#244 expands the borrowck implementation for what we want to reach in the future: more precision, less errors, etc.
In doing that, we aren't modeling NLLs anymore, and for differential testing purposes with the compiler, we would want some way to also add back the outlives constraints needed to accurately model the NLLs compiler implementation, when we need it.
Similarly, we'll also want to model our other algorithms, and they each have their own precision tradeoffs. A solution that would work for these multiple cases will be needed.
This issue is to track this FIXME, and record this work we need to do after that PR is merged.
#244 expands the borrowck implementation for what we want to reach in the future: more precision, less errors, etc.
In doing that, we aren't modeling NLLs anymore, and for differential testing purposes with the compiler, we would want some way to also add back the outlives constraints needed to accurately model the NLLs compiler implementation, when we need it.
Similarly, we'll also want to model our other algorithms, and they each have their own precision tradeoffs. A solution that would work for these multiple cases will be needed.
This issue is to track this FIXME, and record this work we need to do after that PR is merged.