We've started enforcing the use of `--styleCheck:usages` the moment it landed 
in the compiler - ie we don't allow using different style on use than the 
declaration: in secure software, using different styles is a good way to sneak 
in unexpected identifiers that don't look similar at a glance, and therefore 
can be used to mislead a reviewer or code auditor - specially because of nim:s 
global scope, this is a significant issue in any larger codebase.

Similar issues happen with lookalike or empty unicode characters on a regular 
basis - these get reported as CVE:s of fairly high severity: using different 
styles for the same identifier is just one more way to make things inconvenient 
for anyone that's reading the code, and a good way to introduce security bugs.

Reply via email to