On December 6, 2002 01:19 pm, Alexandre Julliard wrote:
Sure, it should run from the build directory (though we should fall back to using the path if we don't find it there). Isn't that what the script already does?
I don't remember now, there was something I didn't like. :) Nevermind that, now the question is: should I include the script in winewrap? Otherwise, we have to install wineapploader, and I don't quite see the point in that. We can have a --loader option that allows you to use an external script if need be, in which case we can maybe even simplify the generated script a bit.
The 32-bit ones do, so it would be possible to identify them. But actually I could hack winebuild to do that itself, and then simply get rid of the -r option.
I think this is a better option, and the reason is that it would allow people to more easily use winebuild directly, which is preferable because the generated files will be in sync with the Makefile targets (whereas with winewrap, we're generating files the Makefile doesn't know about).