Blocksize

[Cconvert2G36]

4MB max in 2016, doubling at the halvings. (2MB would/could/should have been the limit at the drop to 25btc block rewards.)

Growth in the max could be halted or slowed by soft fork if alternative scaling solutions present themselves and are proven to be both functional and desired by the market.  

Code:

Max Blocksize	Block Reward	Year(~)		Native TPS(~)
1MB		50		2009		3
2MB		25		2013		6	
4MB		12.5		2016		12
8MB 		6.25		2019		24
16MB		3.125		2022		48
32MB		1.5625		2026		96
64MB		0.78125		2029		192
128MB		0.390625	2033		384
256MB		0.1953125	2037		768
512MB		0.09765625	2041		1536

There's something nice about the available space in a block for fee paying txs doubling at the same time the reward is cut.

[1713894136]

1713894136

[1713894136]

1713894136

[1713894136]

1713894136

[1713894136]

1713894136

[1713894136]

1713894136

[adamstgBit]

IMO they need to figure out some limit which reflects the limits of technology ( today's avg internet speeds ) and just hardcore the limit to that, that's it.

[2112]

My favorite one - the solution that was suggested at the conference: introduce the second type of blocks that contain transactions only and issued 60 times more often, every 10 seconds or so. Those blocks are not for determining winner of reward, so they won't cause orphan problems. This solution will give us 60 times more space for transactions and reduce confirmation time 60-fold!  I think it is the perfect solution, even better than lightning, since it's not hub-based.

This is the solution that came out of people associated with p2ppool: fold p2ppool protocol into core protocol as a proof-of-propagation through proof-of-work. It is at over a year old, I've discussed it on this board in October 2014:
 
Re: Increasing the block size is a good idea; 50%/year is probably too aggressive 
https://bitcointalk.org/index.php?topic=815712.msg9245823#msg9245823

and I don't recall if I haven't seen it elsewhere before.

But then as usual within the Bitcoin milieu: it is more interesting to try to follow who was opposing that type of proposal and with what kind of argumentation; e.g.

a) is vulnerable to sybil attacks
b) smothers the incentive to include any transactions in blocks: why should I (as a miner) include a tx if the fee would go to someone else?

Also it seems both  are too disruptive  to be implemented in bitcoin.
Anything this much different would take an altcoin to be tried.

[Cconvert2G36]

1 MB forever

A loner, Dottie. A rebel.

 

[Blue_Tiger73]

I'm not 100% convinced that it is necessary.. I just hope they don't go and screw it up..

Then vote stay the same.

I think a move to 8MB and hold is a safe move.

Bitcoin is going fine as is. Do we really want to change this greatness not knowing what is going to happen? The 1MB block size was made to stop spam and we should keep it that way. Bitcoin doesn't need to change so why are we arguing about this.

[AndySt]

Thoughtless increase of the block size leads to more monopolization of mining. So I'm more inclined to option 2 megabytes. This is the best option at the moment.

[spazzdla]

Perhaps 8MB is too much then... hum..  2MB just seems like such a small step. 

[Cubic Earth]

4 or 8 to start.  I submitted a proposal to the conference (which was rejected), 4-8-16-32, doubling every year.  If it started at 8, it could double every two years.

I'm a firm believer that in the short term, it would be really stupid for us to cap user growth.  It would be stupid for us to cap user growth in any time frame, but there are other scaling solutions that will work or are highly likely to work, they just happen to be one, two or three years away from being implemented.  That is the gap we need to bridge.

I am completely comfortable capping at 32MB for now (for the foreseeable future), and to understand why, lets look at some rough numbers.

32-MB would allow for 60-Million tx per week, and 3-Billion tx per year.  I see the Lightning Network as a certainty, but one that is understandably taking time to implement.  Lightning will have a multiplier effect, so that each on-chain transaction will be able to back some greater-number of lightning transactions.  What is unclear is what this multiple will be, and the multiple itself will like grow with time as the lightning network grows and matures.  It could be 10-1 or 100-1 within a few years of being implemented.  So in a sense, the total bitcoin transaction rate could be as high as 1,000 tx/sec or 10,000 tx/sec using those two multiples as examples.  That is based off the idea of a 32-MB block yielding 100 on-chain tx/sec, multiplied by 10 and 100.

Even though the number of tx/sec could essentially be unlimited with lightning, the number of USERS who have access to the blockchain would effectively still be finite.  Lets imagine Lightning taking on the role of today's credit cards.  Users would need to settle on occasion, at least monthly.  32-MB blocks would allow roughly 200-Million people to each make a single monthly transaction.  Considering the fee market under such conditions, the idea of paying $1 for a monthly settlement transaction seems reasonable, though perhaps on the expensive side, and $0.33 per tx shouldn't dissuade any phone / computer user from participating.  Those two fee rates would generate $3-billion and $1-billion per year in total fees, respectively, which will go a long way towards paying for network security.

Bitcoin-NG.  I've read the paper (and understand 95% of it).  I think it will work, and it solves a few of the trickiest scaling problems. The Bitcoin-NG system eliminates miner-bandwidth differentials as a source of centralizing pressure.  Single-packet latency sill matters, as it does currently, but such is unaffected by block size.  With NG, it becomes reasonable to use a significant fraction of a user or miners total available bandwidth and processing power to process blocks.  A 20% capacity on a 20-Mbit/s connection can yield 30-MB / minute, which would give the equivalent of 300-MB blocks under the current system.  Such would have a native, on-chain capacity of 1,000 tx/sec, for 600-million monthly users, who should each be able to perform 10's to 100's of Lightning transactions per month.

And Sidechains... they hold great promise as well, but they are not as well defined as Lightning or NG.

Hopefully we can all see that 32-MB blocks can get us very far when combined with Lightning, so it should be safe for Bitcoin as medium or long-term cap  Bitcoin-NG will allow Bitcoin blocks to effectively be some multiple larger than they otherwise would, say 5x to 10x.  In the short term, we must expand capacity to grow the system.  Choking off user growth is the biggest security risk to the network.  Big blocks have their own issues, but 4 or 8 MB would not materially change the centralization pressures today, and 32-MB will not either, say 4 or 6 years out, as electricity costs continue to dominate the equation.

[cohnhead]

4 or 8 to start.  I submitted a proposal to the conference (which was rejected), 4-8-16-32, doubling every year.  If it started at 8, it could double every two years.

I'm a firm believer that in the short term, it would be really stupid for us to cap user growth.  It would be stupid for us to cap user growth in any time frame, but there are other scaling solutions that will work or are highly likely to work, but they are one, two or three years away from being implemented.  That is the gap we need to bridge.

I am completely comfortable capping at 32MB for now (for the foreseeable future), and to understand why, lets look at some rough numbers.

32-MB would allow for 60-Million tx per week, and 3-Billion tx per year.  I see the Lightning Network as a certainty, but one that is understandably taking time to implement.  Lightning will have a multiplier effect, so that each on-chain transaction will be able to back some greater-number of lightning transactions.  What is unclear is what this multiple will be, and the multiple itself will like grow with time as the lightning network grows and matures.  It could be 10-1 or 100-1 within a few years of being implemented.  So in a sense, the total bitcoin transaction rate could be as high as 1,000 tx/sec or 10,000 tx/sec using those two multiples as examples.  That is based off the idea of a 32-MB block yielding 100 on-chain tx/sec, multiplied by 10 and 100.

Even though the number of tx/sec could essentially be unlimited with lightning, the number of USERS who have access to the blockchain would effectively still be finite.  Lets imagine Lightning taking on the role of today's credit cards.  Users would need to settle on occasion, at least monthly.  32-MB blocks would allow roughly 200-Million people to each make a single monthly transaction.  Considering the fee market under such conditions, the idea of paying $1 for a monthly settlement transaction seems reasonable, though perhaps on the expensive side, and $0.33 per tx shouldn't dissuade any phone / computer user from participating.  Those two fee rates would generate $3-billion and $1-billion per year in total fees, respectively, which will go a long way towards paying for network security.

Bitcoin-NG.  I've read the paper (and understand 95% of it).  I think it will work, and it solves a few of the trickiest scaling problems. The Bitcoin-NG system eliminates miner-bandwidth differentials as a source of centralizing pressure.  Single-packet latency sill matters, as it does currently, but such is unaffected by block size.  With NG, it becomes reasonable to use a significant fraction of a user or miners total available bandwidth and processing power to process blocks.  A 20% capacity on a 20-Mbit/s connection can yield 30-MB / minute, which would give the equivalent of 300-MB blocks under the current system.  Such would have a native, on-chain capacity of 1,000 tx/sec, for 600-million monthly users, who should each be able to perform 10's to 100's of Lightning transactions per month.

And Sidechains... they hold great promise as well, but they are not as well defined as Lightning or NG.

Hopefully we can all see that 32-MB blocks can get us very far when combined with Lightning, so it should be safe for Bitcoin as medium or long-term cap  Bitcoin-NG will allow Bitcoin blocks to effectively be some multiple larger than they otherwise would, say 5x to 10x.  In the short term, we must expand capacity to grow the system.  Choking off user growth is the biggest security risk to the network.  Big blocks have their own issues, but 4 or 8 MB would not materially change the centralization pressures today, and 32-MB will not either, say 4 or 6 years out, as electricity costs continue to dominate the equation.

you seem to have a handle on this....what happens when there is a paradigm shift in electrical generation....granted not 4-6 years out ..but maybe in 20 yrs or so when nuclear fusion could be a reality.

[Cubic Earth]

you seem to have a handle on this....what happens when there is a paradigm shift in electrical generation....granted not 4-6 years out ..but maybe in 20 yrs or so when nuclear fusion could be a reality.

There are too many variables at play to know what 20-years out looks like, and dozens of different hypotheses would probably all be just as likely.  It's really important to keep in mind that if bitcoin is still around in 6 to 10 years from now, the user base will be much, much bigger.  There will probably be a thousand core-devs.  Governance of protocol and code bases will probably be handled very differently than they are now.  With different stakeholders and different methods, Bitcoin will most likely move in directions that today do not seem possible.  We need to keep the torch going so that we may pass it on, and with all luck, things will head in a direction that would make us all - or most of us at least - proud.

Bitcoin is already on it's second generation of users and custodians.  It is not the same group that was using it in 2010-2012.  Sure, many of us are still here, and with big roles in today's community, but we have had to welcome, or grudgingly accept, the inclusion of banks, V.C. firms, governmental rulings, and large bitcoin companies in the process.  There is also far more research and thinking going on, so the technology keeps getting better too.

[tss]

everyone's opinion is still the same.

bitcoin will fail if we increase blocksize
and
bitcoin will fail if we dont increase blocksize.  

i can't say i am glad to see this post.  blocksize debate is getting tired.

You're the one who's wrong and who isn't reading anything. If we increase with the right proposal Bitcoin will not fail. If we do not do anything Bitcoin will also not fail. Bitcoin has already succeeded.

i am merely describing the sentiment of the agenda and fan boys (either side).

"If we do not do anything Bitcoin will also not fail. Bitcoin has already succeeded. "

snip
I'm a firm believer that in the short term, it would be really stupid for us to cap user growth.  
snip

in order to cap user growth as you say it first has to be at a maximum.  which it is nowhere near.