Piotr Fusik

To avoid the problem of interrupts overwriting the return address, have a subroutine that returns it: PLA TAY PLA TAX PHA TYA PHA RTS and maybe write it on ZP...

The reason is that Java and JavaScript do not provide 32-bit unsigned arithmetic. The separate type only documents that negative values are not expected. I'd like to have a true...

```csharp int a = 0xFFFFFFFF; // or uint a = 0xFFFFFFFF; ``` > And this is not compilable unless I add `unchecked`. Your code is invalid. You get a legitimate...

Thanks! I have no plans for concurrency yet. I don't know a design that could fit all the targets (think JavaScript, Python and OpenCL). `cito` has no own garbage collector....

Atomic operations seem very portable between the languages that support them. OpenCL does have them. I'm interested in having them in Ć. Also, mutexes seem like a low-hanging fruit.

Yes, that's the plan! I'd love to have `cito`: - without the dependency on .NET - running right in your browser I've made some progress already: added `HashSet` and `Stack`...

Current status: lexer fully ported to Ć, most of the AST and the parser ported to Ć.

All tests are run in Travis CI. C, C++ and Swift tests are run with the Address Sanitizer and `OpAddAssignString` passes. There's this line in the C translation of `OpAddAssignString`:...

Yes, it's a known issue. I've just added a test and will fix it. Interpolated strings are even worse when targetting C++, because they use C++20 ``. Even though C++20...

`cito` cannot determine all the dangling reference scenarios. If it were easy, C and C++ compilers would do it, given five decades of their development. `Consume(Create())` would be valid code...