You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I get 2 errors when compiling lambda_runtime_error v0.1.1:
error[E0277]: the trait bound (): LambdaErrorExt is not satisfied--> $HOME/.cargo/registry/src/github.com-1ecc6299db9ec823/lambda_runtime_errors-0.1.1/src/error_ext_impl.rs:145:9
error[E0277]: the trait bound (): std::error::Error is not satisfied --> $HOME/.cargo/registry/src/github.com-1ecc6299db9ec823/lambda_runtime_errors-0.1.1/src/error_ext_impl.rs:145:9
Hrm, thanks for the report. That's... not great. I'm not 100% sure what the cause is, but I think this relates to ! being stabilized around that time. As a result, type inference for () and ! changed, resulting in the error you're seeing.
That being said, we should be able to dump the lambda_runtime_error crate now. If you'd like to send a PR removing it (and replace error reporting via std::any::type_name), I'd gladly accept the PR. I don't think the error type in a Lambda function needs to be constrained to anything now—it can just be an unbounded E.
Until the end of the year, I haven't the time to do it. I'm doing my Master Thesis which is extremely time-consuming. If someone else could fix it, until then, it would be great. Otherwise, I will look into it, in January.
I modified my implementation so I can use the stable toolchain instead. I tried to compile with the nightly (cargo 1.42.0-nightly (6e1ca924a 2020-01-06)) toolchain and it worked too.
So as suggested by @davidbarsky I closed the issue.
Activity
davidbarsky commentedon Nov 26, 2019
Hrm, thanks for the report. That's... not great. I'm not 100% sure what the cause is, but I think this relates to
!being stabilized around that time. As a result, type inference for()and!changed, resulting in the error you're seeing.That being said, we should be able to dump the
lambda_runtime_errorcrate now. If you'd like to send a PR removing it (and replace error reporting viastd::any::type_name), I'd gladly accept the PR. I don't think the error type in a Lambda function needs to be constrained to anything now—it can just be an unboundedE.jmendes92 commentedon Nov 28, 2019
Thanks for your feedback.
Until the end of the year, I haven't the time to do it. I'm doing my Master Thesis which is extremely time-consuming. If someone else could fix it, until then, it would be great. Otherwise, I will look into it, in January.
fixes aws#175
tsal commentedon Jan 2, 2020
I'm not sure if it's related, but this seems to no longer fail (it did for me when I checked a month ago). Even without #182 being merged.
Granted, I'm not using the
lambda_runtime_errorcrate.davidbarsky commentedon Jan 3, 2020
@tsal missed that fix, oops. the stabilization of
!got reverted, so this issue can probably be closed.jmendes92 commentedon Jan 9, 2020
I modified my implementation so I can use the
stabletoolchain instead. I tried to compile with thenightly (cargo 1.42.0-nightly (6e1ca924a 2020-01-06))toolchain and it worked too.So as suggested by @davidbarsky I closed the issue.