On Wed, Jun 8, 2011 at 3:45 PM, Amos Jeffries <[email protected]> wrote: > On 09/06/11 01:17, Kinkie wrote: >> >> On Wed, Jun 8, 2011 at 3:11 PM, noc wrote: >> >>> Changes: >>> >>> [Francesco Chemolli] Removed some dead code in case of disabled >>> authentication >> >> Fascinating. This code seems to break the build on mswin, and it >> passed the test-suite on my natty box. >> Uncommitting; sorry for the spam. >> > > #if USE_AUTH must go underneath config.h. ;)
Whoops :) That change is valid so.. recommitting after a check > FWIW I went through the mswin branch and have been trying to bisect the > successful changes back into trunk. There is a lot of mess in that diff. > Some of it may explain the weird build behaviour left on mswin. A bit of bitrot, unfortunately. Thanks for your effort (I'll keep on banging my head into win-specific issues in the meantime). > I spent most of this evening looking up how to get Jenkins to commit after a > successful build so we can do full continuous integration. Nothing easy. I'm not sure I understand. What would you like to do? If a build succeeds on Windows and Linux, merge the change to trunk automatically? > On the other side, there are cppcheck and other static analysis tools with > plugins to scan build logs for issues. I'm using Eclipse for that... double click on the error, open the offending line ;) -- /kinkie
