From: Xiaofan C. <xia...@gm...> - 2010-04-23 01:26:42
|
On Fri, Apr 23, 2010 at 7:58 AM, Michael Plante <mic...@gm...> wrote: > Correct me if I'm wrong, but I figured that's nearly what will be released > anyway (a ./configure script already generated). And I thought that solved > the problem, since I'm told that ltmain is the issue here. Of course, I > have no idea, but I just want to be sure what you're saying really differs > from what's been previously suggested. You are right. Either if you include Pete's patch or if you have the latest libtool, then the generated configure script will be okay. If you do not include Pete's patch, and since most people do not have the latest libtool, then it is required to issue the "export LANG=C" before running Make from MinGW. > Now I think Graeme at least is saying cut autotools out entirely, which I > don't think is going to fly. If we were only supporting MSVC/DDK on Windows > and no other platforms, then maybe. > I understand that is not going to fly. However, I still intend to try out to see if I (with the help of others) can come out with a libusb-win32 style Makefile for libusb-1.0 Windows backend (even with some limitations). Just a personal learning experience. I think it can be done. Afterall, MSVC/DDK do not need the auto-tools. IMHO autotools are more applicable to Linux (or BSDs) and Cygwin, and not that compatible with the initial design goal of MinGW/MSys. -- Xiaofan http://mcuee.blogspot.com |