Hi tony, [...] > I'm not able to reproduce the autopkgtest failure locally running in > clean sid chroots. First, I build the dojo source package and ran the > autopkgtest against those binaries. When that didn't fail, I pulled the > binary packages from the archive and ran the autopkgtest against those. > Again, no failures. > > I see the autopkgtest failure when I run against a bookworm chroot. > > So it seems like the migration of rhino will resolve the test failure. > (Or I'm missing something fundamental.)
Strange. I downloaded the source package and ran the autopkgtests manually. I symlinked js.jar and shrinksafe.jar into util/shrinksafe and then I executed the runner.sh script. I got the same error message "Cannot set property "dojo" of null to "[object Object]". Anyway, are the autopkgtests really useful if they prevent rhino from migration to testing every time we update the package, even if everything works as expected? The same tests already run at build time.
signature.asc
Description: This is a digitally signed message part
