Closed gitmachtl closed 2 years ago
Probably mistake in well known info...
https://github.com/txpipe/oura/blob/d9dc5421f2237464531a7c1759cb86db0afdd5c0/src/framework.rs#L33
Its also off 19 seconds on mainnet.
Event timestamps are inferred from the slot number of the block, but there seems to be some ambiguity on how to translate the interval length between blocks (~20 secs ATM). I found this comment by @KtorZ very insightful:
https://github.com/CardanoSolutions/ogmios/issues/3#issuecomment-619786434
TBH, I'm not happy with the current approach in Oura. I'll do some reverse engineering on db-sync to understand the logic it uses.
Testnet Example: The output for block
3226895
is reported with the timestamp1641849006
. If i convert this back into an actual date likedate --date="@1641849006"
i get the dateMon, 10 Jan 2022 22:10:06
. DB-Sync/Explorers reportMon, 10 Jan 2022 22:10:25
So there is a 19 seconds offset. And its for all the reported blocks. So either i am converting it wrong, or there is a Bug?