i agree, well done
polishing a website is an endless swamp, but the easy 80% is pretty much done
I have been playing with lightning network
How's your experience?
I found `lnd` to be significantly more complex than your average Bitcoin full node, especially to program against. And I failed to make any significant amount of money through routing fees.
That said, it's still cool.
yeah, seems to be fairly advanced
I might look into integrating snowblossom with lightning. lnd is designed for multiple coins already.
It would be a lot of work, snowblossom does things rather differently than ltc and btc
We would have to make a go library of key functions
I think the cross-chain overlay possibilities is one of the coolest things about lightning.
Ah, another port.
And we're missing relative locktimes, right?
that can be added. not saying it would be trivial integration.
We have txout lock times based on clock time and block number
True, that's SIP3 right?
probably need tx level as well
Not having script is kind of elegant in its own way.
Yeah, I think so.
but I had always planned on adding that eventually
I wouldn't rush, the lightning network is still in an early phase really.
true
Although I'm curious to see how it'll pan out if it's really used as a "scaling solution". Can't get over the fact that the game theory is... weird. From a miner perspective.
I might do some load tests on it for my own entertainment
hah, who doesn't love load tests
yeah, it is like a PoS layer on top of a PoW coin
and if lightning every gets really big (like 1 million nodes) the ring routing stuff I've done for snow channels would actually solve routing there
without each node having to have an idea of all the channels
I should write that up more science like and send it out for people to throw rocks at
The DHT concept?
I'm still trying to wrap my head around the ins and outs of lightning routing, to be honest
it's hard to find neutral information too, since people are quite fanatically pro- or anti- LN
yeah, with the DHT each node maintains links across the ring based on its own position
lower density far from self, higher density closer
so to route, you pick the neighbor closest to target and use that
and that neighbor will have higher density links in the area of the target, since it is closer to it
that seems superior to dijkstra's algo, at least on a large scale
plus DHT's have proven quite resilient at these scales already
curious to see how that'll work out in practice, though, some clever ideas for sure
it would have some similiar issues to routing in Freenet I suppose
and generic data streams don't deal with some of the issues payment channels do, e.g. capacity
LN routing for anything that isn't small change is still pretty dramatic
https://github.com/dianerey/lnsim hm, this is pretty nifty Lightning Network Simulator
Any update on the release date for the Android wallet?
Is there an ios wallet in the works?
You can try the beta version of the android wallet now if you like
No IOS in the works