Author

Topic: Gold collapsing. Bitcoin UP. - page 166. (Read 2032248 times)

legendary
Activity: 1764
Merit: 1002
June 30, 2015, 02:46:49 PM
I think tvbcof means to say Bitcoin wasn't ever going to be useful for point of sales. (I assume that means Bitcoin by itself.)

In the interest of not repeating old points, this is what Hearn has to say on this: https://medium.com/@octskyward/replace-by-fee-43edd9a1dd6d

Ah, Mr. Hearn.  Cypherdoc's giant among men.  I like Peter Todd's latest rendition which seems to be something like:  'The storm is coming but won't arrive until all non-hackers have had a fair chance to get on the lifeboats currently under construction.'



lol, your thick skull acts like i am the "only" one out there arguing this.  there is a whole legion of Bitcoin merchants, including Coinbase, out there who are making the exact same argument.  in fact, RBF is a non starter even for the core devs.  Todd is the only one pushing this.

lol.
legendary
Activity: 1162
Merit: 1007
June 30, 2015, 02:41:13 PM
I think tvbcof means to say Bitcoin wasn't ever going to be useful for point of sales. (I assume that means Bitcoin by itself.)

In the interest of not repeating old points, this is what Hearn has to say on this: https://medium.com/@octskyward/replace-by-fee-43edd9a1dd6d

Right.  I think the people pushing full-RBF are more interested in making on-chain Bitcoin transactions less useful for point-of-sales than the often-stated goal of allowing users to "unstick" stuck transactions.  Since we can "unstick" a TX in ways that are first-seen safe, why break a system (zero-confirm TXs) that many people currently enjoy (well unless you're of the ideology that Bitcoin should only be a settlement backbone)?    

...or unless your transaction can wait for some non-trivial amount of time to confirm to the level of reliability that is necessary for it.


I don't follow.  It sounds like you think it would be a good thing to make zero-confirm transactions less secure.  Is this true?
legendary
Activity: 4690
Merit: 1276
June 30, 2015, 02:40:03 PM

with RBF, does the sender have the ability to completely change the destination address back to himself?

What's your problem with that?  It's his money until it's not.

Oh, Bitcoin is not a real-time system in spite of years of propaganda trying to make it so?  How sad.  I think I'm going to cry.

Except that Bitcoin does function as a real-time system for most use cases today perfectly fine. There is an entire Bitcoin ecommerce industry that proves you wrong.

precisely.

Then why are you all fritzed out about the RBF 'attack'.  Real-time Bitcoin not so robust as you've been led to believe?

legendary
Activity: 1036
Merit: 1000
June 30, 2015, 02:39:42 PM
Quote
Because it can be economically expected that most transactions now are low priority, since fees are cheap. This means the actual transaction volume that likely would be taking place given only somewhat more expensive fees is much, much less. It's should be obvious enough that if transactions were completely free people would spam like crazy. That's not a judgment on the validity of those transactions; it's just economic reality. People start to economize dramatically when there is even a marginally-above-zero cost versus when there isn't.

This is in no way suggesting 1MB is sufficient, but it's not helpful to pretend that transaction volume when fees are negligible will remain the same if fees go to merely tiny. In no other market would we expect such a thing, and for all we know we could easily be talking about order-of-magnitude differences in volume from a fee increase that still presents no real pain for any major use case.

but cheap to who, you?  i remember an Indian standing up at the San Jose conf during Q&A and complaining about high tx fees for ppl in his country.  and they are, given their relative income levels.  

we should be promoting Bitcoin to these 3rd world countries where a .0001 minimum fee is arguably expensive.  that's why we have the  economic fee choice of .00001 which oftentimes still results in rejection.  to take it further, do you believe that Bitcoin should offer even these Indians the ability to buy a cup of coffee for cheap fees?  i think so but probably you think different.  i know tvbcof and iCE would flat out say no way.  but then that goes back to my argument that for maximum decentralization and to become digital gold Bitcoin needs to service these ppl.

when you say ppl will spam like crazy, don't you trust that the miners can react to filter that spam, if they so choose?  the problem with the 1MB choke approach is that yes, new users will be forced to economize on their tx's.  OR just leave.

This isn't relevant to the question of whether having a fee market would help. I'm not saying 1MB is adequate, and I'm not saying that fees shouldn't be cheap enough for any particular group. I'm saying that no matter what blocksize cap we have, or even if we have no cap, we will still need a fee market for maximum scalability on chain. And we could easily be talking about order-of-magnitude differences in scalability contingent on the presence or absence of a proper fee market. As for miners blocking spam, sure, but as you say: who's to tell what is really spam?
legendary
Activity: 4690
Merit: 1276
June 30, 2015, 02:36:30 PM
I think tvbcof means to say Bitcoin wasn't ever going to be useful for point of sales. (I assume that means Bitcoin by itself.)

In the interest of not repeating old points, this is what Hearn has to say on this: https://medium.com/@octskyward/replace-by-fee-43edd9a1dd6d

Right.  I think the people pushing full-RBF are more interested in making on-chain Bitcoin transactions less useful for point-of-sales than the often-stated goal of allowing users to "unstick" stuck transactions.  Since we can "unstick" a TX in ways that are first-seen safe, why break a system (zero-confirm TXs) that many people currently enjoy (well unless you're of the ideology that Bitcoin should only be a settlement backbone)?    

...or unless your transaction can wait for some non-trivial amount of time to confirm to the level of reliability that is necessary for it.

legendary
Activity: 1512
Merit: 1005
June 30, 2015, 02:33:03 PM
blocks full again with TPS unacceptably high --> unconf tx set ~ 3000, more than double normal.  pools taking defensive action.  when will Blockstream devs do something?:



Why do you call this "defensive" action?

when Chinese pools face a series of what they consider to be "large blocks", which in this case means full blocks as you can see from the data, they automatically switch to mining "header" blocks with 0 tx's during the time it takes to process all the tx's in the preceding large block.  this is b/c large blocks take a bit more time to process and check all the signatures so the argument goes that they can't afford to waste that precious time so for defensive purposed they just go ahead and start hashing  the next block with only the  "header" that in this case contain no tx's that might have included an invalid input from the preceding large block.

How about: When they see a new block, they immediately start mining based on that, with an empty transaction list. When they have removed the transactions in the last block from their own list of transactions to include, they mine with the new list. If a block is found quickly, it will be the empty block, else it will be a block with the new list.

You can see that the zero transaction blocks always come a short time after the previous. But they should not need minutes to establish a new list, so I am not sure.

Maybe some miner can chime in and say what really is going on.


your first paragraph was correct.

i'm not sure why the 0 tx blocks come so fast after the previous block.  i keep asking this.

it's the sigop checking of a large block, i believe, that takes the extra time.

Well my first paragraph suggested a reason. They need time to construct a new transaction list. When that is done, new transactions can more easily be added. The first list is the time consuming part.

Edit: The block finding is completely random. It is just as probable to find a block the first second, as after several minutes.
legendary
Activity: 4690
Merit: 1276
June 30, 2015, 02:32:23 PM
I think tvbcof means to say Bitcoin wasn't ever going to be useful for point of sales. (I assume that means Bitcoin by itself.)

In the interest of not repeating old points, this is what Hearn has to say on this: https://medium.com/@octskyward/replace-by-fee-43edd9a1dd6d

Ah, Mr. Hearn.  Cypherdoc's giant among men.  I like Peter Todd's latest rendition which seems to be something like:  'The storm is coming but won't arrive until all non-hackers have had a fair chance to get on the lifeboats currently under construction.'

legendary
Activity: 1764
Merit: 1002
June 30, 2015, 02:29:20 PM

with RBF, does the sender have the ability to completely change the destination address back to himself?

What's your problem with that?  It's his money until it's not.

Oh, Bitcoin is not a real-time system in spite of years of propaganda trying to make it so?  How sad.  I think I'm going to cry.

Except that Bitcoin does function as a real-time system for most use cases today perfectly fine. There is an entire Bitcoin ecommerce industry that proves you wrong.

precisely.
legendary
Activity: 1153
Merit: 1000
June 30, 2015, 02:28:07 PM
The replace-by-fee mechanism allows the sender to change a transaction and increase a fee, while the child-pays-parent mechanism allows a receiver to add a fee to an existing transaction.

In most cases (and all cases with a smart wallet), child-pays-parent allows both sender or receiver to add a fee to an existing transaction.  The reason is that in most cases, a transaction contains an output that returns change to the sender.  This means the sender can create a child transaction that re-spends this change output, thereby increasing the parent's effective fee.  

Quote
And this is why RBF is completely contrary to the ethos of Bitcoin. Bitcoin was designed so that transactions can not be canceled and so that once a receiver takes control of BTC they have full control. The P2P network rejecting 2nd seen transactions is one aspect of that mechanism, the blockchain record is another.

Agreed.

Quote
RBF transfers power from the receiver and puts it back in the hands of the sender. With RBF after two people make an exchange, the sender now has the ability to reach into the receivers pocket and take the sent BTC back. This is horrifying.

In the case of non-FFS, I agree.  

I was referring to RBF as full-RBF in your terminology, we are in agreement on the above.
legendary
Activity: 1162
Merit: 1007
June 30, 2015, 02:27:04 PM
I think tvbcof means to say Bitcoin wasn't ever going to be useful for point of sales. (I assume that means Bitcoin by itself.)

In the interest of not repeating old points, this is what Hearn has to say on this: https://medium.com/@octskyward/replace-by-fee-43edd9a1dd6d

Right.  I think the people pushing full-RBF are more interested in making on-chain Bitcoin transactions less useful for point-of-sales than the often-stated goal of allowing users to "unstick" stuck transactions.  Since we can "unstick" a TX in ways that are first-seen safe, why break a system (zero-confirm TXs) that many people currently enjoy (well unless you're of the ideology that Bitcoin should only be a settlement backbone)?    
legendary
Activity: 1764
Merit: 1002
June 30, 2015, 02:26:48 PM
if you're saying where we're going is Moon, then how can a 1MB cap (assuming SC or LN take a year) even with wallet fee capability, result in a smoothly adjusting fee increase as we consistently hit the 1MB cap?  if anything, it will go parabolic in a chaotic fashion and result in extremely poor user experience and subsequent rejection.

Because it can be economically expected that most transactions now are low priority, since fees are cheap. This means the actual transaction volume that likely would be taking place given only somewhat more expensive fees is much, much less. It's should be obvious enough that if transactions were completely free people would spam like crazy. That's not a judgment on the validity of those transactions; it's just economic reality. People start to economize dramatically when there is even a marginally-above-zero cost versus when there isn't.

This is in no way suggesting 1MB is sufficient, but it's not helpful to pretend that transaction volume when fees are negligible will remain the same if fees go to merely tiny. In no other market would we expect such a thing, and for all we know we could easily be talking about order-of-magnitude differences in volume from a fee increase that still presents no real pain for any major use case.

but cheap to who, you?  i remember an Indian standing up at the San Jose conf during Q&A and complaining about high tx fees for ppl in his country.  and they are, given their relative income levels. 

we should be promoting Bitcoin to these 3rd world countries where a .0001 minimum fee is arguably expensive.  that's why we have the  economic fee choice of .00001 which oftentimes still results in rejection.  to take it further, do you believe that Bitcoin should offer even these Indians the ability to buy a cup of coffee for cheap fees?  i think so but probably you think different.  i know tvbcof and iCE would flat out say no way.  but then that goes back to my argument that for maximum decentralization and to become digital gold Bitcoin needs to service these ppl.

when you say ppl will spam like crazy, don't you trust that the miners can react to filter that spam, if they so choose?  the problem with the 1MB choke approach is that yes, new users will be forced to economize on their tx's.  OR just leave.
legendary
Activity: 1153
Merit: 1000
June 30, 2015, 02:26:40 PM

with RBF, does the sender have the ability to completely change the destination address back to himself?

What's your problem with that?  It's his money until it's not.

Oh, Bitcoin is not a real-time system in spite of years of propaganda trying to make it so?  How sad.  I think I'm going to cry.

Except that Bitcoin does function as a real-time system for most use cases today perfectly fine. There is an entire Bitcoin ecommerce industry that proves you wrong.
legendary
Activity: 4690
Merit: 1276
June 30, 2015, 02:25:39 PM

BTW, iCEBREAKER, thanks for getting me to finally look up what the Dunning-Kruger effect was by applying it to Frap.doc here.  nAILED-It.

typical response from you.

having no counter argument, you default resort to ad hominem.

There is really nothing here which bears 'arguing.'  When a differential in capabilities is above a certain level it simply makes no sense to engage in a conversation.  Rest assured, some people get it and have benefited by our conversation even if you have not.

legendary
Activity: 1036
Merit: 1000
June 30, 2015, 02:22:02 PM
I think tvbcof means to say Bitcoin wasn't ever going to be useful for point of sales. (I assume that means Bitcoin by itself.)

In the interest of not repeating old points, this is what Hearn has to say on this: https://medium.com/@octskyward/replace-by-fee-43edd9a1dd6d
legendary
Activity: 1162
Merit: 1007
June 30, 2015, 02:21:10 PM
There are two versions of replace-by-fee: "first-seen safe" and "full."  Peter Todd is pushing "full" replace-by-fee which breaks the security of zero-confirm transactions.  When F2Pool added replace-by-fee, they initially used the full version, but then switched to the first-seen safe version when the recognized the problem.

Which version do you think is good?  What does replace-by-fee do that child-pays-parent can't?

I am not an expert on the technicals, I thought it would fuck up zero conf anyway. You spend the same output with an increased fee. Miners can implement whatever selection criteria they want,  as has always been. You could complain, but when something is buried deep in the blockchain there is not much you can do. Making rules would change bitcoin, in my opinion. Anyway, not expert, I like to drivel on the economics side, and I think it is useful, but not helping with pile-ups, and not needed.


It sounds like you like the idea of being able to increase a transaction's effective fee to get it "unstuck" if necessary.  If that can be accomplished in a way that doesn't break first-seen-safe, would you consider that preferable?

Can someone explain to me how full-RBF offers any benefit over child-pays-parent or FFS-RBF?  As far as I can tell, the only "benefit" is that it breaks zero-confirm transactions and, like sticking with the 1 MB limit, provides impetus to use off-chain solutions.  
legendary
Activity: 1764
Merit: 1002
June 30, 2015, 02:18:55 PM

with RBF, does the sender have the ability to completely change the destination address back to himself?

What's your problem with that?  It's his money until it's not.

Oh, Bitcoin is not a real-time system in spite of years of propaganda trying to make it so?  How sad.  I think I'm going to cry.


whoosh.  completely over your head, isn't it?

the only reason a vendor has the confidence to let someone walk out the store with a cup of coffee instantly is that the tx is irreversible for the most part.  by forcing RBF, now the vendor has to stop everything else he's doing, like serving MORE cups of coffee, to watch that the buyer doesn't walk out of the store before receiving 1 confirmation.  

you'd make a great marketing strategist.  Roll Eyes

BTW, iCEBREAKER, thanks for getting me to finally look up what the Dunning-Kruger effect was by applying it to Frap.doc here.  nAILED-It.



typical response from you.

having no counter argument, you default resort to ad hominem.
legendary
Activity: 4690
Merit: 1276
June 30, 2015, 02:16:34 PM

with RBF, does the sender have the ability to completely change the destination address back to himself?

What's your problem with that?  It's his money until it's not.

Oh, Bitcoin is not a real-time system in spite of years of propaganda trying to make it so?  How sad.  I think I'm going to cry.


whoosh.  completely over your head, isn't it?

the only reason a vendor has the confidence to let someone walk out the store with a cup of coffee instantly is that the tx is irreversible for the most part.  by forcing RBF, now the vendor has to stop everything else he's doing, like serving MORE cups of coffee, to watch that the buyer doesn't walk out of the store before receiving 1 confirmation. 

you'd make a great marketing strategist.  Roll Eyes

BTW, iCEBREAKER, thanks for getting me to finally look up what the Dunning-Kruger effect was by applying it to Frap.doc here.  nAILED-It.

legendary
Activity: 1512
Merit: 1005
June 30, 2015, 02:16:08 PM

I think replace by fee is good. It is a way to unstick a transaction that is stuck with too low fee, when you are in a hurry.  It does not change the protocol in any way, it is just the miner chooses the transaction with the higher fee, making the old one illegal. It does not fill up blocks, only the network.

On the other hand, I don't think it is critically important, necessary or even specially useful. I think the market will sort out fees, making normal high-paying transactions fill up only half the space available in blocks.


There are two versions of replace-by-fee: "first-seen safe" and "full."  Peter Todd is pushing "full" replace-by-fee which breaks the security of zero-confirm transactions.  When F2Pool added replace-by-fee, they initially used the full version, but then switched to the first-seen safe version when the recognized the problem.

Which version do you think is good?  What does replace-by-fee do that child-pays-parent can't?

I am not an expert on the technicals, I thought it would fuck up zero conf anyway. You spend the same output with an increased fee. Miners can implement whatever selection criteria they want,  as has always been. You could complain, but when something is buried deep in the blockchain there is not much you can do. Making rules would change bitcoin, in my opinion. Anyway, not expert, I like to drivel on the economics side, and I think it is useful, but not helping with pile-ups, and not needed.

I know nothing about child-pays-parent, but but but... the miner must mine the parent if he wants the fee from the child. Ok, sounds good, need no change in protocol, go ahead. Why not both? But still not needed IMO, the market will solve the problem of pile-ups as long as the fee mechanism is there.

legendary
Activity: 1764
Merit: 1002
June 30, 2015, 02:09:49 PM

with RBF, does the sender have the ability to completely change the destination address back to himself?

What's your problem with that?  It's his money until it's not.

Oh, Bitcoin is not a real-time system in spite of years of propaganda trying to make it so?  How sad.  I think I'm going to cry.



whoosh.  completely over your head, isn't it?

the only reason a vendor has the confidence to let someone walk out the store with a cup of coffee instantly is that the tx is irreversible for the most part.  by forcing RBF, now the vendor has to stop everything else he's doing, like serving MORE cups of coffee, to watch that the buyer doesn't walk out of the store before receiving 1 confirmation. 

you'd make a great marketing strategist.  Roll Eyes
legendary
Activity: 1036
Merit: 1000
June 30, 2015, 02:08:39 PM
if you're saying where we're going is Moon, then how can a 1MB cap (assuming SC or LN take a year) even with wallet fee capability, result in a smoothly adjusting fee increase as we consistently hit the 1MB cap?  if anything, it will go parabolic in a chaotic fashion and result in extremely poor user experience and subsequent rejection.

Because it can be economically expected that most transactions now are low priority, since fees are cheap. This means the actual transaction volume that likely would be taking place given only somewhat more expensive fees is much, much less. It's should be obvious enough that if transactions were completely free people would spam like crazy. That's not a judgment on the validity of those transactions; it's just economic reality. People start to economize dramatically when there is even a marginally-above-zero cost versus when there isn't.

This is in no way suggesting 1MB is sufficient, but it's not helpful to pretend that transaction volume when fees are negligible will remain the same if fees go to merely tiny. In no other market would we expect such a thing, and for all we know we could easily be talking about order-of-magnitude differences in volume from a fee increase that still presents no real pain for any major use case.
Jump to: