Document v11 features.

[cristianoc]

This will follow from the v11 blog posts.

• Untagged Variants See e.g. @unboxed fails if a constructor has more than one argument rescript-lang/rescript-compiler#6272
• New Record Features
• Uncurried mode
• First-class Dynamic Import Support

[fhammerschmidt]

Here are some changes from the changelog that are still missing in the v11 docs:

• Freely configurable suffix for generated .js files. Freely configurable suffix for generated .js files rescript-lang/rescript-compiler#6472
• Allow coercing unboxed variants with only strings (now including with a single payload of string) to the primitive string. Allow coercing unboxed with payload to primitive when applicable rescript-lang/rescript-compiler#6441
• Allow coercing strings to unboxed variants that have a catch-all unboxed string case. Support coercing string to elgible untagged variants rescript-lang/rescript-compiler#6443
• Allow coercing int to float. Allow coercing int to float rescript-lang/rescript-compiler#6448
• Support the async component for React Server Component in JSX V4. Support async component for React Server component in JSX v4 rescript-lang/rescript-compiler#6399
• Support rescript.json configuration file and deprecate bsconfig.json. Support rescript.json rescript-lang/rescript-compiler#6382
• Stop mangling object field names. If you had objects with field names containing "_" or leading "", they won't be mangled in the compiled JavaScript and represented as it is without changes. Remove obj keys mangle rescript-lang/rescript-compiler#6354
• Record type spreads: Allow using type variables in type spreads. Both uninstantiated and instantiated ones. Allow type variables when spreading record type definitions rescript-lang/rescript-compiler#6309
• Introduced a new %ffi extension (experimental - not for production use!) that provides a more robust mechanism for JavaScript function interoperation by considering function arity in type constraints. This enhancement improves safety when dealing with JavaScript functions by enforcing type constraints based on the arity of the function. Add %ffi as variation of %raw with arity checking. rescript-lang/rescript-compiler#6251

[zth]

Feels like a good amount of these don't need any actual documentation, unless the documentation explicitly states something contradicting the changes.

[fhammerschmidt]

Consider them as hints to know where to check if the docs are still up to date.

[zth]

Crossed off 2 that were handled in this PR: #740

[zth]

Don't think we need to mention these in the docs, they were bug fixes:

• Support renaming fields in inline records with @as attribute. Support renaming fields in inline records rescript-lang/rescript-compiler#6391
• Support renaming object fields of @obj external ppx with @as attribute. Add @as support for obj ppx rescript-lang/rescript-compiler#6412

[zth]

Same with these, seem irrelevant to me for docs:

• Update watcher rules to recompile only on config and .res/.resi/*.ml/.mli file changes. Solves the issue of unnecessary recompiles on .css, .ts, and other unrelated file changes. Update watcher rules rescript-lang/rescript-compiler#6420
• rescript build will always build its dependency by default. The argument -with-deps is not needed anymore. Make build -with-deps a default behaviour rescript-lang/rescript-compiler#6350
• $$default is no longer exported from the generated JavaScript when using default exports. Don't export $$default rescript-lang/rescript-compiler#6328
• Parse assert as a regular function. assert is no longer a unary expression. Example: before assert 1 == 2 is parsed as (assert 1) == 2, now it is parsed as assert(1 == 2). Parse assert as a regular function rescript-lang/rescript-compiler#6180
• Support doc comments on arguments of function types. Parse doc comments in type parameters. rescript-lang/rescript-compiler#6161
• Made pinned dependencies transitive: if a is a pinned dependency of b and b is a pinned dependency of c, then a is implicitly a pinned dependency of c. This change is only breaking if your build process assumes non-transitivity.

[fhammerschmidt]

Thanks. Updated the list.