I have recently stumbled in a few tweets regarding the Lightning Network.

I am always been a LN fan, and I have a long history of reporting on various metrics, and I would like to start again.

First, an overview of the network so far:



The brown has been exponential, considering also that BTC has increased in value.

Looking at the LN capacity we see that much of the grow has been registered on the last few weeks. Growth has gone parabolic:



If we look at the following graph we have a detail on the balances added during the last months:



Of course, adding 250 BTC of capacity when BTC is at 30k is totally a different story than adding 300 BTC capacity when BTCUSD trades at 8,000 as in March 2019.

So, pretty a neat increase in capacity. Just wondering if the real usage of such capacity is gaining traction. I guess the “low” fees in the main chain are not pushing adoption, but it’s important to invest in this invisible assets.

I will try do add news, reports and commentaries about LN to this thread as soon as they are available or worth mentioning.
Glad if you help!

---

Resources on Lighting Network:

Bitcointalks Threads:

• The Lightning Network FAQ
• Basics of the Lightning Network
• Lightning Network -- Basics (documents, data pages, files, plots) by tranthidung

Dashboards:

• txstats.com
• BitcoinVisuals.com

External Resources:

• Jameson Lopp's LIGHTNING NETWORK RESOURCES
• Lightning Network Stores

Absolutely.
My intention on this thread is not to provide a technical reference, or a reference directory. Rather I would like this to become a nontechnical, casual talk thread where to discuss LN and his development.

I am adding your thread, together with some other, in the OP.

total capacity is meaningless, much like bitcoins market cap is meaningless
as it doesnt show actual utility

EG. i have no intent on spending anything for next 6 months. but i could. if i wanted to bloat up the total capacity by multiples of current total capacity. by just locking alot of my coin to LN and doing absolutely nothing with it

..
a better metric to follow is the average capacity per channel
because then you can see the potential liquidity amount of what payment routes could achieve
(EG the success rate of payments)
50% of network only has $200 liquidity per channel
(success rate is low for weekly rent/grocery payments but ok for coffee and gum)
 
in short
it doesnt matter if the network say its has 1 million bitcoin.... because payments are only a regular success for chewing cum amounts

this is beautiful - reminds me of the pictures comparing the internet in the beginning and how it looks now

well 50% of all channels had a max capacity of ~$200
so even without crying about exact stats of how many micropayents..
you know for sure that there were bottlenecks of anyone trying to spend more then $200 in one go
and if say average was $2 spend per payment. 50% would have had to close and re-org their funds after just 100 payments

think of it this way

Alice [$4000><$200] bob [$200><$200] chuck [$200><$200] Dave [$200><$200]starbucks
alice might be in the top 10% with $4k... but do you think she will ever get to spend her $4k without closing

imagine scenarios.
dave has made 10 payments for himself of $4 each for HIS starbucks
Alice [$4000><$200] bob [$200><$200] chuck [$200><$200] Dave [$160><$240]starbucks
chuck done the same
Alice [$4000><$200] bob [$200><$200] chuck [$160><$240] Dave [$120><$280]starbucks
bob done the same
Alice [$4000><$200] bob [$160><$240] chuck [$120><$280] Dave [$80><$320]starbucks

now how much can alice spend if she was a caffiene addict
trust me.. its not $4k .. its not $200..
ill give you a hint.. its only $80

some might say $80 is still good
but now your forgetting the reality that
bob doesnt just have alice and chuck as partners. 'others' may use bob to route to starbucks. thus diluting the dave outbound to starbucks even more

chuck doesnt just have bob and dave as partners. 'others' may use chuck to route to starbucks. thus diluting the dave outbound starbucks even more

..
point being if you know that 50% of the network only has $200 capacity..
then you can work out that alot of outbounds are going to get exhausted very quickly
(average $2 payment = exhausted in 100 payments)
(average $4 payment = exhausted in 50 payments)
(average $8 payment = exhausted in 25 payments)
(average $20 payment= exhausted in 10 payments)

you can also work out if you know it will cost about $3 lock-in and $3 unlock of a onchain fee..
thats $6 of the $200 the people will want to recoup even to allow their balance to be used by others
so you can work out that as a minimum break even (logic/common sense) they are going to charge 3% fee
($6 of $200=3%)

heck.. you can then do some math. if you know the average is 3 hops to the destination
you can work out that it ends up costing the spender 9% (3% per hop)
so you start to get a picture about how expensive it is within LN compared to just buying coffee onchain vs buying a single $80 giftcard via fiat

so yea there is alot you can learn from knowing what the majority of the networks channel liquidity is. even without knowing each precise payment

heck if you know that some 'lightning pizza' promotion is being done for $25 of pizza
you know that the network will start showing massive amounts of payment failures within 8 payments along routes
('fold' lightning pizza' experiment had 90% payment failure rate)
https://decrypt.co/5321/folds-bitcoin-lightning-pizza-turns-out-to-have-been-a-success

bank notes made gold viable(the promotion)...... until people didnt want to withdraw their gold out and instead opted to swap their bank notes for nickel and brass coins
'coz gold too heavy to manage'
'coz gold too expensive to withdraw with'
'coz convenient'

see the pattern
when LN devs are literally screaming it in your face
'bitcoin onchain fees expensive'
'atomic swap to altcoin'

LN does not make bitcoin viable for average joe. it makes average joe viable for altcoins, while custodians take the btc

Interesting point made to a user answering this tweet:



https://twitter.com/kerooke/status/1417961486253764609

This is the usual update by @kerooke about network growth.

An user pointed out a few interesting points, corroborated by some numbers:


https://twitter.com/pablogallegos1/status/1418202181803544581

Is centralisation necessarily a problem?

Ah! Not at all, this experiment would be pronte to the “Chamberlain argument”!

I know the subject has been debated many times. But still there is a lot of issues on the matter.

I think it’s very useful if a payment system is centralised/decentralised/peer-to-peer. Each of those network structures has a set of features, there might not exists an absolute winner.
But honestly recognising the status of the network could be the first step for an analysis.

Having said that:

• After 4 years from launch the 90% of the liquidity is provided by 1% of the nodes. I hardly see any explaination of why this should be the product of free market, rather than a sort of accident, whose causes and effect I ignore.
• Given the above structure, and the stability of the capacity of the median channel, I struggle thinking LN is gaining traction. Something I think it would be a net positive effect.
• I don’t think this centralisation is a problem. It is very similar to the hash power being concentrated at 90% in China: did I like it? No, of course. Was it a problem? Not at all: when China “banned bitcoin” mothers network readjusted by itself. Something probably would happen if the 1% of node would drain their liquidity overnight.
• Which is the best measure for network decentralisation? Capacity? Network usage? Transaction numbers? Which of those metrics can help us assess the LN success and how reliable are them?

For now it's more of a concern.  It's all about potential.  Centralised points of failure aren't generally perceived as a problem until they fail.  So it's better if something could be done to encourage a wider distribution of the load.  But what that "something" might be, I sadly have no idea.  Hopefully it's something that happens organically through market forces and competition, but there's no way to tell right now.

Something things do have to go wrong before people adjust their behaviour.  When mistakes happen, some people learn right away and change their habits, while others don't.  But gradually, over time, there are fewer central points of failure and the whole thing becomes more resilient.  A similar situation could arise here.  If any of these major nodes have failures, people may be more likely to use smaller nodes in future.

All I can suggest is keep a close eye on it, see what develops, then make contingency plans if it looks like it's getting worse.

yep this option
lots of services offering "free inbound balance" using 'stake' they locked months-years ago where the 'stake' is not locked to any onchain multisig funding tx between the service and the temporary channel user

offering temporary(30 day) and instant(no confirmed) inbound balance:- without any onchain activity needed
(there are many security flaws with this method, but i wont rant.. this time)