Hello, Thanks for this nice PEP, I have just read it. I think it is a good thing to add to Python and will enable cleaner code. I support the idea to have common sentinels suggested by Christopher Barker. For example, a common confusion is to identify None with NULL in DB, but then when you are filtering rows in a SQL query generated by a framework function, you have to do convoluted things because most frameworks do not distinguish "Do not filter" and "Keep only rows where value is NULL", because the vocabulary of sentinels is too limited. Some sentinels could be created also for some constants of R language for example. Anyway, extending the language with common sentinels can be done later. But it would be nice to really think about it. Regarding the type of Sentinels, the proposition in the PEP is not perfect. Recently, I used MY_SENTINEL = object() TypeMySentinel = Literal(MY_SENTINEL) in my code. But what I would have like to be able to do is just : TypeMySentinel = MY_SENTINEL exactly like None "is" its own type. However, it was rejected on import of the corresponding file, when using it in function signature. It would be really nice if all sentinel values could have this feature to be their own type. Last, I spotted a typo in the PEP : "Most common exiting idioms" -> "existing' Thanks, best regards, Laurent Lyaudet, PhD