#zerobasefee — Public Fediverse posts

@praetorrian @Peaceandprosperity @adam There are a lot of nodes that are doing #zerobasefee now. Podcast Index's node has zero fee channels with some other nodes we trust.

The issue you run into is that the lightning routing algo will always prioritize lowest fees. So, as soon as you drop to zero on a certain channel, the whole world wants to use you for routing and it drains your liquidity fast.

Running #zerobasefee #dropthebase
https://basefee.ln.rene-pickhardt.de/
@renepickhardt @_einundzwanzig_ @HonigdachsPod
---
RT @aantonop
lncli updatechanpolicy --base_fee_msat 0

#zerobasefee
#dropthebase

https://basefee.ln.rene-pickhardt.de/
https://twitter.com/aantonop/status/1425907445717774340

The concern remains with #zeroBaseFee:
to never route free (not even the 1 msat shards) the minimum feerate setting would be 1 000 000 ppm which is a 100% fee.
If it is not applied a payer with free CPU and and disk IO capacity would be incentivised to flood with msats.

Correction:
Watch the "msat" in the OP.
The "rounding down" only applies to the sub-millisat earnings. HTLC and routing fees are accounted in msats.
This means with #zerobasefee at 100 ppm routing 9 sats (or 9999 msats) would be free.
At 1 ppm this would be 999 sats.

@harding Yes, unfortunately the base fee is the floor, the presence of which does not work with their calculations.
As a payer it would be better to get rid of a variable, but I felt extremely annoyed by seeing hundreds of 0 fee forwards through my node today, hours after dropping the base fee.
#zerobasefee incentivises every payer to break up the amounts to many small pieces.

@openoms I don't think Pickhardt's main claim is simpler routing; I think it's much more effective routing---i.e. fewer retries when sending significant amounts, leading to faster payments and fewer failures (all other things being equal). If it's as big a deal as he claims, I think it's probably worth considering making other changes to LN policies to allow it to work.

That said, I had a similar concern when I first heard he needed a #zerobasefee. There needs to be some sort of floor.

The research is great but after seeing this action made my mind up quickly: https://twitter.com/openoms/status/1416089044967215106?s=19
A simpler model at the cost of routing incentives does not worth it.
---
RT @openoms
If you have #zeroBaseFee and 100 ppm feerate still routing 9999 sat payments for free.
Where you would count on 100ppm your node charges nothing for 100*9999.
With 1 sat basefee you would get exactly 100 sats both when rout…
https://twitter.com/openoms/status/1416089044967215106

If you have #zeroBaseFee and 100 ppm feerate still routing 9999 sat payments for free.
Where you would count on 100ppm your node charges nothing for 100*9999.
With 1 sat basefee you would get exactly 100 sats both when routing 1*999900 or 100*9999.
#runTheNumbers #lightning ⚡
---
RT @alexbosworth
LN routing policy rounds down percentage fees and the lowest ppm is 1, so sometimes you can get totally free routing just by using multi-pat…
https://twitter.com/alexbosworth/status/1415774383298121731

Congrats for this @renepickhardt and @stefanwouldgo and thank you for publishing for all our benefit!
I am aboard the #zeroBaseFee ship.
Make ⚡ pathfinding easier and optimise for more successful payments!
---
RT @renepickhardt
Optimally Reliable & Cheap Payment Flows on the Lightning Network

the joint work with @stefanwouldgo is now (as promised) in the #open: https://arxiv.org/abs/2107.05322

This method should allow us to
⚡️ se…
https://twitter.com/renepickhardt/status/1414895844889960450