pull down to refresh

0 sats \ 6 replies \ @k00b 9h \ on: Stacker Saloon
Currently exploring alternatives given #1317180 and it sucks. Most probing abstractions - either via LND's api or ones provided in the libraries we use - suffer from this not-a-bug-bug. I have to write my own probing logic now, which has its advantages, but it's not how I want to spend my time.
write
preview
0 sats \ 4 replies \ @k00b 9h
As another alternative, I can just not probe I guess but I don't like inconveniencing senders if I don't have to.
reply
1050 sats \ 3 replies \ @SimpleStacker 8h
what happens if you don't probe?
edit: what a weird sentence
reply
1000 sats \ 1 reply \ @k00b 3h
You should be receiving sats for zaps again.
I removed probing. In the meantime we're using LND's route prediction stuff instead which doesn't suffer from this problem.
reply
0 sats \ 0 replies \ @SimpleStacker 3h
cool, thanks! hope the lack of probing doesn't cause too many issues
reply
0 sats \ 0 replies \ @k00b 8h
lol
In the case of there being no route from SN's node to the receiver's, rather than a probe failing, the sender's payment will fail.
The probe is helping us predict the weather. Absent a weather report, we risk the sender stepping out into the rain.
reply
0 sats \ 0 replies \ @k00b 6h
After implementing a custom probe, it would appear that the probe causes the route to be marked as invalid temporarily. Merely attempting the probe, by itself, causes the payment to not succeed.
Yikes. I guess we'll just stop doing probes in the meantime and rely on mission control's guesses.
reply