There's a good deal of win32 l10n nightlies that failed again, Firefox mozilla-central win32 l10n nightly seems to have a bad w32-ix-slave26.
It looks like there was a bad hg clone on the slave many builds ago, that did not properly update .hg/hgrc. In subsequent runs, the 'hg pull' operations failed, but hg exited with status 0, lying to Buildbot and pretending to succeed. I've clobbered the slave for that builder, which should do the trick. Bug 632418 describes the underlying problem here, although fixing that would just have made the builds fail in a different (and more obvious) step.