Closed njgheorghita closed 5 years ago
@pipermerriam Getting to the point where that revamped populus you were talking about would be incredibly useful for testing that resolving link_dependencies from nested packages is implemented correctly. I'm thinking about coming up to boulder wed/thurs if you have some time either day to pair on it?
Lets try to do Wednesday and Thursday as I think there's more than a single day of work hiding here.
@pipermerriam Works for me, i'll plan on 930ish @ galvanize wed/thurs
Is this still valid, should I review?
@pipermerriam it's waiting on releasing pytest-ethereum
to pypi - i'll ping you once that happens
@pipermerriam pytest-ethereum
is only python>= 3.6
- it also seems as though web3 is moving on from 3.5 - is it a fair time to deprecate 3.5 in py-ethpm
?
and this pr's ready for a review whenever you get the chance
is it a fair time to deprecate 3.5 in py-ethpm?
Yes
requires https://github.com/ethereum/pytest-ethereum/pull/5 to be merged and released before passing CI
What was wrong?
Validation of linked references was needed, along with validation of tx_receipt data if present in deployment data
renamed:
runtime_link_refs
=>linked_references
deployment_link_refs
=>unlinked_references
How was it fixed?
Added validation that for deployments with link-dependencies, their on-chain bytecode references the same instances specified in their manifest.
Cute Animal Picture
fixes #110