Ok, I did misunderstand you.
Thank you.
Then we agree that using _WIN32 for pure Windows stuff is superior to HAVE_THIS_OR_THAT guards which *only* benefit that platform.
> Your patch cripples FileInfo. Agree? The code in FileInfo.C
works perfectly well under MinGW, so your proposed changes to FileInfo.C AS THEY STAND are going to be a step backward on MinGW.
He, he, MinGW is a step backwards, so what?
No, I agree. I understand that FileInfo is a mess, and I'd like to clear it up eventually, either by ripping out features or by refactoring. But the question is this:
1) Do you want me to do this kind of janitor stuff? 2) or do you want me to find and fix the difficult bugs in the kernel?
It's your choice.
If you choose 1), I'd just forget about integrating the patches into CVS for now, and just do 2) anyway with the portability changes being local to my computer.
Andre and others, which have a desire for working with a decent Windows toolchain, can then ask for the patches and them themselves.
Regards, Asger
