Hi, I can reproduce this problem, though I also don't see the WAR repackaged if I make changes to src/main/webapp/test.html. This is on OS X 10.6.6 with buildr 1.4.4 on MRI 1.8.7.
$ ls -l target/ total 8 -rw-r--r-- 1 rsutphin wheel 637 Jan 18 10:17 buildr-1.0.0.war $ ls -l src/main/webapp/ total 8 drwxr-xr-x 3 rsutphin wheel 102 Jan 18 10:16 css/ -rw-r--r-- 1 rsutphin wheel 11 Jan 18 10:29 test.html (Note that test.html was modified after the war was last built.) $ buildr package (in /private/tmp/buildr, development) Building buildr Packaging buildr Running integration tests... Completed in 0.005s $ ls -l target/ total 8 -rw-r--r-- 1 rsutphin wheel 637 Jan 18 10:17 buildr-1.0.0.war Despite buildr indicating that it's packaging, the war isn't changed. I have noticed intermittently that some the subproject dependencies in a complex build aren't rebuilt after a change (usually a resource change). I hadn't taken the time to isolate the problem (since a clean build of the affected subproject is a cheap workaround), but this might be the same problem. Rhett On Jan 18, 2011, at 10:04 AM, Ingo Schmidt wrote: > Hi Antoine, > > I hadn't included the buildfile, because I had included the steps how to > create it (I let buildr create it). Anyway, here it is for convenience: > --------------- snip --------------- > # Generated by Buildr 1.4.3, change to your liking > # Version number for this release > VERSION_NUMBER = "1.0.0" > # Group identifier for your projects > GROUP = "buildr-test" > COPYRIGHT = "" > > # Specify Maven 2.0 remote repositories here, like this: > repositories.remote << "http://www.ibiblio.org/maven2/"; > > desc "The Buildr-test project" > define "buildr-test" do > > project.version = VERSION_NUMBER > project.group = GROUP > manifest["Implementation-Vendor"] = COPYRIGHT > > define "app-war" do > package(:war) > end > > end > --------------- snap --------------- > >> since we use timestamps to control what changed, Buildr probably didn't do >> anything the second time you asked, as the modification occurred last than a >> second before. > > No, that is definitely NOT the case. I checked that. Trust me, I made quite > some tests before I came here and asked. > > This was just an example so that it is easy for people to reproduce it. If > you could create this tiny project and try for yourself? > > It happens on Windows and Linux, by the way. > > > Cheers, Ingo =;->
