Foundation contest - kenta's results

• safeTransferFrom: We have had a few suggestions on this. ATM we are going to continue with transferFrom because it lowers risk of reentrancy. But we may revisit this in the future.
• Incorrect comment: Thanks for pointing this out, we have removed the old comment.
• delete unnecessary code: This code is used - e.g. when canceling an auction.
• delete unused variable name: We added return names to help with documentation, but then it creates an inconsistent style. So we have updated to use the named params instead of return.
• missing inputs validation: There were also a couple other reports of this, we have added a check to asset the lengths match.
• address is not necessary: Yes, we have removed the unnecessary cast. This helps with readability.
• emit: Thanks for the suggestion, but this is for the Treasury contract so out of scope of this contest.

Unadjusted score: 35

• Replace storage with memory: We did some testing of the gas impact and found that storage is most often less gas than memory. The key here seems to be which slots are accessed. When memory is used it makes a full copy, so all the data is read. However often in logic we are able to return after reading just one of the datapoints in the struct. By using storage we only read the data that's required. As a general best practice we have been using storage unless memory is required in order to access the original data (e.g. to emit after a delete).
• Cache at the beginning: That suggestion does seem to improve readability, but we tested that change and it increases gas costs by ~2.3k gas in the happy case (where we return before the memory line).
• No need to cache: Generally these lines were added to make the code easier to follow. We tested a couple of the suggestions and found <100 gas savings.
• Code duplication: The examples provided are just reading from a map. I'm not sure using the getter helps with readability. Tested the change and it did not seem to impact gas or size, so this is just a style preference.
• Initial value for uint256: We tested making this change and did not find a measurable impact on gas. We'll keep the default values, which are redundant, for improved readability.

Without testing the gas savings, it's unfortunately difficult for wardens to ensure that the gas savings are real. Food for thought.

Score: 100