"moderate" makes sense here; those are issues that needed fixing, but they wouldn't give someone privileged access they shouldn't have, and they occur in non-default configurations.
> A highly-privileged user (able to run commands as other users, or as root, through sudo) who knows one password of an account they are allowed to run commands as, would be able to run commands as any other account the policy permits them to run commands for, even if they don't know the password for those accounts.
> A common instance of this would be that a user can still use their own password to run commands as root (the default behaviour of sudo), effectively negating the intended behaviour of the targetpw or rootpw options.
The drive to rewrite existing, tried and tested code in the new trendy language is crazy. Hopefully problems like this will be a bit of a reality check for those cheerleading it.
> A partially typed password would be output to standard input if a timeout occurred when Defaults pwfeedback was not enabled (GHSA-q428-6v73-fc4q).
> Timestamp files did not take into account the setting of the Defaults targetpw and Defaults rootpw (GHSA-c978-wq47-pvvw)
"moderate" makes sense here; those are issues that needed fixing, but they wouldn't give someone privileged access they shouldn't have, and they occur in non-default configurations.
> access they shouldn't have
It does, quotes from https://github.com/trifectatechfoundation/sudo-rs/security/a... below:
> A highly-privileged user (able to run commands as other users, or as root, through sudo) who knows one password of an account they are allowed to run commands as, would be able to run commands as any other account the policy permits them to run commands for, even if they don't know the password for those accounts.
> A common instance of this would be that a user can still use their own password to run commands as root (the default behaviour of sudo), effectively negating the intended behaviour of the targetpw or rootpw options.
I stand corrected, you're right. I misread the description of the vulnerability.
I'm surprised how little attention this is getting on HN.
Not sure what's news-worthy here? Safe Rust eliminates certain classes of bugs, not all bugs. The editorializing of the title seems like strawmanning.
The drive to rewrite existing, tried and tested code in the new trendy language is crazy. Hopefully problems like this will be a bit of a reality check for those cheerleading it.
A bug in Ubuntu's standard sudo executable is newsworthy.
Updates are probably required for a large number of servers.