Repository navigation
Suggestion Backlog Slog, 6/8/2016 #9034
Description
Activity
- Suggestion: Only create global variable for module if something tangible is exported #8668 Only create global variable for module if something tangible is exported
- Why not use an IIFE?
- Qualification
- Does future extension matter?
- Don't see why
- Too complex; workaround exists
- Why not use an IIFE?
- Bug: Accessing private statics in a class via its derived class is allowed #8624 Accessing private statics in a class via its derived class is allowed
- Agree
- Breaking change
- Should awaiting a non-Promise value be an error? #8310 Should awaiting a non-Promise value be an error?
- No
- It's sometimes useful
- Also OK because consistency with union types
- And generics
- Do not rename imports from ambient modules #8227 Do not rename imports from ambient modules
- Agree
- User must write alias and then rename
- Fix is to restrict syntactic forms of rename invocation sites
- Suggestion: stricter operators #7989 Stricter binary operators
- Clearly suspect for numeric/bitwise operands; make these an error
string + [whatever]is still OK, or at least less clear- What about add an option that flags coercion #7746?
- Proposal: Disallow implicit truthiness checks for non-null types in strictNullChecks mode?
- What constitutes a truthiness operand?
ifand other control flow expressions, and the left operand of||/&&, operand of!
- What constitutes a truthiness operand?
- Other behavior listed here is idiomatic
- Proposal: Disallow implicit truthiness checks for non-null types in strictNullChecks mode?
- Make exported consts better minifiable #7565 Make exported consts better minifiable
- 👍 and wait for new emitter to land
- make the strict equality comparison operator a typeguard for string literal types #7447 make the strict equality comparison operator a typeguard for string literal types
- Breaking change: This would (badly) restrict the domain of variables whose type was inferred from
value - 👎 without use cases to justify break
- Breaking change: This would (badly) restrict the domain of variables whose type was inferred from
- Consider re-ordering Array#reduce overloads in lib.d.ts #7014 Consider re-ordering Array#reduce overloads in lib.d.ts
- Try it and see wha thappens
- Support override keyword on class methods #2000 Support override keyword on class methods
- What about
overrideproperties?- Sure, no reason not to
--noImplicitOverridemeans what you think it does- Should this apply to declaration files?
- Argument for yes: correctness for the same reason
- Argument for no: convenience, no need to mess with declaration emitter
- If no: Declaration emitter will need to emit
overridemodifier (same wayanyworks) if one was not present explicitly - Decision: Not enforced
- Should this apply to declaration files?
- Nothing for interface implementations at this time
- No special type behavior at this time
- Accepting PRs
- What about
- No compile error thrown when
thisreferenced before call tosupercompletes #8060 Re-order emit ofvar _this = thiswhen arrow func appears in super call- 👍, Community
- Extended type parameters don't work as constraints with overloads #7378 Extended type parameters don't work as constraints with overloads
- Overload/implementation signature checking is already pretty bad anyway
- Too Complex
Out of time
- Allow ArrayLike as a spread source #7596 Allow ArrayLike as a spread source (and figure out emit semantics)
- Suggestion: Only create global variable for module if something tangible is exported #8668 Only create global variable for module if something tangible is exported
- locked and limited conversation to collaborators
on Jun 19, 2018
thisreferenced before call tosupercompletes #8060 Re-order emit ofvar _this = thiswhen arrow func appears in super call