# Backward looking:
Folks, it is ... well not good that this got stuck and lost in between a
great backport (thanks Trent), several users testing (thanks Linus and
Kevin and whom I might miss).
Trent (thanks again) went further than normal by finding, setting up and
running the upstream test suite - yes there are issues but they are not
regressions as far as we know.
But
It should be a bug on its own, even more important for resolute than jammy to
look at what the upstream testsuite says there and if we can/need to improve
anything.
I think what really took the energy from this was the mention in comment
19 to no more be required. And while that is true for whatever the
original reason was that isn't for many users that we see here on the
bug and more that I know and probably even more that haven't identified.
So all that is good right?
And also if I might be more bold for a minute, a regression against what are we
talking about here.
Let us also see the alternative reality - If it would not have been removed for
an FTBFS it would not have been tested by anyone except a few dep8 tests which
are good and just be there.
And in our actual reality where we are not replacing a formerly working with a
potentially non-working version. We are replaying "nothing" with a likely
working well version (and surely better than nothing).
I'd say let us pick this case back up and get it over the finishing line
together please, for the benefit of everyone and not mourn over why it
hasn't been for so long :-/
---
# Forward looking:
I've talked to some more people ...
Weii Wang offered yet another real-use case test of the basic functions
in a real setup - thanks!
Weii also offered to help having a into the failures reported by the
test suite.
I'll ask Jonas to join this effort - especially having a look at
comparing pre-jammy, jammy and resolutes behavior in regard to the
testsuite. I'm quite hopeful that this might either show something to
fix or more likely confirm we are not looking at a regression anyway.
When these results comes in and are not shattering what we think we
know, it might be time to release it then - verified-done here we come.
And if it truly shows that it is hard-broken, well then we know instead
of continuing this "held back by uncertainty" state.
---
# And action:
I'll re-assign bug-tasks like this:
- Weii to the Jammy case for yet another verification as he offered. As new
smoke test to run the basic functions and exercise most components? Create a
test that installs everything, runs a backup, and performs a restore, which
should exercise most general components of bacula (bacula-fd, bacula-sd,
bacula-dir, bconsole).
[As much as you can join Jonas in a look at the testsuite]
- Jonas - please have a look at the testsuite like Trent did. But also
do so Jammy-1 (were they regressions or always there or new tests) and
Resolute (we might want to know that for the future).
** Changed in: bacula (Ubuntu)
Assignee: Trent Lloyd (lathiat) => Weii Wang (weii-wang)
** Changed in: bacula (Ubuntu Jammy)
Assignee: Trent Lloyd (lathiat) => Jonas Jelten (jj)
** Tags added: server-todo
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1973322
Title:
Bacula for 22.04/Jammy
To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/bacula/+bug/1973322/+subscriptions
--
ubuntu-bugs mailing list
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
