Overview
Request 1227296 superseded
Add avahi-CVE-2024-52616.patch: Backporting 1dade81c from upstream: Properly randomize query id of DNS packets. (CVE-2024-52616, bsc#1233420)
- Created by qzhao
- In state superseded
- Supersedes 1227262
- Superseded by 1227528
-
Open review for
gnome-maintainers
I do appreciate the step away from update-alternatives here, which seems indeed useless - BUT doing so without any word of mention in the changelog, masking under a CVE fix, does not sound right
OK. no problem, thank you. will update.
I remember the original requirement that tiny changes should not detailed record all into the log.
eliminating usage of update-alternatives is not 'tiny' :)
Request History
qzhao created request
Add avahi-CVE-2024-52616.patch: Backporting 1dade81c from upstream: Properly randomize query id of DNS packets. (CVE-2024-52616, bsc#1233420)
gnome-review-bot accepted review
Check script succeeded
Add avahi-CVE-2024-52616.patch: Backporting 1dade81c from upstream: Properly randomize query id of DNS packets. (CVE-2024-52616, bsc#1233420)
FTBFS
btw: with stripping the alternatives: how do you handle the different pythonXXX packages not conflicting?
Somewhat as expected
Do we need the different variants? Should we perhaps switch the package to only build for the "Current standard" python at buildtime?
It's python library/module, they should be fine for all tw supported pythons.
We could move the binaries out and only install them as primary
oh I do not doubt that they are fine, what I'm asking is if it makes sense that we build it for all pythons, would it not suffice with the "main" one. Does any use case exist where one would need/use a non default python module?
Hi:
I have updated in 1227528. Current openSUSE env could compile it successfully without the change of python alternatives. I think this way is correct.
Thank you for review again.