Gavin Andresen Proposes Bitcoin Hard Fork to Address Network Scalability

[LeMiner]

A lot of the motivation for the 1MB limit was a fear that if it weren't limited we'd have people building gigantic blocks just to clog up the system.  If someone with major hashing power started building giant blocks every time he got a block, he  could DoS the system just by making the block chain too large for most people to store.  

How exactly would this be done? Where are these transactions coming from? Could this "bloating" attack be detected and these blocks rejected?

It would be better to prevent this some other way and remove the limit entirely.

This has already been addressed by Gavins original post. Maximum Block size would for example grow 50% per year to allow for more transactions in each block, per year, while still keeping storage costs relatively low. A totally unrestricted block size has never been spoken about.

[1714025692]

1714025692

[1714025692]

1714025692

[1714025692]

1714025692

[1714025692]

1714025692

[1714025692]

1714025692

[asdf]

A lot of the motivation for the 1MB limit was a fear that if it weren't limited we'd have people building gigantic blocks just to clog up the system.  If someone with major hashing power started building giant blocks every time he got a block, he  could DoS the system just by making the block chain too large for most people to store.  

How exactly would this be done? Where are these transactions coming from? Could this "bloating" attack be detected and these blocks rejected?

It would be better to prevent this some other way and remove the limit entirely.

This has already been addressed by Gavins original post. Maximum Block size would for example grow 50% per year to allow for more transactions in each block, per year, while still keeping storage costs relatively low. A totally unrestricted block size has never been spoken about.

I meant how would the attack be done...

[Q7]

I just worry about the negative impact when the hard fork is taking place especially put into consideration the size of network we are having right now. I mean what if somebody plans a out scheme that manipulates the network during the switch over. But then if we think again, if we don't do it now, then the question is when are we going to do it. As the network grows there can never be the right time.... and we really need to plan, prepare and anticipate for the future.

[Cryddit]

A lot of the motivation for the 1MB limit was a fear that if it weren't limited we'd have people building gigantic blocks just to clog up the system.  If someone with major hashing power started building giant blocks every time he got a block, he  could DoS the system just by making the block chain too large for most people to store. 

How exactly would this be done? Where are these transactions coming from? Could this "bloating" attack be detected and these blocks rejected?

It would be better to prevent this some other way and remove the limit entirely.

It's a simple matter of having the miner send one (or more) of his unspent txouts to a new txout -- which he also controls, of course.  He can even put in a generous tx fee -- because he's the miner and he'll collect it.  They won't cost him anything even if another miner gets the block.  The other  miner hasn't heard about them yet (because the guy bloating blocks didn't broadcast them), hence the attacker won't even owe any tx fees. 

So, yes, in principle a miner can bloat a block with as many bogus transactions as he wants, it won't cost him a darn thing to do so beyond the possibility that his huge block broadcasts more slowly and gets orphaned, and there's really not much anybody can do to stop him. 

[BitMos]

...

K.I.S.S. = Keep It Stupidly Simple, leads to the Moon and beyond...

[QuestionAuthority]

"Only invest what you can afford to lose" applies just as well to any complete technology as it does to something that's still in beta.  

Remember, Bitcoin is the initial implementation of a protocol.  Even if Bitcoin is a mature technology and no longer experimental, you should still invest only what you can afford to lose, because some other implementation of the protocol (yes, an altcoin) may eventually be the one that businesses settle on as being friendlier to their interests.  

I think investing is different from daily use. If the only users were speculators I wouldn't have an issue with them losing everything. If you invest in junk bonds and lose it all it's your fault. The altcoins don't have much of an economy. Those are really investments (so to speak) and their users are out for a quick return. Some of the altcoins are trying new things and might have the ability to overtake Bitcoin while being more stable and safe. In that case, I hope they do leave Bitcoin. I seriously doubt big business, at this point, will jump ship and attach to an altcoin with no name recognition and no established economy.

I know it's impossible to see many problems before they happen. Some issues can be seen ahead of time and corrected. For the most part the dev team has corrected the issues quickly. Sometimes crooks can use real systemic issues as an excuse to steal as Karpeles did with transaction malleability. Is the system so well prepared now that we can scale up and invite even more use and possibly more loss until there's no one left on the earth that's willing to trust Bitcoin?

It's an easy decision for businesses to accept another payment method. It's especially easy when companies like Coinbase are offering the first million dollars of exchanges to fiat for free. Users are a different story. Users need to trust the system and benefit from it. I know low fees are a benefit and this will help. Should the system be throttled limiting the amount of new entry until it's ready for prime time, is it ready now or is there no way to know until it's ramped up? Ever increasing use means ever increasing losses when failure happens. I would hate to be the lead dev knowing I had this huge experimental market sitting on my shoulders. I'm really not trying to be a prick. I'm just giving you food for thought. Mr. Andresen likes to reach out to the community for consensus and that's a good thing. Maybe the word community to him means only the dev team. In that case, he should really just send out an email instead.

[BitMos]

by reading this post I realized that most of you are against competing currencies... if you could kill the king$ you would, just because you have high stakes in btc... But the great wise and smart speculators will all know your little plan before you are even born. Why? they are so basic, make btc the only global currency, and you, first time hoarders the new master of the Earth... you really believe it would be that simple...   Long Live the Alt. Btc must stay unforked. the speculators, again will find its right place... to extract the maximum value of it. by learning to code you made a trade off. You can't escape the consequences of this choice.

Edit: forking would be the greatest thing to happen to the alts since... well... bitcoin.

[iluvpie60]

"Only invest what you can afford to lose" applies just as well to any complete technology as it does to something that's still in beta.  

Remember, Bitcoin is the initial implementation of a protocol.  Even if Bitcoin is a mature technology and no longer experimental, you should still invest only what you can afford to lose, because some other implementation of the protocol (yes, an altcoin) may eventually be the one that businesses settle on as being friendlier to their interests.  

I think investing is different from daily use. If the only users were speculators I wouldn't have an issue with them losing everything. If you invest in junk bonds and lose it all it's your fault. The altcoins don't have much of an economy. Those are really investments (so to speak) and their users are out for a quick return. Some of the altcoins are trying new things and might have the ability to overtake Bitcoin while being more stable and safe. In that case, I hope they do leave Bitcoin. I seriously doubt big business, at this point, will jump ship and attach to an altcoin with no name recognition and no established economy.

I know it's impossible to see many problems before they happen. Some issues can be seen ahead of time and corrected. For the most part the dev team has corrected the issues quickly. Sometimes crooks can use real systemic issues as an excuse to steal as Karpeles did with transaction malleability. Is the system so well prepared now that we can scale up and invite even more use and possibly more loss until there's no one left on the earth that's willing to trust Bitcoin?

It's an easy decision for businesses to accept another payment method. It's especially easy when companies like Coinbase are offering the first million dollars of exchanges to fiat for free. Users are a different story. Users need to trust the system and benefit from it. I know low fees are a benefit and this will help. Should the system be throttled limiting the amount of new entry until it's ready for prime time, is it ready now or is there no way to know until it's ramped up? Ever increasing use means ever increasing losses when failure happens. I would hate to be the lead dev knowing I had this huge experimental market sitting on my shoulders. I'm really not trying to be a prick. I'm just giving you food for thought. Mr. Andresen likes to reach out to the community for consensus and that's a good thing. Maybe the word community to him means only the dev team. In that case, he should really just send out an email instead.

If this thread needed more conjecture I would provide it(and in a more amusing way).

I have seen your posts and you do troll. I could go back to a thread from something and post your quote where you admit that you troll for the fun of it. You treated me badly before in a thread from the past and it is whatever honestly.


Bitcoin will need to adapt and change. I fully agree with Gavin after reading more into his posts and more on the bitcoinfoundation site. It is easy for everyone here to criticize something when they aren't even in charge of more than a few hundred K per year.

I personally Forecast/Plan for 2 production facilities and the yearly total of production is $80,000,000 to $150,000,000(just giving a range real number is in there). Supply and Demand is KING for what I do. As a Forecaster I can see Bitcoin needs this change because all signs point to there being a higher demand for transactions sometime soon, therefore supply of transactions will increase by X% as demand grows. On the supply slide of things you can then infer that transactions will grow larger in size and eventually hit a point where it is bigger than 1 MB blocks, meaning fees will need to be higher. It is not intended for fees to be so high early on, that comes later. It is a simple concept and if it is wrong there isn't exactly any consequences as it can be changed to something else that would better suit it.

Bitcoin Protocol is largely correct in many areas, other areas get improved, and the rest of the slack will be taken up by something else or another coin. Bitcoin cannot be everything any anything at the same time. Focusing on what Bitcoin does best is what Gavin is talking about for this hard fork. The main principle that wider adoption will come from ensuring fees stay low is key here. Especially since the fees are a direct result of reward + transaction volume(size and amount). So if there are too many transactions going through that it goes over 1MB consistently only the higher value fees will be taken which makes fees go up in price.

[BitMos]

meaning fees will need to be higher. It is not intended for fees to be so high early on, that comes later.
... So if there are too many transactions going through that it goes over 1MB consistently only the higher value fees will be taken which makes fees go up in price.

tada the protocol already found the answer itself... no need for fork. Who said the most precious numbers were cheap to transact? use an alt for your tipping.

edit: you can put a k for thousand and a M for millions... but I don't think you need to learn the next letters... even in your fantasies...

[Lauda]

When did this turn into a discussion about TBF, Gavins credibility and such?
This is about Network Scalability. Gavin is right, and you know that he is.

[lunarboy]

response from Gavins AMA today

Q: (NedRadnad) - Do we really need to hard fork the chain to achieve scalability? When do you plan on making the fork?

  gavinandresen  A: Yes, I think we do. There is still at least a month or two of work before I'd be willing to write a patch to increase the maximum block size, and then probably a month or two more of arguing. So, early next year at the earliest before even starting the hard-fork process (which must roll out to miners – they will control when the fork actually happens).

estimates on timeframe suggest Q2 next year at the earliest.

[wolfYella]

I just worry about the negative impact when the hard fork is taking place especially put into consideration the size of network we are having right now. I mean what if somebody plans a out scheme that manipulates the network during the switch over. But then if we think again, if we don't do it now, then the question is when are we going to do it. As the network grows there can never be the right time.... and we really need to plan, prepare and anticipate for the future.

I don't think a planned fork would make the network any more vulnerable to a potential attack then a time when the network is not about to get hard-forked. I would even argue that the network would be slightly more secure during this time as many people would be paying very close attention to the network and various aspects about the network

[1echo]

i like this hardfork idea

[fa]

If a hard fork is inevitable, lets just embrace it. We should give BTC protocol a chance to improve itself. Just like we have Html 5 and CSS 3 after Html 4 and CSS2.

[tvbcof]

Probably THE most critical thing would be predictability, and the set-in-stone increase in transaction rate achieves that.  Without predictability it would be much less easy to justify investments in infrastructure among other forms of planning and it would likely set the solution back a fair bit.  So, hats off to Gavin on this thing at least.  (No shit.)

I still don't understand why, if the goal is to support the world's beings with a nearly free (in monetary terms) solution to their economic exchange needs, the expansion rate is so low.  It makes me suspicious that there is something else going on...but then I'm a naturally suspicious guy.  I'd take a long hard look at the deltas which make 1.0, especially with the likes of Mike hanging around.  Maybe it's just a matter of getting people used to hard forks the solution can be more easily modified at will (or on demand) going forward.

[deluxeCITY]

Probably THE most critical thing would be predictability, and the set-in-stone increase in transaction rate achieves that.  Without predictability it would be much less easy to justify investments in infrastructure among other forms of planning and it would likely set the solution back a fair bit.  So, hats off to Gavin on this thing at least.  (No shit.)

I would strongly agree with this. It is unquestionable that the block size will need to be increased in order for Bitcoin to facilitate the number of transactions it would likely see when it becomes successful. If the protocol is forked multiple times then uncertainty would be caused

[BitMos]

if they fork once they will fork twice... 

[The00Dustin]

I have seen your posts and you do troll. I could go back to a thread from something and post your quote where you admit that you troll for the fun of it. You treated me badly before in a thread from the past and it is whatever honestly.

No, that was me, and I still say you were trolling first in that thread.  However, in this thread, I agree with you.

if they fork once they will fork twice... 

If you don't like that, then you need to get a time machine and go back in time to stop them from forking the first time.

[BitMos]

I have seen your posts and you do troll. I could go back to a thread from something and post your quote where you admit that you troll for the fun of it. You treated me badly before in a thread from the past and it is whatever honestly.

No, that was me, and I still say you were trolling first in that thread.  However, in this thread, I agree with you.

if they fork once they will fork twice... 

If you don't like that, then you need to get a time machine and go back in time to stop them from forking the first time.

It's funny how none of you want to accept the existence of alts, and how they could be integrated in the p2p payment system... but nooo, you want your precious bitcoin to take over everything, including alt... interesting.

[bitcoinpiece]

if they fork once they will fork twice... 

You cannot change that if forking is what they do.