Topic: The Lightning Network FAQ - page 57. (Read 33452 times)

If I show you the channel he opened with me, can you tell if all the conditions are met or not?

There is an unspendable channel reserve. You need to spend about 2-3% of the channel's capacity before you will be able to receive anything. Theoretically, you should be able to receive slightly less than 30k sats through friend's B second channel since all funds are on his side of the channel at the moment. There might be some problem with friend's B channels. He might not have enough inbound capacity because of the reserve.

Hello once more.

As I'm new to the Lightning network, I need some help to clear a few questions I have about opening channels and inbound capacity.

So, the scenario is this:

I have a c-lightning node running on my laptop and I have Spark-Wallet installed and connected to my node.
I have Bluewallet installed in my smartphone.
I have opened 2 channels with 2 friends of mine (friend A and friend B) of 20k sats (for the sake of experimenting)
Then I sent 100 sats to friend A.
After that I sent 1000 sats to my Bluewallet. I know this transaction was routed through friend B node.
Then friend B opened a channel with me with 30k sats.
Now I'm trying to send sats from my Bluewallet to my Spark-Wallet, but no matter the value I choose, I always get an error of inbound capacity not available!

Shouldn't I be able to receive up to 100 sats from friend A node and up to 1k sats through friend B node?

What am I missing and what is that reserved value I know it gets kinda locked when channels are opened?

Not necessarily, the Eclair wallet provides for buying inbound capacity.  I think the Zap wallet has a similar function.

If part of that purchase was (example) you paid for an inbound 0.05 channel and part of that 0.05 was sent to you (say 0.04) your cost might be (random numbers out of my head) say 10% on top of the 0.04 plus the TX fee to open the channel - rounded up to 0.005.  In the hypothetical example you might pay (in advance) 0.045 and get a 0.05 channel opened to you with 0.01 inbound plus 0.04 outbound capacity which would no doubt be a healthy start to any new node.

It's also a matter of perception of use.
I have no issues opening 2 very large channels to some providers and a bunch of smaller ones to other places. As I stated elsewhere I was actually using 1 node to connect to the big peers and the another to connect to the small ones and then connecting to the node with the smaller ones.

And as d5000 said you can manage fees by opening channels at certain times. But if I do keep a bunch of channels open AND fees stay high at some point I am going to empty the smaller channels and only deal with larger liquidity providers. That is going to be the issue.
Because, yeah I can say open this with 1 or 2 sat fee because I know it will open overnight on a weekend. But at some point who wants to plat that game.

-Dave

Yes, there is a thin ledge between "fees too high for LN" and "fees too low for LN".

However, the good thing is that opening a LN channel is not time-critical, so people can wait for it for moments with low fees. Let's say the fees for a instant confirmation rise to $50 at weekdays at daytime, but go down to $5-10 at night (western hemisphere). Then, observing the current mempool data*, at weekends it's likely that they go under $1 and cheap channels can be opened. And once LN adoption is boosted and many services using it, pressure on the on-chain fees should become lower.

But yes, I guess that if LN adoption depends only from fees, then it can fail, as fees could rise too fast and opening LN channels would be too expensive, like you wrote. So I guess there are additional efforts needed. Service providers, above all wallets and exchanges, should really realize that if they support LN now when fees are still low, they also contribute to a healthier and more sustainable Bitcoin ecosystem.

*For example, according to Johoe, in the week between October 5 and October 11, fees were relatively high, with instant-confirmation rising over 100 sat/byte in several occasions. However, even in this situation at the daily minimum (night in the western hemisphere) they could always be confirmed for 8 sat/byte or even lower, and at Sunday 11, 1 sat/byte again was enough. So there were 2 (decimal) orders of magnitude between the minimum instant-conf fees and the maximum instant-conf fees.

I'm afraid that's not going to happen. If fees reach 1000 sat/byte again, opening and closing a LN channel is going to cost $25 or (much) more. That won't be worth is, just like I couldn't use Bitcoin for normal transactions at the end of 2017.
Unless you can use an exchange to buy Lightning Bitcoin directly so they open large channels instead of having users open small channels, but that means you can only use custodial wallets.

For me, there are mainly two conditions for the Lightning Network to really take off:

1) Providers of massively used services must support it.
2) It must be clear that transaction costs using LN are much smaller than without it.

Condition 1 is a chicken-and-egg problem. If there are many users, there will be many services, and vice versa. Ideally, some service providers could create something like a "beta area", where people can use LN to deposit and withdraw coins at their own risk but keeping the wallets of on-chain and LN activity strictly separated to minimize exposure to threats related to the alpha/beta state of LN software. But as it was discussed some pages ago this will probably only happen on a small scale.

So it is possible that a LN "takeoff" is more likely to come from a new increase of transaction fees due to fuller blocks (condition 2). This would make it again attractive for service providers to jump on the LN boat.

Obviously usability is also a factor, and thus I consider the Electrum support for LN a major step forward, even if its still limited. It may have contributed to the new records set in September/October for LN capacity.

BTW: I was looking for a good, updated list of LN services. A list of lightning services/stores seem to be at Lightning Network Stores, but it is possible that it is outdated as all wallet providers are from 2018 or 2019. Is there a better list anywhere?

I admire the optimism, but I don't know how an ordinary user will be expected to be in it long term, unless he/she is incentivized from the level of specialization in maintaining a Lightning node. But I see a debate for altruism to boot-strap the network.

I can choose from many different wallets, pay a growing number of services and receive payments in seconds. LN is doing just fine.

You ask a lot of questions on topics you abandon afterwards. Why's that?

Would it be more mature from our community and from the developers of Lightning Network to state that it has failed? It's fine, not everything needs to be a success in this world. The 2% isn't happening.

Yes, here you're totally right, security wise it is more risky. But most will choose to not use the funds that they would hold alternatively in a cold wallet, but a relatively small part. What they normally would hold in a hot wallet anyway. And they can use part of it for transactions.

If only 2% of all Bitcoin hodlers (estimated at 25 million worldwide, e.g. here) would participate with 2% of their funds, LN liquidity would see a massive boost already. Currently we have only about 33-35K LN nodes and ~1100 BTC capacity. The "2% of 2% scenario" would make LN grow to 500K nodes and to a capacity of 0.02 * 0.02 * 20 million = 8000 BTC.

Agree, as it's actually a virtuous circle: the more participants/services join, the lesser the opportunistic cost. The lesser the opportunistic cost, the more new people join.
Not even mentioning protocol upgrades that help this move forward (eg splicing out).

Strongly disagree ! You are leaving your coins in a hot wallet on alpha software (and protocol!) here.
So it *is* different, just a trade-off (that a lot of us choose to make, at least for part of their funds).

I don't really see big opportunity costs for Bitcoin hodlers if they provide liquidity for Lightning, because you don't "lock" the capital really, but more "freeze" it. Because you can always close the channel, and you are also not forced to be online 24/7 (although it would be ideal, obviously, if you did). So it's not really different if you hold the BTC (improductively) in your wallet.

There is more a direct cost involved, and that is the one of transaction fees (when opening and closing a channel). But you can do the opening transaction in a period with low fees for 1 sat/byte (currently always possible on weekends and most of the days also at night in the Western hemisphere). And even if you use your BTC regularly, you should be able to know which part of your balance you only need long term, so you can use that for LN.

And I agree with Carlton Banks: long-term there is a reward, it's not only altruism. It can be future fees, a rising aggregated value of LN (which adds value to the BTC network), etc. So I think for hodlers it is rational to support LN with an own node.

there is an incentive. It's more long-term, but the reward is bigger.

The people that provide liquidity to those who need it most will be rewarded. Those who do not will need to hope they get lucky, and that a dissimilarly self-less person provides to them some incoming liquidity.

Leading on from that topic, does electrum lightning allow capacity to be added (it'd not just lnd) I found a wire in the back of my router I was going to use to put a node on a pi, I could instead open 30 channels with 0.003 btc on each capacity wise to help fund the network a bit more and I think it's a pretty cheap way to do it as well... (this is all dependent on whether I remembered to bring a hdmi cable with me as lockdown will make it difficult for me to source one I think).

Altruism. But reality is, without incentives, Bitcoiner's can't be expected to take/accept the opportunity cost to lock their capital, Bitcoins, in Lightning channels.

Which leads to more programming / work on the time of the exchange and more places for errors / vulnerabilities to come in.



It really becomes something of a loop.
If I want to use my node a lot I have to connect to the larger more connected nodes.

If I want to help the network I connect to smaller nodes and then have to hope that through enough hops I get to a better connected one with more funds.

So I start by connecting to ln.pizza or ln.big and then work my way down.
But then I have to worry about how much BTC I am locking away or opening channels or keeping in software that the developers themselves keep saying is beta and don't put a lot of funds in.

I do both. I have my "dave's private node" connected to some big liquidity providers and then a smaller one (not online at the moment but that's a different story) that connects to a few smaller ones and my node that is connected to the big providers. But, lets be honest how many people are doing that?

-Dave

or "everyone"

those who refuse to provide incoming liquidity to others are not helping the network (including those who charge a fee to do so, but that's my opinion and their prerogative)

we can all help the situation by identifying nodes with low or no incoming liquidity, and providing it to them by opening a channel. I realize that there's potentially more fees to be earned opening channels to nodes with higher routing throughput, but ultimately, a better connected network is stronger (and more profitable in all aspects) than a centralized network. There will doubtless be a limit to how may channels it's useful for a single routing node to have, although I am unsure what that limit is (it's likely related to the amount of traffic on the network considered over a long timeframe, as well as how much BTC a node has to put into the network).

I know. I was pointing out companies like Lightning Labs, as they are "the company", or one of the companies behind Lightning Network development. Plus Blockstream too. What happened to Blockstream and Lightning?