I agree. Negative shadow prices are incredibly important for optimality of
constrained network markets where flows in opposite directions cancel (as is
the case with lightning). See for example FTRs. It’s unclear to me how well
the analogy holds, but it’s worth considering.
—Will
On Tue, Jan 16, 2018 at 3:32 PM, Benjamin Mord <b...@mord.io> wrote:
> Thanks. It sounds like it was dropped due to difficulty in the routing
> protocol. Is that difficulty documented somewhere I can review? If so, I
> might take a crack at a solution to it. But regardless I suggest the protocol
> should support negative fees, even if an individual routing implementation
> prefers to treat as 0 for simplicity. That should be up to the implementation
> I think, and not a protocol constraint.
_______________________________________________
Lightning-dev mailing list
Lightning-dev@lists.linuxfoundation.org
https://lists.linuxfoundation.org/mailman/listinfo/lightning-dev
