Topic: i have proven the Lightning Network can't provide decentralized scaling. (Read 2328 times)
legendary
Activity: 1302
Merit: 1008
Core dev leaves me neg feedback #abuse #political
July 16, 2017, 10:47:51 PM
yes , I replied on her blog post. she proves my point as her model requires 14 open channels per user and each person can only send 0.01 BTC
July 05, 2017, 04:11:36 PM
July 05, 2017, 03:56:22 PM
If Jonald was indeed a paid BU shill, he works hard for the money LOL 
Yes good work but I do not agree with the statement, that a partially centralized external layer is dangerous.
From a practical viewpoint, if some of the LN hubs want to buttrape me (e.g. KYC documents for their services) I will stop using those.
If all of them do it, I can fall back to onchain (so coins will never become worthless) and the community will abandon LN and move on to something else. LN is not irreversible....max risk IMHO is losing the Bitcoins that are blocked in a specific channel...which is much much less than what people put in exchanges.
Correct me if I'm wrong.
Cheers
leopard2
Yes good work but I do not agree with the statement, that a partially centralized external layer is dangerous.
From a practical viewpoint, if some of the LN hubs want to buttrape me (e.g. KYC documents for their services) I will stop using those.
If all of them do it, I can fall back to onchain (so coins will never become worthless) and the community will abandon LN and move on to something else. LN is not irreversible....max risk IMHO is losing the Bitcoins that are blocked in a specific channel...which is much much less than what people put in exchanges.
Correct me if I'm wrong.
Cheers
leopard2
July 05, 2017, 02:47:49 PM
And UaSF try to force themselfe into their lightning - sounds bit esotheric and helps bitcoin homeopaticaly.
Some good hashpower from above might get them back to the mainchain soon, hush hush.
Some good hashpower from above might get them back to the mainchain soon, hush hush.
July 03, 2017, 08:10:23 AM
the reality people need to realise is
FIAT spenders use debit cards 40 times a month on average, but because bitcoin is not acceptable in every retailer
the average spender only uses bitcoin once a week/once a year.(5 a month) based on person-merchant
the only real niche LN has is the faucet raiders and exchange day traders that will want to arbitrage daily. (average 10 a day)
knowing people cannot predict spending habits beyond a couple weeks and wont risk locking entire hoards into year/eternity long channels. the reality is channels will only get funded with pocket money amounts and for only a couple weeks lock periods.
these things limit the available funding for route hopping to occur
LN is not.. and i emphasise this.. is not about the funds of [A-B] $120 actually moving out of the channel. LN is about having 'pockets' of funds that 2 parties share. and they have a couple pockets dedicated to share
think of it like a 3 legged race where the legs are tied together with a fanny/bumbag. where the money stays in the fanny/bumbag. but who owes what % of whats inside the fanny/bumbag is agreed by the 2 people tied to it.
so get 5 people and play a game
EG
[A-B] [B-C] [C-D] [D-E]
imagine BCD only trust $60 for 2 weeks
A need to deposit $60 (1 channel with B of $60)
B need to deposit $120 (2 channels of $60 with A and C)
C need to deposit $120 (2 channels of $60 with B and D)
D need to deposit $120 (2 channels of $60 with C and E)
E need to deposit $60 (1 channel with D of $60)
[$60-$60][$60-$60][$60-$60][$60-$60]
now imagine if A wants to pay E $60
[$0-$120][$60-$60][$60-$60][$60-$60] then
[$0-$120][$0-$120][$60-$60][$60-$60] then
[$0-$120][$0-$120][$0-$120][$60-$60] then
[$0-$120][$0-$120][$0-$120][$0-$120] now finally after 4 hops E has $60 extra thanks to the 'routing'/hops
technically B, C, D still has $120 but its not spread over the channels
for instance B has $120
but has $120 in [A-B] but $0 in [B-C] channel
meaning if B wants to pay D.. B can no longer user the [B-C][C-D] route because B doesnt have funds to pay C to pay D.
instead B needs to go backwards through [A-B] because thats where B's funds are, so...
B needs to give A.. then A needs to find a completely new route or create a new channel that is funded and has a new route to E to get E to pay D. or B needs to set up a new channel to fun that has a new route to E.
.. it may seem complicated but its alot easier visually if you just get 5 friends and 3 fanny/bumbags and put $120 of monopoly money in each bag ($60) each person... and without physically taking the paper money out.. agreeing who owes what inside each bumbag and then finding the best chain of hops to fund the group..
and see how long the 'payments' last by making up scenarios of buying each other things.. you soon learn the limitations of LN
next time you are at a bitcoin meet-up.. try it with the other attendee's. as its a very helpful visual display
FIAT spenders use debit cards 40 times a month on average, but because bitcoin is not acceptable in every retailer
the average spender only uses bitcoin once a week/once a year.(5 a month) based on person-merchant
the only real niche LN has is the faucet raiders and exchange day traders that will want to arbitrage daily. (average 10 a day)
knowing people cannot predict spending habits beyond a couple weeks and wont risk locking entire hoards into year/eternity long channels. the reality is channels will only get funded with pocket money amounts and for only a couple weeks lock periods.
these things limit the available funding for route hopping to occur
LN is not.. and i emphasise this.. is not about the funds of [A-B] $120 actually moving out of the channel. LN is about having 'pockets' of funds that 2 parties share. and they have a couple pockets dedicated to share
think of it like a 3 legged race where the legs are tied together with a fanny/bumbag. where the money stays in the fanny/bumbag. but who owes what % of whats inside the fanny/bumbag is agreed by the 2 people tied to it.
so get 5 people and play a game
EG
[A-B] [B-C] [C-D] [D-E]
imagine BCD only trust $60 for 2 weeks
A need to deposit $60 (1 channel with B of $60)
B need to deposit $120 (2 channels of $60 with A and C)
C need to deposit $120 (2 channels of $60 with B and D)
D need to deposit $120 (2 channels of $60 with C and E)
E need to deposit $60 (1 channel with D of $60)
[$60-$60][$60-$60][$60-$60][$60-$60]
now imagine if A wants to pay E $60
[$0-$120][$60-$60][$60-$60][$60-$60] then
[$0-$120][$0-$120][$60-$60][$60-$60] then
[$0-$120][$0-$120][$0-$120][$60-$60] then
[$0-$120][$0-$120][$0-$120][$0-$120] now finally after 4 hops E has $60 extra thanks to the 'routing'/hops
technically B, C, D still has $120 but its not spread over the channels
for instance B has $120
but has $120 in [A-B] but $0 in [B-C] channel
meaning if B wants to pay D.. B can no longer user the [B-C][C-D] route because B doesnt have funds to pay C to pay D.
instead B needs to go backwards through [A-B] because thats where B's funds are, so...
B needs to give A.. then A needs to find a completely new route or create a new channel that is funded and has a new route to E to get E to pay D. or B needs to set up a new channel to fun that has a new route to E.
.. it may seem complicated but its alot easier visually if you just get 5 friends and 3 fanny/bumbags and put $120 of monopoly money in each bag ($60) each person... and without physically taking the paper money out.. agreeing who owes what inside each bumbag and then finding the best chain of hops to fund the group..
and see how long the 'payments' last by making up scenarios of buying each other things.. you soon learn the limitations of LN
next time you are at a bitcoin meet-up.. try it with the other attendee's. as its a very helpful visual display
July 03, 2017, 06:28:12 AM
RECAP :
1) Jhonny (bless him..) posts his document telling us that LN won't work and that he has proven it mathematically.
2) Bram Cohen (Inventor of BitTorrent) replies :
3) Jhonny retorts :
4) Bram retorts :
5) Jhonny fires back :
....
.. as always.. please fasten your seat belt and make up your own mind.
ps.. in 5).. Jhonny is saying that the fact you have to make a single on chain txn to start using LN (and then potentially do infinite txns), negates LN. I'm not kidding.
1) Jhonny (bless him..) posts his document telling us that LN won't work and that he has proven it mathematically.
2) Bram Cohen (Inventor of BitTorrent) replies :
Quote
In this analysis nodes aren’t bothering to make sure that they have any connections at all. Shockingly, if a single node has no connections whatsoever then no amount of other connections in the system will make it routable. If instead every node enforce that it has at least three connections then these problems evaporate and everything works fine. I’m not being flippant here, this analysis really is that dumb.
Also there’s nothing wrong with long routes. They settle out in the middle just fine, despite the author’s dismissiveness to the possibility that they can.
This is not to say that a completely random homogenous network is an ideal or even necessarily a good way of setting up routing, but it works fine as long as every node makes sure that it has a minimal level of connectivity.
Also there’s nothing wrong with long routes. They settle out in the middle just fine, despite the author’s dismissiveness to the possibility that they can.
This is not to say that a completely random homogenous network is an ideal or even necessarily a good way of setting up routing, but it works fine as long as every node makes sure that it has a minimal level of connectivity.
3) Jhonny retorts :
Quote
Bram, are you sure you understand the article? “At least 3 connections” means you have to divide your money up “at least 3 ways”. With only 3 open channels, you’d be then relying on long routes which make actually getting your payment through exponentially less likely, as the math illustrates. Not sure what “settle out in the middle” means to you.
4) Bram retorts :
Quote
‘Settle out in the middle’ means using the smart transactions bit of Lighting so that there’s no lending happening in the middle. Maybe you should learn about how technologies work before posting tirades about them.
As for your math, I can’t tell if you’re disingenuous or just stupid, but this is a very simple concept: The probabilities of different routes working aren’t independent of each other, because the target node will notice if they have no connections at all and form some, so if some routes don’t work that increases the probability of other routes working.
If you assume that peers don’t make sure they have connections then your math is mostly correct, but your claim is that ‘lightning network can’t work’, not ‘lightning network doesn’t work with this asinine implementation I came up with’, which is what you actually showed.
As for your math, I can’t tell if you’re disingenuous or just stupid, but this is a very simple concept: The probabilities of different routes working aren’t independent of each other, because the target node will notice if they have no connections at all and form some, so if some routes don’t work that increases the probability of other routes working.
If you assume that peers don’t make sure they have connections then your math is mostly correct, but your claim is that ‘lightning network can’t work’, not ‘lightning network doesn’t work with this asinine implementation I came up with’, which is what you actually showed.
5) Jhonny fires back :
Quote
With all due respect, I believe it is you who is being “disingenuous or just stupid.” The target node cannot form new connections without doing an on-chain transaction, which defeats the whole point: Why not just send an on-chain transaction in the first place, without the LN.
....
.. as always.. please fasten your seat belt and make up your own mind. ps.. in 5).. Jhonny is saying that the fact you have to make a single on chain txn to start using LN (and then potentially do infinite txns), negates LN. I'm not kidding.
Ouch - that (and more above) is killing LN. I wait for first LN Fan Boy triade here wanting SW and LN work despite of this AND missing out any usual financial product disclaimer in case I might lose my bitcoin on that - nobody can tell there is 100% security on that - even bitcoin on-chain tx are not 100% secure but much closer to than anything known yet.
July 03, 2017, 04:31:42 AM
RECAP :
1) Jhonny (bless him..) posts his document telling us that LN won't work and that he has proven it mathematically.
2) Bram Cohen (Inventor of BitTorrent) replies :
3) Jhonny retorts :
4) Bram retorts :
5) Jhonny fires back :
....
.. as always.. please fasten your seat belt and make up your own mind.
ps.. in 5).. Jhonny is saying that the fact you have to make a single on chain txn to start using LN (and then potentially do infinite txns), negates LN. I'm not kidding.
1) Jhonny (bless him..) posts his document telling us that LN won't work and that he has proven it mathematically.
2) Bram Cohen (Inventor of BitTorrent) replies :
Quote
In this analysis nodes aren’t bothering to make sure that they have any connections at all. Shockingly, if a single node has no connections whatsoever then no amount of other connections in the system will make it routable. If instead every node enforce that it has at least three connections then these problems evaporate and everything works fine. I’m not being flippant here, this analysis really is that dumb.
Also there’s nothing wrong with long routes. They settle out in the middle just fine, despite the author’s dismissiveness to the possibility that they can.
This is not to say that a completely random homogenous network is an ideal or even necessarily a good way of setting up routing, but it works fine as long as every node makes sure that it has a minimal level of connectivity.
Also there’s nothing wrong with long routes. They settle out in the middle just fine, despite the author’s dismissiveness to the possibility that they can.
This is not to say that a completely random homogenous network is an ideal or even necessarily a good way of setting up routing, but it works fine as long as every node makes sure that it has a minimal level of connectivity.
3) Jhonny retorts :
Quote
Bram, are you sure you understand the article? “At least 3 connections” means you have to divide your money up “at least 3 ways”. With only 3 open channels, you’d be then relying on long routes which make actually getting your payment through exponentially less likely, as the math illustrates. Not sure what “settle out in the middle” means to you.
4) Bram retorts :
Quote
‘Settle out in the middle’ means using the smart transactions bit of Lighting so that there’s no lending happening in the middle. Maybe you should learn about how technologies work before posting tirades about them.
As for your math, I can’t tell if you’re disingenuous or just stupid, but this is a very simple concept: The probabilities of different routes working aren’t independent of each other, because the target node will notice if they have no connections at all and form some, so if some routes don’t work that increases the probability of other routes working.
If you assume that peers don’t make sure they have connections then your math is mostly correct, but your claim is that ‘lightning network can’t work’, not ‘lightning network doesn’t work with this asinine implementation I came up with’, which is what you actually showed.
As for your math, I can’t tell if you’re disingenuous or just stupid, but this is a very simple concept: The probabilities of different routes working aren’t independent of each other, because the target node will notice if they have no connections at all and form some, so if some routes don’t work that increases the probability of other routes working.
If you assume that peers don’t make sure they have connections then your math is mostly correct, but your claim is that ‘lightning network can’t work’, not ‘lightning network doesn’t work with this asinine implementation I came up with’, which is what you actually showed.
5) Jhonny fires back :
Quote
With all due respect, I believe it is you who is being “disingenuous or just stupid.” The target node cannot form new connections without doing an on-chain transaction, which defeats the whole point: Why not just send an on-chain transaction in the first place, without the LN.
....
.. as always.. please fasten your seat belt and make up your own mind. ps.. in 5).. Jhonny is saying that the fact you have to make a single on chain txn to start using LN (and then potentially do infinite txns), negates LN. I'm not kidding.
July 03, 2017, 03:41:14 AM
The projection is real mass business.
Of course it is. That's why we need to consider other options than linear on-chain scaling.
Even a tiny share for 2nd layer central crap might cost you a bunch of bitcoin because of some lower security you cannot grasp yet.
[citation needed]
Corps and banks are way more picky on that and can afford huge risk analysis. They will all stay off bitcoin and you stay nichy.
Bitcoins biggest challenges when it comes to business and banks are legal and economic. The technical aspects are miniscule.
Good bye
Have a nice day
Tx !
Entry point for big business is always risk / reward. RIsks: There a lots of: Operational / Market. (Reputation, Systemic, Counterparty, Legal.../ Volatility) Hope you know this a bit.
If your riskmanagers throw the red flag on only one of these parts and you cannot fix / control / MINIMIZE* them - bitcoin is OUT - byebye.
Either blockstream knows all that shit and want to do only a quick buck - or they kill bitcoin not knowing better.
Think bigger
*MINIMIZE risk = MAXIMIZE on-chain scaling - easy to get.
July 03, 2017, 02:46:44 AM
jonald_fyookball, we have seen that you have "proven" cannot provide decentralized scaling. Without any arguments made from the developers of LN around, let us assume that you are correct. Then what is your proposal on how to do it right?
July 03, 2017, 02:39:31 AM
Fair points for sure, but yeah, even users like me will be happy to use LN for smaller payments. I'd be able to transact more often and with risks on those small amounts. I don't because there isn't a cheap and fast way.
But why crippling the on-chain usage? No one has answered this question reasonably.
July 03, 2017, 02:30:48 AM
The projection is real mass business.
Of course it is. That's why we need to consider other options than linear on-chain scaling.
Even a tiny share for 2nd layer central crap might cost you a bunch of bitcoin because of some lower security you cannot grasp yet.
[citation needed]
Corps and banks are way more picky on that and can afford huge risk analysis. They will all stay off bitcoin and you stay nichy.
Bitcoins biggest challenges when it comes to business and banks are legal and economic. The technical aspects are miniscule.
Good bye
Have a nice day
July 02, 2017, 02:16:47 PM
Again, not fully right. If you take the protocol layer, tell me what might help to scale: thin protocol, few simple rules, many teams competing - or that what we have now ?
A thin protocol with a few simple rules is what he have now. Moving scaling details to a second layer would help keep it that way. Or are you arguing in favour of adding complexity to the base layer?
Software separate isn't the issue, but rather economics. Proof of Work secures the base layer. Any money for fees, etc that's not on the base layer is draining security funding from that base layer.
Miners still get their block rewards and transaction fees. Even in the far future when block rewards are gone they still would get at least the amount of transaction fees that they are getting now (or like 6-12 months ago before blocks started to get full).
The projection is real mass business. Even a tiny share for 2nd layer central crap might cost you a bunch of bitcoin because of some lower security you cannot grasp yet. Corps and banks are way more picky on that and can afford huge risk analysis. They will all stay off bitcoin and you stay nichy. Good bye
legendary
Activity: 1302
Merit: 1008
Core dev leaves me neg feedback #abuse #political
July 02, 2017, 02:07:16 PM
Again, not fully right. If you take the protocol layer, tell me what might help to scale: thin protocol, few simple rules, many teams competing - or that what we have now ?
A thin protocol with a few simple rules is what he have now. Moving scaling details to a second layer would help keep it that way. Or are you arguing in favour of adding complexity to the base layer?
Software separate isn't the issue, but rather economics. Proof of Work secures the base layer. Any money for fees, etc that's not on the base layer is draining security funding from that base layer.
Miners still get their block rewards and transaction fees. Even in the far future when block rewards are gone they still would get at least the amount of transaction fees that they are getting now (or like 6-12 months ago before blocks started to get full).
Yes there is plenty of money for miners for now. But I dont' see any reason we need to force second layers.
July 02, 2017, 01:51:27 PM
Again, not fully right. If you take the protocol layer, tell me what might help to scale: thin protocol, few simple rules, many teams competing - or that what we have now ?
A thin protocol with a few simple rules is what he have now. Moving scaling details to a second layer would help keep it that way. Or are you arguing in favour of adding complexity to the base layer?
Software separate isn't the issue, but rather economics. Proof of Work secures the base layer. Any money for fees, etc that's not on the base layer is draining security funding from that base layer.
Miners still get their block rewards and transaction fees. Even in the far future when block rewards are gone they still would get at least the amount of transaction fees that they are getting now (or like 6-12 months ago before blocks started to get full).
legendary
Activity: 1302
Merit: 1008
Core dev leaves me neg feedback #abuse #political
July 02, 2017, 10:32:39 AM
I cannot agree in full here. He also says wise words about scarcity at all. If you spend recources to on side, there is always another side suffering.
Translate
If you spend things on 2nd layer dev, on chain scaling does not get full support. Choose what you want bitcoin to be. I vote for on chain as much as possible. Its our best security.
Translate
If you spend things on 2nd layer dev, on chain scaling does not get full support. Choose what you want bitcoin to be. I vote for on chain as much as possible. Its our best security.
That's not how software development works. The whole point of separate abstraction layers (or protocol layers in this case) is that you can have separate development efforts working on different parts of the problem without interference. Either way, if on-chain scaling would be as simple as changing the blocksize variable, there'd be no development effort needed for on-chain scaling in the first place.
Software separate isn't the issue, but rather economics. Proof of Work secures the base layer. Any money for fees, etc that's not on the base layer is draining security funding from that base layer.
July 02, 2017, 10:11:23 AM
I cannot agree in full here. He also says wise words about scarcity at all. If you spend recources to on side, there is always another side suffering.
Translate
If you spend things on 2nd layer dev, on chain scaling does not get full support. Choose what you want bitcoin to be. I vote for on chain as much as possible. Its our best security.
Translate
If you spend things on 2nd layer dev, on chain scaling does not get full support. Choose what you want bitcoin to be. I vote for on chain as much as possible. Its our best security.
That's not how software development works. The whole point of separate abstraction layers (or protocol layers in this case) is that you can have separate development efforts working on different parts of the problem without interference. Either way, if on-chain scaling would be as simple as changing the blocksize variable, there'd be no development effort needed for on-chain scaling in the first place.
Again, not fully right. If you take the protocol layer, tell me what might help to scale: thin protocol, few simple rules, many teams competing - or that what we have now ?
July 02, 2017, 10:07:07 AM
I cannot agree in full here. He also says wise words about scarcity at all. If you spend recources to on side, there is always another side suffering.
Translate
If you spend things on 2nd layer dev, on chain scaling does not get full support. Choose what you want bitcoin to be. I vote for on chain as much as possible. Its our best security.
Translate
If you spend things on 2nd layer dev, on chain scaling does not get full support. Choose what you want bitcoin to be. I vote for on chain as much as possible. Its our best security.
That's not how software development works. The whole point of separate abstraction layers (or protocol layers in this case) is that you can have separate development efforts working on different parts of the problem without interference. Either way, if on-chain scaling would be as simple as changing the blocksize variable, there'd be no development effort needed for on-chain scaling in the first place.
July 02, 2017, 09:46:42 AM
You may think I begin trolling now 
... but I mostly agree with your analysis. I have discussed, above all, the "channel exhaustion" (=channels that lose the "routing power") problem with other users and no solution (that doesn't involve a centralized entity) has been found.
For me, LN was never more than a semi-centralized micropayment solution that is better than a totally-centralized online wallet because the "bank" cannot run away with my money. I would be happy if I can manage all my 0-20USD payments via LN and having at least 10 different hubs to chose from.
... but I mostly agree with your analysis. I have discussed, above all, the "channel exhaustion" (=channels that lose the "routing power") problem with other users and no solution (that doesn't involve a centralized entity) has been found. For me, LN was never more than a semi-centralized micropayment solution that is better than a totally-centralized online wallet because the "bank" cannot run away with my money. I would be happy if I can manage all my 0-20USD payments via LN and having at least 10 different hubs to chose from.
As much as we all are for "as much decentralisation as possible", there is not going to be "true decentralisation". The evidence for and against all solutions seem to hinge around this idea: "it won't work because that's centralisation".
Fair points for sure, but yeah, even users like me will be happy to use LN for smaller payments. I'd be able to transact more often and with risks on those small amounts. I don't because there isn't a cheap and fast way.
July 02, 2017, 09:01:16 AM
Check this video at 3:09
LN is a Mesh with 80 Hubs. always attackable. Anyone get this explained?
https://bitcointalksearch.org/topic/is-matonis-wright-analyse-this-video-and-build-your-settled-view-1997310
And yes, if you do not concentrate to on chain scaling, you allow starting that less secure crap around, weakening bitcoin network security...
LN is a Mesh with 80 Hubs. always attackable. Anyone get this explained?
https://bitcointalksearch.org/topic/is-matonis-wright-analyse-this-video-and-build-your-settled-view-1997310
And yes, if you do not concentrate to on chain scaling, you allow starting that less secure crap around, weakening bitcoin network security...
Bitcoin network security can't get weakened by layers you put on top of it, unless you change the very core of the Bitcoin protocol itself.
I cannot agree in full here. He also says wise words about scarcity at all. If you spend recources to one side, there is always another side suffering.
Translate
If you spend things on 2nd layer dev, on chain scaling does not get full support. Choose what you want bitcoin to be. I vote for on chain as much as possible. Its our best security.
July 02, 2017, 07:53:07 AM
Check this video at 3:09
LN is a Mesh with 80 Hubs. always attackable. Anyone get this explained?
https://bitcointalksearch.org/topic/is-matonis-wright-analyse-this-video-and-build-your-settled-view-1997310
And yes, if you do not concentrate to on chain scaling, you allow starting that less secure crap around, weakening bitcoin network security...
LN is a Mesh with 80 Hubs. always attackable. Anyone get this explained?
https://bitcointalksearch.org/topic/is-matonis-wright-analyse-this-video-and-build-your-settled-view-1997310
And yes, if you do not concentrate to on chain scaling, you allow starting that less secure crap around, weakening bitcoin network security...
Bitcoin network security can't get weakened by layers you put on top of it, unless you change the very core of the Bitcoin protocol itself.
The bit at 03:09 is assuming that Sybil attacks are viable. Possibility does not equal viability. Claiming that the Bitcoin is fully distributed doesn't mean anything if you equate non-mining full nodes with the handful of mining pools that are actually confirming transactions.
Jump to: