As I previously mentioned, I committed a change a bit too late last night that
was meant to be a small cleanup of our rather big earlier change, but it
created a few problems with the regression scripts causing them to all fail.
I believe that it is now all fixed in the current SVN trunk (no problems with
Branch-2.4). However, I recommend that everyone who is running nightly
regressions tests to manually do the following if possible. I am not 100%
sure it is required, but it certainly won't hurt.
After doing the above, you will have essentially re-initialized the regression
test base, so that hopefully the nightly regressions will run.
There is still one script that will fail, but it is not serious.
FYI, the regression tests should normally build Bacula using shared objects
for the libraries rather than static compiled in libraries, so there may
still be a few problems to be ironed out.
Thanks for your help testing, please keep it up, we need it more than ever
before in order to ensure that our development version is solid.