Traverse 🚅

Traverse

14 February 2023 update I will be closing Traverse down on or after 21st February. Channels will be cooperatively closed at a time when on-chain fees are low.  If you have an on-going liquidity arrangement, then i'll compensate you at a rate of 100ppm for each remaining month. Please send me an invoice by 28th Feb for 100ppm x months remaining x liquidity amount. I'll process them after 28 Feb. The invoice must point to the public key of your server. Make sure our invoice has a suitably long expiry time! I opened Traverse in December to understand more about the Lightning network and discover ways to improve it. I discovered that there is a strong price gradient across the network, and the incentives for node operators don't lend themselves to reducing that price gradient. The price gradient tends to cause liquidity to be trapped in channels, and most is unlikely to take part in transactions. Indeed, under these circumstances, the liquidity is a lure for potential transaction routes, but it is available only in one direction, and overwhelmingly the wrong direction causing routing failures. I attempted sociological means to encourage node operators to pass liquidity through the network to improve the whole system. This met with only limited success. I attempted to reach out to LND developers. but they seem to be conflicted; offering services to supposedly reduce the problems as opposed to implementing the necessary protocol-level changes. Observations: It is not good for the network for node operators to be overly concerned with node centrality. Liquidity balance in channels is more important. Node recommenders need to re-focus on price gradients which leads to directional lack of liquidity and high prices. Node operators tend to hold on to liquidity as opposed to passing it on for the good of the network, and active encouragement is needed even in a swap agreement with such a stipulation. The protocol must incorporate a dual-funding mechanism by default, where the channel opening takes on-chain funds from both ends. This can certainly be implemented by requiring submarine swaps on channel openings. Other, better methods may exist. This will flatten the price gradient across the network and make channel opening more fair. A unidirectional channel balanced via routing half the channel opening funds through the network is worse than useless to the network. Price increases in channels block them for hours due to the lethargic propagation of  fee information.  LN has great potential. LN developers providing potentially profitable services as a band-aid for the shortcomings of the network protocol hold back the development of the network. Such services form a conflict of interest. I may come back to LN to contribute ideas and try to fix these problems, but having coins in a hot wallet mostly stuck tight in the ends of liquidity channels doesn't compensate for the attendant risk of keeping coin in a hot wallet, so it would be on a much smaller scale. Thank you friends!

Hubness Rank: 922

Weighted for channel sizes: 602

Lower numbers are better. Measures influence of a given node in the network. Better ranks imply a well-connected node that is linked to other well-connected nodes.

Hopness Rank: 970

Weighted for channel sizes: 957

Lower numbers are better. Measures how many hops it takes to reach any node on the network. The better the rank, the fewer the hops are required to reach other nodes.

Betweenness Rank: 1660

Weighted for channel sizes: 1499

Lower numbers are better. Measures how often this node falls on the shortest path between other nodes. The better the rank, the more likely the node will route a payments.

Metrics by LN node insight