now handling ERC1271 validation in the library as well

Refactor ready @sohkai: https://github.com/aragon/aragon-apps/pull/580/commits/345a9d8aa4f411df6d2b5044a801b2cb7d9ca6ee

@jorge was running the agent coverage tests locally and for some reason the signature tests seemed to not work

@jorge about this:
https://github.com/aragon/aragon-apps/pull/657#discussion_r255435114
Tests are not passing now. Do you want me to raise it to 6.4e6, which would be closer to 80% of current limit, or to lower down the number of max allowed tokens?

Might as well use 6.5e6 and give that a shot

ok, sounds good (I said 6.4 because it’s 80% of 8 and because it’s now using 6.37, but I’ll do 6.5)

for some reason the signature tests seemed to not work

hmm but the tests were passing?

@jorge Would be great if you could take a look at https://github.com/aragon/aragon-apps/pull/686

@all Does anyone using VSCode like to try out a security-oriented solidity extension? One of the AGP-18 security partners built https://github.com/tintinweb/vscode-solidity-auditor and it looks really, really helpful 🔥. Feel free to file issues on github!

Installation instructions: https://github.com/tintinweb/vscode-solidity-auditor#installation

The next two lines give File import callback not supported when compiling using 0x dev-tools: https://github.com/aragon/aragonOS/blob/dev/contracts/lib/ens/ENS.sol#L4
https://github.com/aragon/aragonOS/blob/dev/contracts/lib/ens/PublicResolver.sol#L3
I fixed it using double quotes. Is this a bug?

Good question; the docs officially only use double quotes, so we should use those instead

Perhaps it was old syntax from earlier versions of Solidity that have since been removed, but the newer contracts in ENS are now using the double quotes too :shrug:

https://blog.ethereum.org/2019/03/26/solidity-optimizer-and-abiencoderv2-bug/

https://medium.com/@mykelp/introducing-polaris-ced195dd798e

pretty interesting price feed for tokens that uses uniswap prices and prevents manipulation by not taking the current price but a median of checkpoints. we should check it out for payroll

cc @sohkai @facuspagnuolo @lkngtn @bingen

it is live actually: https://github.com/marbleprotocol/polaris

https://github.com/ethereum/EIPs/pull/1884

SLOAD might go from 200 to 800 gas (not ridiculous, but a 4x)

Hey guys, I need some Solidity expertise advice. I want to avoid this if:
https://github.com/aragon/aragon-court/blob/getLastPresent/contracts/lib/HexSumTree.sol#L196

• because it makes the code more expensive (even if it’s just an if)
• because past value never changes once the initial function is called, so it’s the same for every iteration of the loop and every recursion step
• because for the case sortition (where I have the same problem) adding this param makes me hit the stack too deep error
• beacuse it makes the code uglier
  First idea I had is to just copy the whole function changing only that line, but this sucks of course (too much repeated code)
  As far as I know there’s nothing similar in solidity to passing functions as parameters or dependency injection or any other fancy technique.
  I also thought of overloading the function get, but it comes from Checkpointing and there is no Library inheritance either.
  Any ideas?

IMO it is okey as it is, definitely cleaner than copying the entire function as in _sortition and _sortitionLastPresent

if this ends up being limited to just the current term (as I think we should do), then we could remove the option of doing it in the past all together and just allow linear searches

yes, that _sortition duplication is temporary, it was a quik work around for stack too deep issue, and I was waiting as you said to see if we really need both options or stick to just one

yes, that _sortition duplication is temporary, it was a quick work around for stack too deep issue, and I was waiting as you said to see if we really need both options or stick to just one

Is truffle supposed to compile dependencies? If so, why? security? 😕
(trying to update truffle to v5 which uses solc-js v0.5 and the compilation fails)

Moving to truffle@5 is going to be a longer process across the entire development pipeline

@danielconstantin Moving to truffle@5 is going to be a longer process across the entire development pipeline

I imagine we could start doing it for the boilerplates though

You should also be able to specify the solc version (it will have to be 0.4.24) in the truffle config

This is super cool: https://github.com/cgewecke/eth-gas-reporter/blob/master/docs/codechecks.md