Niels Provos wrote: > On Sat, Nov 22, 2008 at 2:10 PM, Matthew Weigel <[EMAIL PROTECTED]> wrote: >> After all that, I can build libevent but none of the test programs, so it's >> hard to tell whether it works at all or not. I'm still plugging away, and >> I'd >> be happy to share the newly-created .sln files, but does anyone have more >> positive experiences? Is there something fundamental I'm missing? > > I am not really a Windows user or developer.
I know the feeling, I lean heavily on Emacs and Cygwin for development on Windows. On the other hand, I still use Microsoft's toolchain to actually build software. Have you considered switching to nmake for building on Windows? > With 1.4.8, you should > have vcproj files for both libevent and the regression program. Yes, but not event_test, signal_test, or time_test. > Using those I can build the library and run the regression program > without problems. You should be able to verify that there are at least *some* problems with the included project files in the 1.4.8 tarball; for instance, line 292 of WIN32-Prj/libevent.vcproj specifies that the file "..\WIN32-Code\event-config.h" be part of the project, but no such file exists. Like I said, I had to move event-config.h from the main source directory into WIN32-Code/, and modify it. Note that copying the file was insufficient: the version in the main directory is found before the one in WIN32-Code/, with its default configuration which doesn't work on Windows. I've also found some more things I've needed to comment out in event-config.h, it's been trial and error as I find problems since there isn't a Win32-specific one included. When I'm done I'll simply provide the file that *does* work. So far, most of the work to get the regress project to build has been fiddling with event-config.h, include directories, library dependencies, that kind of thing. My best guess is that a) WIN32-Code/event-config.h simply didn't make it into the tarball, and b) you may have set some options globally for your Visual Studio installation rather than specifying them in the projects themselves. With all that done, I've got regress building and all tests passed except IPv6 DNS resolve: for my purposes, more than sufficient. I'll try to have a tarball of changes up later today. -- Matthew Weigel hacker unique & idempot.ent _______________________________________________ Libevent-users mailing list Libevent-users@monkey.org http://monkeymail.org/mailman/listinfo/libevent-users
