PR # 39 - Document the imagecraft.yaml configuration file
Merged.
Threat modeling
WIP.
Had some discussions with other teams also doing Threat Modeling.
I am actively trying to borrow some time from some Security Team members to review the current state and make sure I am heading in the right direction.
Implement the spec to add needed features to craft-archives to use it in imagecraft
Done and reviewed once. Comments were fixed. Now waiting for another review.
ubuntu-image
Review several contributors PRs
PR # 227 - Set gpt partition as bootable if associated to system-boot role
Working on a test case
Deep dive into partition/volume/disk handling in ubuntu-image to find root causes for at least 3 classes of bugs. Next Pulse I will implement at least several fixes to cover the majority of the known (and newly discovered) bugs on this.
Updates to the FIPS patches in the openssl packages and @schopin re-reviewed 3.2.1 and uploaded it, thanks!
started looking at the obligatory autopkgtest failures; it’s still unclear how many are flaky tests or testbed failures
Work on crypto-config for system-wide configuration of cryptography
Updated code against latest specification, fixed several small issues
Added temporary code to be able to not be a dependency of the configured packages
Started preparing demonstration profiles
looked at gnutls’ configuration handling to add drop-ins support
looked at nginx’ configuration which prevents disabling TLS versions after they’ve been enabled once
Sent python-google-api-core for review in Debian ; if you want it more quickly in Ubuntu, feel free to pick it up!
Misc
Exactly 0 changes to my better update_excuses.html page but I’m definitely longing for the time it will display updated and past test results! (that’ll be September at the earliest ='( )
Opened PR: #2015 - Fix for the systemd-networkd-wait-online bug (LP: #2063331)
Slowly chipping away at some further improvements we could make to the network handling in Subiquity that I discovered during my investigation of the above.
Resuming my investigation into MOK enrollment and how we can use umockdev to develop and test it in QEMU in the absence of affected hardware.
Merged PR: #2016 - instructions to test netboot with custom ISOs
Subiquity Bug triage and PR review
Distro
Merged MP:#467744 - add netboot tests to the Desktop ISO tracker
Prepared Uploads to SRU the Subiquity Apport hook changes to Noble and Jammy - LP: #2067775
Misc
Finished travel planning for Ubuntu Summit and Engineering Sprint
Completed my first training objective: a beginner course on C. Next up is some more advanced C topics!
I broke the archive a bit when uploading my glibc and base-files merges (once I figure out exactly what happened I’ll try to write it up). Those uploads have been kicked out of -proposed to fix the immediate issue, and I’ll need an AA to look into this bug before I can do a new attempt.
As a consequence of my archive breaking mistakes, I hacked a bit on ppa-dev-tools to add support for pocket selection.
Started working on fixing the apport test suite for non-amd64 architectures (unfinished due to the aforementioned debacle)
highlighting dotnet/aspnetcore/issues/56472: for .NET 9 minified js files will be in a seperate repo and the Microsoft ASP.NET team continues to work on a fully offline source buildable mechanism that may land for >= .NET 10 (it’s crazy to me what a huge mess building from source is when npm is involved :/)
Ubuntu Packaging Guide
MIR team approached me regarding general MIR & packaging documentation (irc log)
reviewed & merged PR #57 – feat(README.rst): add ReadTheDocs link
reviewed PRs #56, #59, #60 – fix(README.rst): visually broken link
merged #60
triaged Issue #58 – docs: PDF documentation broken
reviewed PR #55 – Fixed broken link and add ReadTheDocs link in README
My improvements for netplan status --diff were merged upstream PR#466
Investigated a report about docker builds that started to fail after the last netplan.io updates. It’s not really a netplan.io issue though. LP#2071333
Currently updating my pending SRUs and including the security fixes
continued work on octave-dicom - I have a concern that the use-after-free seen in octave-dicom testing may be a stdc++ bug. The following test is a more minimal version of the failure: