On Tue, 2011-05-24 at 21:00 -0400, Charles Lepple wrote: > Please be sure to merge r2990 and any other recent changes (r2993, > r2994?) onto windows_port as well. (The patch file dated "24-May-2011 > 16:10" seems to revert those changes.) A good way to catch things like > this is to indicate the range of revisions which you are merging in > the commit message, and check the last merged revision against the > trunk. (Newer versions of Subversion supposedly automate this process, > but I am skeptical.)
My mistake, I use "svn diff -r HEAD" to generate the patch. I should have used the revision number of the merge with trunk. Anyway I have merged the latest revision now. Thanks for the tip about the revisions range in the commit message. > > Also, please ensure that the C files have Unix-style line endings when > they get checked in. Setting the svn:eol-style property to "native" on > those files will cause LF to be translated to CR+LF on Windows > platforms. I have fixed the eol style now. I am not sure if it is worth adding the property since the building tools on Windows seems to behave correctly with unix style eol. > > I'll start a build of the windows_port branch on a *nix machine > shortly, but have you been able to do any regression testing of that > branch with a non-Windows machine? I have briefly tested the windows_port branch on a ubuntu 11.04 64 bits host with a Eaton-Powerware 9130 connected via USB and I didn't see any obvious regression. Thanks for your feedbacks, Fred -- Team Open Source Eaton - http://powerquality.eaton.com -------------------------------------------------------------------------- _______________________________________________ Nut-upsdev mailing list [email protected] http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/nut-upsdev
