We were automatically falling back to bolt12 decoding, clobbering the
fail message. Ultimately resulting in confusing error
messages (expected prefix lni but got lnbtrc). Now we first determine
which decoding we're trying to do, and then only decode accordingly.
Changelog-Fixed: pay: Report the correct decoding error if bolt11 parsing fails.
We were not aborting if we had routehints, even though all routehints
may have been filtered out.
Changelog-Fixed: pay: `pay` will now abort early if the destination is not reachable directly nor via routehints.
We would happily spin on attempts that are doomed to fail because we
don't know the entrypoint. Next up: remove routehints whose
entrypoints are known but unreachable.
The main responsibility of this new function is to mark a payment
process as terminated and set a reasonable error message, that will be
displayed to the caller. We also skip the remaining modifiers since
they might end up clobbering the message.
As pointed out by @cfromknecht [1] there was no formal standardization of
the featurebit, and lnd would try a keysend whenever TLV was supported
by the recipient. This mimics that behavior by checking only that TLV
is enabled.
Changelog-Fixes: keysend: We now attempt to send with keysend even if the node hasn't explicitly opted in by setting a featurebit.
[1] https://github.com/ElementsProject/lightning/issues/4299#issuecomment-781606865
The semantics don't change, since `lightningd` will use false as
default as well, however setting it to something other than `None`
causes the RPC library to include the parameter in the query, and
since the parameter was introduced only in 0.9.3 and pyln may be used
with older versions this then results in an error about an unknown
parameter.
Setting this to `None` makes sure pyln filters out the argument before
calling.
Changelog-Fixed: pyln: Fixed an error when calling `listfunds` with an older c-lightning version causing an error about an unknown `spent` parameter
We were reporting the failure immediately but still continuing with
the startup. This could happen if an important plugin ends up in a
race with another plugin (important or not) for a contended
resource (CLI option or RPC method name). We would eventually notice
that we were supposed to abort, but at that point we already processed
a couple of blocks, loaded the entire state, etc.
This just aborts early with a sane error message.
Changelog-Added: plugin: If there is a misconfiguration with important plugins we now abort early with a more descriptive error message.
Reported-by: PsySc0rpi0n
Reported-by: Ján Sáreník <@jsarenik>
I had way too much fun with this and got a bit carried away with the
letter writing. The idea is to be helpful when users start the plugin
from the command line, rather than run it under the control of
lightningd. We also print detailed information about the user-visible
things such as the methods and options exposed by the plugin.
Changelog-Added: pyln: Plugins that are run from the command line print helpful information on how to configure c-lightning to include them and print metadata about what RPC methods and options are exposed.
Suggested-by: Rusty Russell <@rustyrussell>
It's been causing me quite some headache, and I don't see the point in
jumping through the hoops for something that can be trivially fixed by
having the required build tools.
The receiving node: ...
- MUST fail the channel if:
- the `witness_stack` weight lowers the effective `feerate`
below the agreed upon transaction `feerate`
> If the peer's revocation basepoint is unknown (e.g. `open_channel2`),
> a temporary `channel_id` should be found by using a zeroed out basepoint
> for the unknown peer.
We consolidate to the latest/singular RFC patch for dual-funding, so
there's just a single patchfile for the change. Plus we move back to the
opener setting the desired feerate, the accepter merely declines to
participate if they disagree with the set rate.
Just a security measure to avoid alternative use-cases of the hsmd
running into the issue that they need to send a `WIRE_HSMD_INIT`
message as first message. If that is not done, the `secretstuff` won't
get initialized and we'd be producing signatures from uninitialized
memory, which are completely useless.
Changelog-None: Internal change only
Looks like #4394 treated a symptom but not the root cause. We were
actually sending the message framed with the WIRE_CUSTOMMSG_OUT and
the length prefix over the encrypted connection to the peer. It just
happened to be a valid custommsg...
This fixes the issue, and this time I made sure we actually send the
raw message over the wire. However for backward compatibility we
needed to imitate the faulty behavior which is 90% of this patch :-)
Changelog-Fixed: plugin: `dev-sendcustommsg` included the type and length prefix when sending a message.
We were getting a memleak error that the open_attempt isnt' being
cleaned up in test_rbf_reconnect_tx_construct. I had some trouble
reproducing it, so I removed the reliance on using `tmpctx` to clean it
up and was more surgical about cleaning it up inline.
