Daniel

For the record, this briefly came up in our meeting again today, and @chriseth mentioned that Sourceify calls the cbor metadata in the bytecode "aux data" these days, which is...

Pinging @bshastry about this one, just to confirm that the interpreter changes won't interfere with fuzzing (since most of it is conditional behind the flag, it shouldn't affect much, but...

I'm not entirely sure I understand what you're getting at, but the memory layout of dynamic arrays other than ``string`` or ``bytes`` is specified to be padded to 32 byte...

It happens a lot on compiler updates and it's better to be able to tell people who then file issues about it (after looking at the situation) that it's safe...

And if pedantic checks are enabled by default and in CI, this entails no safety loss. So I say we can go ahead with this.

https://github.com/ethereum/solidity/pull/12762 pretty much showcases how this can be done with the recent more elaborate semantic information we have - so the same mechanism just needs to be transferred to the...

Staging this as a task ready to be worked on for using the mechanism of https://github.com/ethereum/solidity/pull/12762 to determine when a memory-writing call has a length of zero and, based on...

Yeah, this is a known issue - we have experimental research towards general reasoning- and SMT-based optimizations that are intended to be powerful enough to remove the overflow checks in...

We had an initial discussion about this today and decided that we want to further explore a few options. - @hrkrshnn will check, if it's feasible to accelerate getting an...

The EVM defines division by zero to be zero, not as a revert (as you say yourself) - since our code generation targets an EVM following that specification, I'd not...