On Mon Aug 1 20:33:16 2022 +0000, Jacek Caban wrote:
What interesting use cases do you mean? Note that one may just build widl as PE if it's just for sake of running widl on Windows. So far you only mentioned being able to test parser failures. I think that it would be generally more interesting to make widl itself more compatible with midl.
Well for instance having an open-source, drop-in replacement alternative to midl.exe.
I think that it would be generally more interesting to make widl itself more compatible with midl.
Yes, I agree, though then it'd possibly break backward compatibility. I'm not sure we want that as widl is used publicly elsewhere. Having a builtin midl.exe (and possibly a Unix version of it) would be a better way to get a command-line compatible tool, improving widl at the same time for things that do not break backward compatibility, but eventually deprecating it once the other is in good shape.