)

Yeah... it somehow reminds of him wanting that someone decides what transactions are valid and which not. Illegitime transactions, dust and so on are unworthy to be part of bitcoin. They should be done "off chain". Well, that is the very definition of centralization, insecurity and totally not what bitcoin was created for. ::)

No one wants Hearns crazy ideas. Not even the one that chose xt temporarely.

And why the fear of miners working for free? You realize that mining is so rewarding nowadays that we have 100 times more hashrate than needed to make the whole bitcoin network secure? Bitcoin mining is way too rewarding at the moment. You really should not fear that miners drop their work. Miners NEED to drop their work. It does not make sense that it costs 1.75 average us households daily power usage for one single bitcoin transaction.

Then why do you fear that ominous spam coming when turtle claimed we are not even half full with our block? Shouldn't we have full blocks already when your spam fear would be true?

The blocks are not full with spam because it is costly. And it will be even more costly filling 8 MB blocks. No one would pay that because there would be no reason to do so.

I always wonder where suddenly all the spam should come from that does not even now happen. And the spam attacks that happen now only make sense because there is so few space in the block anymore. Otherwise the cost for spamming would be way way higher. And no one would pay for it.

Well I mostly disgaree with the OP for reasons others have already stated.

But good job for at least presenting some arguments of the other side of
the debate (I'm for bigger blocks).  At least you didn't list "decentralization"
which is a highly dubious argument when one is also advocating to push transactions off
the main chain and into the hands of a trusted party.

The thing about the so called "dusty junk" -- how do you know whats junk
and whats legit?  Maybe I want to send 25 cents to a gambling site,
or maybe I want to tip someone 20 cents on Facebook for an awesome comment.
What's the difference? Who's to say?  Who's to judge?

Why even limit things at all unless we have to?

This is a great argument which clearly shows that some people don't care about bitcoin but instead about some altcoin that they claim is better suited. What does this have to do with bitcoin? Nothing anymore.

We need to make bitcoin better. And that only can be reached with making bitcoin fit for the future.

Nobody awaits bitcoin having as many transactions as visa suddenly. But we know that the bitcoin usage is slowly rising and me need to be prepared for that.

All this fear spread... it would be funny when it wouldn't be sad.

why not let more TX pre block so that a block can have a lot of small 5 cent or less fees TX pay for the network?


its werid everytime i go see https://blockchain.info/ i always see a few full blocks on the six latest blocks. out of the last 6 blocks the 4 are full blocks right now.


would be interesting to get your idea of an want constitute a non-legitimate transaction.


one very cool thing about bitcoin is it made possible miro transaction, 10cent fee might not seem like much, but it's 10% of a 1$ payment. so much for bitcoin TX being free.


only if a significant % of miners not only not upgrade but start to massively attack the new chain with a lot of hashing power is it dangerous. at this point you are in a increasingly small minority when you say you want to keep the 1MB limit.


no comment.

thats weird other day someone said a ~256MB block would allow for about the same amount of TX VISA processes pre day.

...

~300million TX pre day is VISA which = 3472TPS

we need ~300bytes pre TX, so that ~1MB pre second worth of TX so ~600MB blocks should do it.

not bad, maybe miners can "zip" the blocks when trying to send them to other nodes so that might cut down the actual bandwidth needed dramatically.

BTW  3472TPS @ 0.0001BTC fee pre TX is about 208BTC in fees pre block, enoght to keep minners happy?? I think so!  ;D

(emphasis mine)

Oft-cited, rarely questioned, and never satisfactorily argued.  This is a myth.

Miners decide what to include.  You're not making sense saying 'they can be sent for free'.  They can only be sent free if the miner allows it.  I've disussed some of the complexities in this thread.

https://bitcointalk.org/index.php?topic=1162684.msg12361309#msg12361309

increasing block size could actually destroy Bitcoin, rather than help it!

fact, for years bitcoin has been well below block limit and for years most poeple put a fee along with their TX...

fact, 1million TX per block with avg fee of about 1cent is a hell of reward every 10mins.

fact, 1000 TX per block with avg fee of about 1dollar is 10 times less of a reward.

conclusion, small block with high fees, will make bitcoin less useful and less secure in the future

 8)

That's obviously true, but its a completely different context than the one where a blocksize cap can help raise fees (that context is far in the future when subsidies are greatlly diminished)...So It is puzzling why you're talking about them in the same paragraph, almost the same thought, when they are unrelated.

its not obviously true, its obviously false.

back in my day we paid fees for 2 reasons

1) paying a tiny  fee to support the network felt good.
2) the bitcoin client was defaulted to pay this fee. altho easy to change the setting, it was simply not cool and you felt bad doing it.

blocks were never anywhere near full, yet poeple still paid fees.

further more, if miners saw that some % of poeple never payed the tiny fee it would only take 1 miner to refuse to include 0fee TX into his blocks and that would add a big insensitive for everyone to pay this tiny fee to be sure to not let there confirmation time be determined by luck, most likely miners would be smart and include only 10  0 fee TX pre block that they mine, so that free TX in bitcoin are always possible but always much slower to confirm then a paid fee TX.

Proof that this "fee market" BS spread by the block-stream 1MBters are spreading are false! ::
Fact, miners are profit driven.
Fact, majority of miners want to increase block size
Conclusion, the idea that block limit will somehow make miners more money is a funny idea.

I think you might be the one who is confused!

Your statement here will almost certainly be false when the block subsidies go away.  
Fees will be the only thing left to support security costs.
Commercial miners will need to charge fees to stay in business.

The only way it could be true is if all mining is done by volunteers and hobbyists
or perhaps by dual purpose ASICS (mining refrigerators, etc)

@Adam:

You seem a bit confused also.  

The argument that the small-blockers are making (the ones who know what they're talking about --
see Davout on Reddit), is that unless blockspace is made a scarce resource, the free market
price of fees will be too low to provide sufficient security.  The idea is similar to price fixing eggs
at the supermarkets so the farmers earn enough and continue farming eggs.  However, there's
a few issues I have with that argument, which I posted in the other thread:

Here is real data that shows that over the last five years, higher block sizes = more fees overall for miners.  

https://i.imgur.com/MgVxfPe.gif

there argument is invalid for alot of reasons, but here's one.

1million TX per block with avg fee of about 1cent is a hell of reward every 10mins.
1000 TX per block with avg fee of about 1dollar is 10 times less of a reward.

the best is to have as many paying TX's on the network as possible, in that case in the future TX fees going from 1 cent to 3 cents wont mean much the the users but mean 3X more BTC for the minners.

+1

<sarcasm>
I hope to one day see Bitcoin transactions be expensive as hell and still limited to 7TPS
</sarcasm>

It's funny: many small farmers are now locked out the egg market because they can't get a quota.  The quota system has actually caused centralization of commercial egg producers here in Canada.  Small farmers can't sell their eggs on the commercial markets even though some are more efficient than the large operations!

Well, yes thats a possible scenario.

We have no idea how many transactions there will be decades from now,
which makes their line of argument pretty pointless, because we just
don't have enough data to formulate a complete scaling strategy (at least
as far as the fees go) this far ahead of time.  

Is it possible the free market revenue of fees with no block rewards
will provide low security?  Yes, it is.  But how likely it is, is anyone's guess because no
one really knows how adoption is going to unfold.  Therefore,
it would be irrational to limit the blocksize based on that,
at the current time.

https://static.fjcdn.com/comments/Aare+you+retarded+or+were+you+hit+in+_b7e423c9276e2b2b2c14975c6578b1c7.jpg
we can try to use commen sence to go in the right diretions

some how limiting bitcoin to <7TPS and stating that this is the only way miners will collect enough fees in the future seems wrong.

Actually the break from linearity is on the wrong side to support your theory.
(Should be up above the line, not below it).

Anyway, can you explain why you keep harping about the fees when
it will be several block halvings until fees even coming close the block
rewards, and therefore will be a minority of the security cost for
years or decades?

Maybe it's you who is totally misunderstanding the issue or just trolling. (I'm beginning to think of latter)

Block size has changed as you can see from the graph. The block size limit hasn't changed. So far the gap between actual block size and the limit has been so large that it has had no effect on economy whatsoever. (aside from recent "stress test".)


That makes no sense. Total fees have increased so far with increasing block size (and transfer rate). There is no reason to think that total fees would suddenly drop while transaction volume keeps growing. History shows us that exactly opposite is true.


More apt comparison would be that if more oil is transported the transport companies will make more profit in total, even if growing market would drive unit transportation fees down.

If you honestly think that limiting block size will profit miners I have a great plan for you. Lets set block size limit to 0 and let miners enjoy huge profits from zero transaction rate. yay logic! :D

the chart shows that total fees ( not the size of each fee ) per block goes up, because there's more TXs
1,000 TX paid 1 dollar fee each
or
100,000 TX paid 1 cent each
the end result is the same for the miners
case #1 bitcoin has a limited TPS and fees are high
case #2 bitcoin  has unlimited TPS and fees are low
how in the world can you argue case #1 is better????

What are you talking about?  Block size has unquestionably changed (increased) over the last 5 years.  The increase in the block size--such that it's now approaching the anti-spam limit--is what has precipitated this debate. 

Peter your chart is misguiding obviously as it shows an increase in transactions, not block size, all happening under the same limit.

The increase in fees paid is a result of more transactions, not bigger blocks.

You do know that blocks are made of transactions? :) So generally, more transactions means exactly that, bigger blocks.

https://dl.dropboxusercontent.com/u/43331625/feemarket.pdf

now that #1 argument is deemed invalid lets talk about #2


we are hitting the 1MB limit on the majority of block right now.

the proof: https://blockchain.info/ look at the block size on the last 6 blocks mined

the average block size seems low because sometimes a miner puts out a block with like 10KB and pools seem to be limiting themselves to 900ish KB not 1024KB

blocks have about doubled in last twelve months.  could double again in twelve months so if that's "a long time" then he's right,  but it's not, and he isn't.

And how do you get more transactions without fitting them into larger blocks? There is a straightforward linkage between the two in the long run.

Or they can double in the next 3 months if we get some crazy influx of new users. Let's say because of some amazing news, or some government that would recognize us, or just 10 million Chinese that would see a Bitcoin documentary and get interested.

The thing is that future is unpredictable and we never know when and why this might happen. And what we do then if network would get crippled?

There's nothing "misguiding" about that chart.  It clearly demonstrates that more transactions equals more fees collected.  That's why we should alter the blocksize limit to allow more transactions.  If we don't raise the cap, the only way for revenues to continue to rise is to raise the fee itself.  But if that gets prohibitively expensive for the average user, they'll simply transact on another chain and miners will get less revenue than they would with a larger blocksize limit.  Forcing an unnatural fee market will not help scalability and will deter new users.  A cautious, responsible and preferably (IMO) flexible increase that keeps centralisation to an absolute minimum is the best way forward.  The only people still denying that are the ones who envision Bitcoin as an exclusive little club for the wealthy and frankly I can't wait to leave all those people behind.  

Of course more transactions = more fees collected. This does not logically entail anything that you are saying. Mere capacity is not the only concern. And reducing the complexity of a fee market to being "unnatural" and "will not help scalability," or assuming that it is necessarily a deterrent to adoption -- I don't buy it. I'm not convinced by your sectarian perspective, either.

The wealthy will continue to rule Bitcoin and there is no amount of "new users" tears that will change this.

The average user' seat in the block, if he chooses not to claim it, will be taken by anyone willing to pay the cost. It is quite simple: the capital demand for censorship-resistance, at whatever price, is above and beyond any demand for cheap transactions.

Simply said: the movement of capital around the world is not handicapped by cost but by political friction.

The world will not be made wealthier from cheaper transactions but from breaking the shackles of governments power over money.

Once you can understand this aspect of the economy maybe you will come to peace with reality.

Where did I say it was the "only" concern?  I clearly said: "A cautious, responsible and preferably (IMO) flexible increase that keeps centralisation to an absolute minimum".  I've taken on board the considerations of maintaining a sufficiently decentralised system.  I recognise the importance of that, but there is still a balance to be struck against capacity.  I've made -numerous- attempts to be reasonable and discuss solutions that would strike a balance that doesn't result in any reckless increases in blocksize, but it's like banging my head against a wall because certain people won't budge.  They genuinely believe that capacity isn't an issue and the only concern is centralisation.  I'm tired of it.  So if I'm displaying a "sectarian perspective", that's where it's come from.  Too many MP fanboys spouting too much MP guff.

And to further clarify, I'm not ruling out fee markets altogether, I just don't think we should hit a solid wall and ramp the fees up too quickly as a result.  Yet another reason why I've been pushing for a dynamic limit like BIP106 (but with alterations to the doubling/halving bit as that seems excessive).  Something like this (https://bitcointalk.org/index.php?topic=1154536.msg12312664#msg12312664) seems perfectly reasonable, or perhaps my (even more conservative) adjustment described here (https://bitcointalk.org/index.php?topic=1144606.msg12342679#msg12342679). 

How about we eliminate centralization risks forever and still serve your demand for capacity?

https://www.mail-archive.com/bitcoin-development@lists.sourceforge.net/msg07937.html

+1
there are a lot more poeple that think like you do then you realize, i think what you say is more or less what the majority think. the key is finding a good balance for capacity Vs centralisation. there is simply a very loud minority unwilling to budge from there senseless position, at this point i'm calling them trolls. Arguing that 1MB is the absolute limit before block size starts to increase centralisation is just plain incorrect.

Do you have issues with reading comprehension, turtlehurricane?  Try slowly reading it again one more time:


That's what uxgpf said.  The LIMIT has not changed.  But the size of individual blocks have been steadily increasing with more transactions taking place.  As the number of transactions increase, the total amount generated in fees continues to increase with them.  -If- the number of transactions can no longer increase because blocks have reached the 1MB limit, the only way the total amount generated in fees can increase is if fees on individual transactions go up (and there's no guarantee this would generate the same income as simply allowing more transactions through).  It's really not that complicated.  We shouldn't prevent a natural increase in transactions (unless it becomes excessive and threatens decentralisation) because the miners receive greater rewards from supporting more transactions.

There is no "spam" there are no "junk" transactions. There are only "unprofitable" transactions. The more I look at all these "issues", the more it seems that miners are the problem and we need to find another alternative to secure the block chain instead of relying on naked greed.

Sounds like you don't really want a proof-of-work system at all! That's cool, and lots of altcoins have explored alternative incentive regimes. However, this would fundamentally change bitcoin and is unlikely to be supported by those who support the principles it was built upon.

There is indeed spam -- for instance, what is the use of pushing outputs that require significantly more fees than the output is worth in order to re-spend them? It's pure spam.

adamstgBit: People are redeeming outputs from the Coinwallet dust keys:
https://bitcointalk.org/index.php?topic=1175321.msg12373932#msg12373932

Each output = ~75,000 bytes.

I wonder if they will actually release hundreds more keys in the coming days. Hilarity may just ensue. :D

Each output = ~75,000 bytes.

that must mean a nice fee for the miners?

I guess that depends on your definition of nice. :P

I suppose a 0.00447 fee to include a 0.00053 output is pretty nice from a percentage standpoint. But I don't think anyone is getting rich off of these 2.5BTC keys. :D

Fees as a "spam" disincentive isn't working very well then, is it?

If you need to send someone Sataoshis, the recipient doesn't care about the fees-they still demand the Satoshis.

At some point the miners will push for not including transactions with zero or low fees at all, then Bitcoin will be another mainstream banking product.

How so? Most people default to the lowest fee that will secure their transaction on the next block. Do all economic incentives work 100% as in theory? Of course not. Game theory prevails. Those "investing" in spam attacks may indeed have incentives that outweigh the disincentive of fees. I'm not sure where that leaves us, though.

I don't think the former entails the latter at all. Decentralized =/= another mainstream banking product. That's completely impossible. Whether or not the former is an acceptable outcome is debatable, but the association with "mainstream banking" is fallacious.