3 ms·
Coming from an outside reference frame (I have zero knowledge of the - obviously contentious - backstory, and I am holding off on reading that backstory until I
by SolarNet 7y ago
Coming from an outside reference frame (I have zero knowledge of the - obviously contentious - backstory, and I am holding off on reading that backstory until I review this language on it's merits).
On my examination it looks like a very interesting transpiler. I suspect that like most languages it will bloat as it attempts to achieve it's goals, especially with the size of the standard library that is going to be included by implication.
However if it's goal was to be a useful C transpiler for the modern age I would be more excited, and more interested. Things like hot reloading, easy to use REST access, nice syntax, and the ability to fall back to C are all nice features. For a certain class of application (notably the author's Volt application, and video games) this would be very useful.
I would warn the author of throwing away what they have for going down the rabbit hole of language implementation. It is nowhere as simple as they seem to imagine. An anecdotal story, when using LLVM one will regularly come across required parameters that seem like they should be optional and don't seem important. And yet once one starts thinking about compatibility, errors, ABIs, they realize there are hundreds of edgecases that must be accounted for. For example structure returns on the windows ABI change the ordering of function call arguments depending on the size of the structure, and additionally require redundant memory operations (and failing to do both will cause your compiler to work for some libraries but not others).