Not sure what's happening, but I'm not able to make it work. The logs are different, but it seems to still be complaining about msats over sats.
What's the wallet source? This error means lnbits is returning an invoice for an amount larger than we requested. In this case, it's a lot more than we requested.
reply
151 sats \ 10 replies \ @fed 17 Aug
Lnbits using PhoenixD. Here is the invoice it is giving.
reply
It looks like it's charging SN to open a channel, but it might also be poor support for msats in lnbits. We are trying to pay you 2,762 sats and it wants us to pay 27,878 sats.
reply
0 sats \ 7 replies \ @fed 17 Aug
Shouldn't be a channel charge, that phoenixd instance got plenty inbound liquidity. Weird, it's like the invoice further multiplies the sats by a thousand before charging.
reply
It looks like lnbits doesn't support msats: https://github.com/lnbits/lnbits/issues/2009
reply
100 sats \ 1 reply \ @ek 17 Aug
The price needs to go higher for enough people to care about msats, lol
reply
I know, but the lightning specs are all in msats, so all these non-msat APIs will have to be rewritten eventually, and it's weird to have APIs acting like they're receiving direct user input.
reply
I've submitted a fix that doesn't use msats with lnbits. Let me know if that finally does the trick.
reply
1057 sats \ 1 reply \ @fed 18 Aug
I can confirm, auto withdrawal worked now. Thanks!
reply
Thanks for your patience/help troubleshooting!
reply
It'd be a pretty obvious msats/sats confusion if it were 1000x as much, but maybe it's doing something else funky. It works fine in our testing with lnd, but maybe phoenixd doesn't like msats.
I'll review this deeply tomorrow and see if there's something we can do.
reply
10 sats \ 0 replies \ @fed 17 Aug
deleted by author
reply