Read The (Freely Hosted) License
Why you might ask? Here are a few reasons:
Basically every modern language has accepted a repository centered view of licensing with a single top level file ([repo_root]/LICENSE) which integrates with build-system, IDE, code-hosting, indexing, and search.
A license is not a copyright. A license is a document describing the terms under which a piece of software may be used, whereas a copyright is an AUTOMATIC international right granted to all content creators which forbids the unlicensed use/duplication by others.
It is the responsibility of the duplicator of copyright work to ensure that their use is permitted by the copyright holder.
If every time an reader loads up a file they must first take a second to scroll past redundant front-matter, and they do this (conservatively) 5 times a day, and only on work days, by the end of their career thy will have spent an entire day of their life ignoring this garbage.
This means that for the 21 million engineers today, we are spending over 5.5 human lifetimes EVERY YEAR. These are the collective lives of people like you that we will never get back. Which reminds me...
Do you know the copyright date on your files? Copyright is protected for 50 years after the content creators death. Inclusion of copyright date and authorship is the best way to make sure these rights are protected. Are you excited about updating this in every single file in your whole repository? You sure shouldn't be because there is literally a best practice acronym to describe why this sucks...
Unnecessary duplication leads to version drift between copies, poor maintainability, and host of other issues. Without a mechanism to check for consistency these will end up wrong, this weakens protection, wastes space, and has no...
Without a single source of truth your protections are weaker. Which of several competing licenses apply is less clear and overall weakens your protections.
This rtfl.io is here to help.