tapd: add itest for grouped asset channel funding #987
Conversation
guggero
force-pushed
the
group-key-support
branch
2 times, most recently
from
bf764ca to
3c5f6cb
Compare
guggero
force-pushed
the
group-key-support
branch
from
3c5f6cb to
4c38293
Compare
guggero
force-pushed
the
group-key-support
branch
2 times, most recently
from
84bf0f8 to
d30865e
Compare
guggero
force-pushed
the
group-key-support
branch
from
d30865e to
103e485
Compare
|
Updated |
|
@GeorgeTsagk: review reminder |
guggero
force-pushed
the
group-key-support
branch
from
103e485 to
e159f66
Compare
guggero
force-pushed
the
group-key-support
branch
from
684cb4e to
b14f20b
Compare
|
Ready for review, @ffranr and @GeorgeTsagk. |
guggero
force-pushed
the
group-key-support
branch
from
b14f20b to
85ae741
Compare
We remove the Zane node from all tests where its only role was being the universe server. Starting a node takes multiple seconds, so if we can spin up fewer nodes, we can save some time. And functionality wise Charlie can easily be the universe, we just need to define its port upfront so we can configure it to be its own proof courier.
guggero
force-pushed
the
group-key-support
branch
from
85ae741 to
dedf12c
Compare
guggero
force-pushed
the
group-key-support
branch
3 times, most recently
from
2ce063a to
23f300d
Compare
There was a problem hiding this comment.
LGTM!
Non-blocking: Since the group keys on taprpc are already merged, we could enhance testCustomChannelsLiquidityEdgeCasesCore with an extra option similar to groupMode, which would control whether we open single asset ID or group funded channels. This way we could incorporate low-hanging coverage for this PR. Not sure what the diff for that would be, could def be a follow-up PR.
This commit prepares some of the helper functions to be able to handle channels that have multiple asset pieces in them.
We'll be adding a new HTLC force close test where we'll be able to re-use this functionality.
guggero
force-pushed
the
group-key-support
branch
from
23f300d to
21e38f3
Compare
|
Addressed comments. |
| // We use Charlie as the proof courier. But in order for Charlie to also | ||
| // use itself, we need to define its port upfront. | ||
| charliePort := port.NextAvailablePort() |
There was a problem hiding this comment.
Maybe here and in the git commit message (or next to the topology map) we can mention that charlie is effectively a remote universe server for Erin/Fabia. And maybe the same for the other tests.
| // Then we burn everything but a single asset piece. | ||
| assetID1 := mintedAssets[0].AssetGenesis.AssetId | ||
| assetID2 := mintedAssets[1].AssetGenesis.AssetId | ||
| burnAmount1 := mintedAssets[0].Amount - charlieFundingAmount/2 - | ||
| erinFundingAmount/2 - 1 | ||
| _, err := charlieTap.BurnAsset(ctx, &taprpc.BurnAssetRequest{ | ||
| Asset: &taprpc.BurnAssetRequest_AssetId{ | ||
| AssetId: assetID1, | ||
| }, | ||
| AmountToBurn: burnAmount1, | ||
| ConfirmationText: taprootassets.AssetBurnConfirmationText, | ||
| }) | ||
| require.NoError(t.t, err) | ||
|
|
||
| mineBlocks(t, net, 1, 1) | ||
|
|
||
| burnAmount2 := mintedAssets[1].Amount - charlieFundingAmount/2 - | ||
| erinFundingAmount/2 - 1 | ||
| _, err = charlieTap.BurnAsset(ctx, &taprpc.BurnAssetRequest{ | ||
| Asset: &taprpc.BurnAssetRequest_AssetId{ | ||
| AssetId: assetID2, | ||
| }, | ||
| AmountToBurn: burnAmount2, | ||
| ConfirmationText: taprootassets.AssetBurnConfirmationText, | ||
| }) | ||
| require.NoError(t.t, err) | ||
|
|
||
| mineBlocks(t, net, 1, 1) |
There was a problem hiding this comment.
Why are we doing this. Please consider adding a bit more to the comment.
There was a problem hiding this comment.
I think we burn all but a small piece as we don't support fully burning an asset in a single transaction.
Depends on lightninglabs/taproot-assets#1478.