comphead commented on PR #319:
URL: https://github.com/apache/datafusion-spark/pull/319#issuecomment-6923717351

   Two follow-ups on the API compatibility section:
   
   1. **`#[non_exhaustive]` wording.** The attribute applies to structs, enums 
and enum variants, not to fields, so "new `#[non_exhaustive]` fields" can't be 
done. The rule is also stricter than needed: adding a field only breaks a 
struct whose fields are all public. Suggested wording: "add a public field to a 
struct whose fields are all public, or a variant to an enum, unless the type is 
`#[non_exhaustive]`". Adding `#[non_exhaustive]` to an existing public type is 
itself a breaking change, and the section could say so.
   
   2. **The SQL-facing API isn't covered.** Most users reach the crate through 
`register_all` and `all_default_*_functions`, not through Rust items. 
`cargo-semver-checks` can't see changes such as dropping a function from those 
lists, renaming or removing a SQL name or alias, or changing a return type. A 
bullet listing these as breaking would help. So would a note that 
`cargo-semver-checks` doesn't cover them. It would also help to say whether a 
return-type change that makes a function match Spark counts as a bug fix, like 
a change in results.
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to