[c-lightning] Towards an autopilot

[Corné Plooy]

> and say a few days before payday if the capacity of those channels is low on the employeer side could close those channels in preparation for opening them again with new funds just before payday)
IMHO the preferred approach would be for the employer to do
payments-to-self to ensure sufficient employer-side funds on the
employer/employee channel, a few days before payday. Make it a bit more,
so you can still route payments over the channel without getting into
trouble; refuse routing payments that would eat away too much of the
balance. Only if that approach fails, you should consider channel
re-arrangements. This could be, for instance, if the employee is of the
"much saving, little spending" type: then you'll have to keep adding
channel capacity.

I agree to the core argument that in special cases where you anticipate
above-average-size transactions for which it is important they won't
fail due to the inability to find a route, you want to have a direct
payer/payee channel. Large-sized transactions tend to be important, and
they also are more likely to fail on Lightning. For one-off transactions
you might be better off with a non-Lightning (on-chain) transaction, but
for recurring large transactions (such as salary), manually creating a
channel (and keeping it open) makes sense.

CJP