The Lightning Network FAQ

[LoyceV]

Thus, it might seem a bad idea to allow unenforceable amounts

We have the same in euros: most shops reject 1 and 2 cent coins, and yet most prices end on 9 cents. Entire cents, currently worty far more than 100 satoshi, are lost due to rounding. And nobody cares
Why would anyone care about possibly losing a fraction of a satoshi due to rounding?

[Wind_FURY]

OK, is it as simple as changing the most basic unit of account from sats to millisats? How does that "transformation" happen in Lightning. Or is it simply because any rule can be made up in L2 networks?

Yes, and making it unenforceable onchain (as dust outputs once again!). Some are critic regarding that choice.

However, accounting for such small amounts are really handy : for example for the fees. Fees are denominated in msat and most are below the thousand msat, without such an accounting system how could fees that low (or that precise !) be enforced ? Paying one whole sat but once each X forwarded payments ? That's not viable.

OR for example, the possbility of a more than 6 digit per Bitcoin world. Instead of breaking Bitcoin's social contract by adding more decimal places onchain, do it in L2.

[darosior]

Thus, it might seem a bad idea to allow unenforceable amounts

We have the same in euros: most shops reject 1 and 2 cent coins, and yet most prices end on 9 cents.

This is not comparable at all, I think. Shops reject 1 and 2 cents coins because they can't automagically get a 10$ ticket out of a thousand of them. On LN they can.
And this was not about payments but about forwarding fees. Even if you can see an utility for payments too (pay-per-call on an API for example).

Entire cents, currently worty far more than 100 satoshi, are lost due to rounding. And nobody cares
Why would anyone care about possibly losing a fraction of a satoshi due to rounding?

Again, not comparable : my point that these tiny amounts may become valuable in a fraction of seconds. That is again true for a forwarding node or a payee for an autonomous service.

OR for example, the possbility of a more than 6 digit per Bitcoin world. Instead of breaking Bitcoin's social contract by adding more decimal places onchain, do it in L2.

Absolutely, I think the need for those tiny amounts is already present today, and it might be even more with a price increase !

[Rath_]

Yesterday, LND released lnd v0.10-beta. Aside from many small improvements, there are two notable features. Sending up to 0.167 BTC via multiple paths (multipart payments) is now possible. Eclair also has a full MPP support while c-lightning supports only receiving so as long as most nodes are up-to-date, the payments should go through without any problems. The other feature worth mentioning is support for Partially-Signed Bitcoin Transaction (PSBT). It is now possible to fund multiple channels using a single transaction directly from one's wallet if it supports PSBT.

[Wind_FURY]

Yesterday, LND released lnd v0.10-beta. Aside from many small improvements, there are two notable features. Sending up to 0.167 BTC via multiple paths (multipart payments) is now possible. Eclair also has a full MPP support while c-lightning supports only receiving so as long as most nodes are up-to-date, the payments should go through without any problems. The other feature worth mentioning is support for Partially-Signed Bitcoin Transaction (PSBT). It is now possible to fund multiple channels using a single transaction directly from one's wallet if it supports PSBT.

Hypothetical question. If I provide larger channel capacity in LN , but with moderately higher fees required, would you choose to pay my fee to complete payment "in a single shot"? Asking for a friend.

https://github.com/lightningnetwork/lnd/releases/tag/v0.10.0-beta

The splitting algorithm that is employed is a variation of "divide and conquer". First, lnd will try to complete the payment in a single shot. If that is not possible - because channel capacities aren't sufficient or other constraints like the fee limit are hit - an attempt is made for half the payment amount. After that htlc has been launched, a second path is searched for to pay the remaining amount (the other half). If further failures happen, amounts are repeatedly split until a minimum amount limit is reached or the payment succeeds.

We opted for this relatively simple strategy to begin with. It is expected to address various basic liquidity problems that arose in the past, but it is unlikely to always execute payments in a way that is optimal. The field of multi-path planning is still mostly uncharted and contains lots of challenges. Therefore we hope to collect valuable feedback with this release and use that to drive further development.

[darosior]

Hypothetical question. If I provide larger channel capacity in LN , but with moderately higher fees required, would you choose to pay my fee to complete payment "in a single shot"? Asking for a friend.

I think MPP will still only be used as a fallback for some time, so yes (your route will be tried first). However at some point if the sum of the base fee of the MPP routes is cheaper... No.

[jackg]

There's the other question that you could be paying a higher fee to be deanonymised... One of the perks of the ln goes if everyone cna use a certain node to route through.

I know this was their plan and it takes a while for my phone to load all the routes even at the moment (to sync I mean) but is there a way to disable it if you wanted to be more anonymous?

[DaveF]

Hypothetical question. If I provide larger channel capacity in LN , but with moderately higher fees required, would you choose to pay my fee to complete payment "in a single shot"? Asking for a friend.

I am with darosior on this. Not going to pay higher fees for anything I can avoid.

With that being said, I am running two of my own lightning nodes so I may not be the best person to ask. Which, is my next point. Asking people in this thread may not be the best place to ask, because you are going to get the answer from a bunch of people who may or may not run their own node, but who do understand or are trying to understand the LN and will probably not want to pay more fees for very little benefit, and you would loose a bit of privacy (what jackg said).

Stay safe.

-Dave

[darosior]

is there a way to disable it if you wanted to be more anonymous?

To disable what ? MPP ? To be more anonymous how ?

[jackg]

is there a way to disable it if you wanted to be more anonymous?

To disable what ? MPP ? To be more anonymous how ?

Yes, the idea with the lightning network originally was that there were essentially 2 layers of nodes... The first layer deals with how to find people and routing information and the second deals with participating in the network (sending and receiving nodes but not necessarily routing).

Since acinq at least used to have the longest running nodes and were the main developers I was trying to work out if there's a way to disable funds going through any main servers (even though we can probably trust the developers of the ln, no device is completely secure and if you want to be fairly anonymous)...

[Wind_FURY]

Hypothetical question. If I provide larger channel capacity in LN , but with moderately higher fees required, would you choose to pay my fee to complete payment "in a single shot"? Asking for a friend.

I am with darosior on this. Not going to pay higher fees for anything I can avoid.

With that being said, I am running two of my own lightning nodes so I may not be the best person to ask. Which, is my next point. Asking people in this thread may not be the best place to ask, because you are going to get the answer from a bunch of people who may or may not run their own node, but who do understand or are trying to understand the LN and will probably not want to pay more fees for very little benefit, and you would loose a bit of privacy (what jackg said).

Stay safe.

-Dave

For personal use, I believe you're right. For merchants to handle high-volume of transactions, I believe high-liquidity/higher-fee channels will be the right choice. They shouldn't let the customer wait.

[darosior]

Yes, the idea with the lightning network originally was that there were essentially 2 layers of nodes... The first layer deals with how to find people and routing information and the second deals with participating in the network (sending and receiving nodes but not necessarily routing).

Yep I'm aware, but I still don't understand the relation with being "more anonymous"?.. The only part of added privacy I can think of is your direct peer not knowing the full amount you want to transfer

Since acinq at least used to have the longest running nodes and were the main developers I was trying to work out if there's a way to disable funds going through any main servers (even though we can probably trust the developers of the ln, no device is completely secure and if you want to be fairly anonymous)...

Why would you not want the ACINQ node to be in your route ? If they aren't the first node they don't know you are part of it ..

In addition, regarding your sentence about trust in ln developers : you don't need to trust anyone in the network with your funds. Just the open source software you are running.

[Wind_FURY]

For those 1/2/3 sats/byte transactions waiting for DAYS in the mempool. If they were in Lightning, formed like an "organized mempool" waiting to be confirmed back onchain, they would be going back and forth instantly, possibly for lower fees. Why are merchants so slow in adopting it?

https://jochen-hoenicke.de/queue/#0,24h

[DaveF]

For those 1/2/3 sats/byte transactions waiting for DAYS in the mempool. If they were in Lightning, formed like an "organized mempool" waiting to be confirmed back onchain, they would be going back and forth instantly, possibly for lower fees. Why are merchants so slow in adopting it?

https://jochen-hoenicke.de/queue/#0,24h

Old question, same answer :
1. There aren't many payment processor/gateway which accept LN
2. There aren't many wallet which support LN, personally i doubt custodial or web wallet will support LN at all
3. Trade-off between convenience or full-control when choosing LN wallet

1) True. But there are more then there used to be. Off the top of my head there is
    a) BTCPay
    b) coingate
    c) zap

2) Why would you want a custodial wallet? But they are out there.
   a) Wallet of Satoshi
   b) bluewallet

3) Would also add technical knowledge. We as a group are still introducing people to bitcoin. Now we have to explain lightning.

Stay safe.
-Dave

[Chlotide]

It's not only about the merchant but also the consumer.
Companies want profit and pioneering is lower on the list. If your potential buyer is not that LN savvy might not be the right time to implement it I assume.
Most need serious market research before they can pitch LN to the board/investors. That's what you get when you have some stubborn old farts with serious equity in the company.

[LoyceV]

2) Why would you want a custodial wallet?

Compared to opening my own channels, a custodial wallet is:

• easier to use
• cheaper to fund
• better connected
• easier to empty

And for small amounts, I don't mind trusting a custodial wallet.

It's going to be interesting once big exchanges start accepting LN transactions, people trust them with large amounts already.

[DaveF]

2) Why would you want a custodial wallet?

Compared to opening my own channels, a custodial wallet is:

• easier to use
• cheaper to fund
• better connected
• easier to empty

And for small amounts, I don't mind trusting a custodial wallet.

It's going to be interesting once big exchanges start accepting LN transactions, people trust them with large amounts already.

easier to use --> Yes.

cheaper to fund / better connected --> Tough to say, I can put BTC.1 into a channel between my node and a larger peer (say ln.pizza)  and then BTC.01 into a smaller one and be done. Never had any real issues spending / receiving except those that did turn out to be caused by me in the end. Not saying there were no issues, just that they all turned out to be "Dave can't read or follow directions" issues. But yeah, funding bluewallet and being done is probably easier.

easier to empty --> I use Ride the Lightning so it's just a click on a webpage.

Just my view. For a newbie it's probably easier. But, Loyce I have seen the work you do scraping data and putting stuff together it's amazing. You are manipulating a ton of data and have a ton of programming behind it. Don't tell me you can't have a Raspiblitz or mynodebtc server up and running in less then an hour. (Not counting sync time)

Stay safe.

-Dave

[LoyceV]

cheaper to fund / better connected --> Tough to say, I can put BTC.1 into a channel between my node and a larger peer (say ln.pizza)  and then BTC.01 into a smaller one and be done. Never had any real issues spending / receiving except those that did turn out to be caused by me in the end. Not saying there were no issues, just that they all turned out to be "Dave can't read or follow directions" issues. But yeah, funding bluewallet and being done is probably easier.

I haven't installed my own node yet, but was comparing my experiences between BlueWallet and Eclair. And I was thinking about much lower values than 0.1 BTC. In BlueWallet, I can fund it with a low transaction fee. When I tried that in Eclair, it didn't confirm fast enough and only confirmed once the channel closed and basically did CPFP with a much higher fee. A full node probably gives more control, but Eclair felt like I have no idea what's happening with fees.

Just my view. For a newbie it's probably easier. But, Loyce I have seen the work you do scraping data and putting stuff together it's amazing. You are manipulating a ton of data and have a ton of programming behind it. Don't tell me you can't have a Raspiblitz or mynodebtc server up and running in less then an hour. (Not counting sync time)

I can probably get it to work, but never tried. I'd do it if I have a use for it other than paying tiny LN transactions.

[Wind_FURY]

2) Why would you want a custodial wallet?

Compared to opening my own channels, a custodial wallet is:

• easier to use
• cheaper to fund
• better connected
• easier to empty

And for small amounts, I don't mind trusting a custodial wallet.

It's going to be interesting once big exchanges start accepting LN transactions, people trust them with large amounts already.

easier to use --> Yes.

cheaper to fund / better connected --> Tough to say, I can put BTC.1 into a channel between my node and a larger peer (say ln.pizza)  and then BTC.01 into a smaller one and be done. Never had any real issues spending / receiving except those that did turn out to be caused by me in the end. Not saying there were no issues, just that they all turned out to be "Dave can't read or follow directions" issues. But yeah, funding bluewallet and being done is probably easier.

Plus for non-hobbyists/ordinary users, it's very convenient. Running your own node, staking your own coins, and balancing your channels, is a specialization that requires some level of skill, not present in all Bitcoiners.

It's why I believe being a liquidity provider in Lightning should be made financially justifiable to maintain the node.

[LoyceV]

Plus for non-hobbyists/ordinary users, it's very convenient. Running your own node, staking your own coins, and balancing your channels, is a specialization that requires some level of skill, not present in all Bitcoiners.

At some point most of that should be automated from within a wallet.

It's why I believe being a liquidity provider in Lightning should be made financially justifiable to maintain the node.

That's an open market: it's already being sold as a service. Just like some custodial wallets ask a fee for deposits. I like BlueWallet because they don't ask that, but they're advertising other exchanges to earn revenue. Lightning Roulette charges 4% fee to convert on-chain Bitcoin to LN.