Closed alexkroeger closed 4 years ago
@nadavhollander @graemecode the PR is failing tests, but I think the list that is being checked against is not comprehensive.
An example user that failed the test is 0x094b1aa08a107340baa421ba6013dbda722d71ed
. However, we can see in this transaction the user is receiving LINK from Uniswap. The issue might be the list of addresses doesn't account for meta-transactions submitted on behalf of the user.
@nadavhollander @graemecode the PR is failing tests, but I think the list that is being checked against is not comprehensive.
An example user that failed the test is
0x094b1aa08a107340baa421ba6013dbda722d71ed
. However, we can see in this transaction the user is receiving LINK from Uniswap. The issue might be the list of addresses doesn't account for meta-transactions submitted on behalf of the user.
Interesting. Yes, we might have to update the query. We're specifically using from_address
in any traces for transactions submitted to the Uniswap contracts, and do not include those addresses that are just the recipients of tokens. Will consider and update tomorrow.
Hey @alexkroeger,
We updated the bigquery SQL command used to generate the superset of accounts to include all addresses that received a token transfer as part of a transaction that touched uniswap — it includes the example user you supplied, but there are still failing addresses in the list you've supplied. One thing that just occurred to me is that this likely won't catch ether transfers — it would be helpful if you could assist in triaging on this front!
Hey @alexkroeger,
We updated the bigquery SQL command used to generate the superset of accounts to include all addresses that received a token transfer as part of a transaction that touched uniswap — it includes the example user you supplied, but there are still failing addresses in the list you've supplied. One thing that just occurred to me is that this likely won't catch ether transfers — it would be helpful if you could assist in triaging on this front!
@0age I reran the locally and it looks like there is now only 1 offending address--a contract that made a 'fill' with 0 amounts. I can exclude that and I think it should be good to go.
@0age tests passing 👍
Added users of 0x API that traded on Uniswap v1 or v2 indirectly and were not part of the original airdrop.