more events, AnnotatedBlock #2437
Conversation
|
What's the logic behind |
|
theres discussion here IntersectMBO/cardano-db-sync#668 and AnnotatedBlock seemed relatively uncontroversial. I think it could probably be an event or something so we don't need to do all of this plumbing again |
|
Yep, having this data as an event makes lots of sense. I'd rather avoid the current implementation. |
goolord
force-pushed
the
db-sync-wishlist-2
branch
2 times, most recently
from
558b335 to
a891924
Compare
goolord
force-pushed
the
db-sync-wishlist-2
branch
3 times, most recently
from
963bd9a to
0fcdf2b
Compare
goolord
force-pushed
the
db-sync-wishlist-2
branch
from
0fcdf2b to
1a0caf2
Compare
…ecs into db-sync-wishlist-2
|
This somehow has a very large number of unrelated changes integrated into it... please can you rebase just the relevant parts onto master? |
|
oh yeah, i don't know what's up with that. i'll fix that |
goolord
force-pushed
the
db-sync-wishlist-2
branch
from
990165d to
d30b674
Compare
…ecs into db-sync-wishlist-2
| data AnnotatedTx = AnnotatedTx | ||
| { atInputSum :: !Coin, -- Sum of the tx inputs | ||
| atOutSum :: !Coin, -- Sum of the tx outputs | ||
| txFees :: !Coin -- All fields in the superset of all tx types |
There was a problem hiding this comment.
Speaking with Erik, he would also like to have the actual resolved inputs here.
| @@ -64,7 +84,8 @@ class | |||
| Environment (Core.EraRule "BBODY" era) ~ STS.BbodyEnv era, | |||
| State (Core.EraRule "BBODY" era) ~ STS.BbodyState era, | |||
| Signal (Core.EraRule "BBODY" era) ~ (Block BHeaderView era), | |||
| ToCBORGroup (TxSeq era) | |||
| ToCBORGroup (TxSeq era), | |||
| Coercible (Event (Core.EraRule "BBODY" era)) (STS.BbodyEvent era) | |||
There was a problem hiding this comment.
I appreciate this is more general, but for consistency with everything else I'd rather stick with the equality constraint.
There was a problem hiding this comment.
Some instances won't compile if this is in equality constraint. This constraint will get floated outside of typeclass. Maybe it could be an Embed constraint instead, but the Coercible constraint is not superficially more general
| abEpochSlot :: !SlotNo, -- The slot within the epoch (starts at 0 for first slot of each epoch | ||
| abTimeStamp :: !UTCTime, -- The slot number converted to UTCTime | ||
| abEpochSize :: !EpochSize, -- Number of slots in current epoch | ||
| abTxs :: [AnnotatedTx] -- All fields in the superset of all block types |
There was a problem hiding this comment.
Don't understand the comment here
There was a problem hiding this comment.
the AnnotatedTx is meant to contain fields from the TxIn/TxOut/TxBody of each i believe
| pools = _pParams $ _pstate $ _delegationState $ ledgerState | ||
| certs = Foldable.toList (getField @"certs" txBody) | ||
| ins = | ||
| Val.coin (balance @era (eval ((getField @"inputs" txBody) <| _utxo u))) |
There was a problem hiding this comment.
Doing this within the API, rather than during the appropriate ledger rule, means we end up resolving the inputs twice. The correct place for this would surely be in the Utxo rule?
There was a problem hiding this comment.
Is this expensive? I can't imagine this AnnotatedBlock would be very useful if it was split up into several events. I don't see where in Cardano.Ledger.Alonzo.Rules.Utxo.hs would be more appropriate, I feel like I'm missing something
goolord
force-pushed
the
db-sync-wishlist-2
branch
from
76dd5fc to
78d10c5
Compare
goolord
force-pushed
the
db-sync-wishlist-2
branch
from
78d10c5 to
78ef70f
Compare
|
closed in favor of #2652 |
No description provided.