| Commit message (Collapse) | Author | Files | Lines |
|
|
|
|
|
|
|
|
|
|
|
calculations.
|
|
the future, they will be completely optimized away.
|
|
used (they are easier to optimize away).
|
|
|
|
|
|
|
|
- There are still some corner cases where decoding fails. MUST FIX.
|
|
|
|
epoch).
|
|
|
|
|
|
with the relevant signatures.
|
|
finally-tagless approach.
|
|
homogeneous.
|
|
|
|
|
|
|
|
|
|
- Minor refactorings.
|
|
|
|
parts of the whole input, instead of having to clip the input as they lex. thereby doing a lot of unnecessary text allocations.
- Some refactoring.
|
|
escaping).
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
- Fixed some bugs.
|
|
|
|
the actor's state, and to the actor itself.
- When creating channels and promises, the type is now mandatory.
- Minor refactorings.
|
|
|
|
|
|
|
|
- Created ordered dictionary implementation, and based the set implementation upon it.
|
|
- Also did argument swapping for lux/type;application.
|
|
|
|
|
|
|
|
|
|
|
|
- Lexers now rely only on Text, instead of also relying on Char.
|
|
|
|
"writer").
|