Tails

#1 Updated by intrigeri 2016-08-28 03:41:41

• Description updated

#2 Updated by intrigeri 2017-07-24 19:43:42

• blocks Feature #11912: Purge Jessie time-based APT snapshots added

#3 Updated by intrigeri 2017-07-24 19:44:51

• Assignee set to anonym
• Target version set to Tails_3.5

anonym, can you take it during the 3.4 cycle (as part of Foundations Team work) so I can do Feature #11912?

#5 Updated by intrigeri 2017-07-29 06:27:32

• Assignee changed from anonym to bertagaz
• Target version changed from Tails_3.5 to Tails_3.1
• QA Check set to Info Needed

Dear RM, would it feel crazy in principle to do this for 3.1? Rationale: as explained in Bug #13526#note-9 this would allow us to do Feature #11912 soon after the 3.1 release, and solve the recurring disk space problems we’re experiencing on the apt-snapshots volume.

In practice live-build does everything in a chroot it creates (that’s already Stretch anyway), so what would be impacted is only what we do in the basebox itself, i.e. building the wiki (FWIW we already install a version of discount that’s almost the one in Stretch), updating PO files, and running debootstrap (we already install 1.0.89~bpo8+1 while Stretch has 1.0.89) so perhaps it’s not such a big deal.

I’ll try to quickly prepare a branch so we have more info wrt. how much work it requires and how broken the resulting ISO is.

#6 Updated by intrigeri 2017-07-29 07:07:11

• Feature Branch set to feature/11738-stretch-vagrant-basebox

#7 Updated by intrigeri 2017-07-29 17:31:15

• Status changed from Confirmed to In Progress
• % Done changed from 0 to 10

I now have an ISO built from this branch, let’s see what the test suite says. Will report diffoscope results later.

#8 Updated by intrigeri 2017-07-29 18:32:16

• blocked by Bug #13530: ISO builds are broken by weird enigmail situation added

#9 Updated by intrigeri 2017-07-29 19:49:44

• % Done changed from 10 to 50
• QA Check changed from Info Needed to Ready for QA

The subset of our automated tests that we run on Jenkins has passed here (on my own Jenkins).

I’ve diff’ed the two ISOs, one built from this branch and one built from the one for Bug #13530 (nothing else builds right now anyway), and the only differences in the SquashFS are in the local copy of the website (expected as we’re using different versions of ikiwiki due to Bug #13531).

This matches my expectations, i.e. upgrading the OS of the basebox has very little impact on the resulting ISO image. So please review’n’merge.

#10 Updated by bertagaz 2017-08-01 16:45:14

• % Done changed from 50 to 60

intrigeri wrote:
> The subset of our automated tests that we run on Jenkins has passed here (on my own Jenkins).
>
> I’ve diff’ed the two ISOs, one built from this branch and one built from the one for Bug #13530 (nothing else builds right now anyway), and the only differences in the SquashFS are in the local copy of the website (expected as we’re using different versions of ikiwiki due to Bug #13531).
>
> This matches my expectations, i.e. upgrading the OS of the basebox has very little impact on the resulting ISO image. So please review’n’merge.

Couldn’t really believe it, but my local builds/diffs say the same (well, apart from some “bleeding” we already noticed). Good there’s no surprise there.

I’m running the test suite on the ISO just in case, I think I’ll merge this for 3.1 if there’s no unexpected failure.

#11 Updated by bertagaz 2017-08-03 12:57:09

• Status changed from In Progress to Fix committed
• Assignee deleted (bertagaz)
• % Done changed from 60 to 100
• QA Check changed from Ready for QA to Pass

Merged into stable and devel, everything seems fine in Jenkins. Nice to see that upgrading our basebox is easy as that and does not have unexpected side effects!

#12 Updated by bertagaz 2017-08-09 12:38:46

• Status changed from Fix committed to Resolved