Re: git: 66f36c368676 - main - packages: Add meta-package sets [ a pkgbase poudriere jail creation got a populated /usr/local/poudriere/jails/main-amd64/usr/src/ ], not just sys/
- Reply: Mark Millard : "Re: git: 66f36c368676 - main - packages: Add meta-package sets [ a pkgbase poudriere jail creation got a populated /usr/local/poudriere/jails/main-amd64/usr/src/ ], not just sys/"
- In reply to: Mark Millard : "Re: git: 66f36c368676 - main - packages: Add meta-package sets [ a pkgbase poudriere jail creation got a populated /usr/local/poudriere/jails/main-amd64/usr/src/ ], not just sys/"
- Go to: [ bottom of page ] [ top of archives ] [ this month ]
Date: Sun, 14 Sep 2025 18:18:23 UTC
Mark Millard: > On Sep 14, 2025, at 09:04, Lexi Winter <ivy@freebsd.org> wrote: > > Mark Millard: > >> Lexi Winter <ivy_at_FreeBSD.org> wrote on > >>> packages: Add meta-package sets > >> [poudriere installs more packages than before] > > [...] > > please file a PR against poudriere for that. > >[...] > Seems odd that I should be the one specifying interface > changes between how pkg/pkg-static is used by FreeBSD > and poudriere-devel and suggesting how they be handled, > especially by using words someone else supplied. i don't expect you to specify the interface between (pkg)base and poudriere, but understand that we're on a very tight timeframe to get pkgbase ready for the 15.0 release, and i have several other fixes that i need to land before that can happen and limited time to work on those. so, since you discovered this issue, and have a system that can reproduce the issue and test a fix, i thought it would be reasonable for you to report the issue to poudriere. the alternative is that i will put this low-priority issue on my TODO list and get around to reporting it at some point, but that might not be before 15.0-R and it might be not in the timeframe you were expecting the issue to be addressed. i don't want to seem dismissive here, because you have found an actual issue that should be fixed, but i think it's reasonable to assume that someone who runs CURRENT or STABLE would be willing to report issues in third-party software without needing me to act as an intermediary. this doesn't mean you need to take on the responsibility of designing the fix for this issue; if there's really a need for more input from me here i'm sure Bryan is quite capable of handling that.