Hi again Bill,
At 22:51 19.06.2002, William A. Rowe, Jr. wrote: >Einar... > >This has been on my to-do for 3 years ;) Thanks for taking up the challenge. > >If we could set up an evening next week or over the weekend, perhaps we have >an informal Win32/mingw hackathon on irc://irc.openprojects.net/ #apr channel >to trade ideas with anyone interested in attacking this, and turn around >working >1.3 and 2.0 build methods. > >The problem with mingw has been the lack of libtool/autoconf/automake/m4 >support, but I'm hoping to hear that the situation has improved over time :) So when should we do this? From thursday evening (CET time :) tothe 4th of July I'm free most of the time. Then it's summer vacation :) Should we schedule something? Just tell me your timezone too, because we don't want any disagreements about the time! Anyway, I've been looking into the subject a little when I have had the time: for me it seems like the most reasonable approach would be to forget the libtool things, as it's still imperfect :( but rather treat Mingw GCC as just another windows compiler, and port the win32 makefiles to that. This is the path taken in Perl which has very good Mingw support IMO. I've just been looking at why Perl uses dmake as opposed to nmake for the Mingw build before trying to modify the makefiles to any extent. (on another note, my nmake downloaded from an MS site complains when looking at the Apache makefiles.. the '::'s seem to be bugging it.. weird). Anyway, as Mingw exposes the Win32 API and links to MSVCRT.DLL I think the win32 code should be able to run cleanly. Or atleast hope so :) -- Per Einar Ellefsen [EMAIL PROTECTED]
