!!! Multiple package instances within a single package slot have been pulled !!! into the dependency graph, resulting in a slot conflict: dev-python/resolvelib:0 (dev-python/resolvelib-0.8.1:0/0::gentoo, ebuild scheduled for merge) USE="-test" ABI_X86="64" PYTHON_TARGETS="pypy3 python3_10 python3_8 python3_9" pulled in by =dev-python/resolvelib-0.8.1 (Argument) (dev-python/resolvelib-0.5.4:0/0::gentoo, installed) USE="" ABI_X86="64" PYTHON_TARGETS="python3_10 python3_8 python3_9" pulled in by <dev-python/resolvelib-0.6.0[python_targets_python3_8(-),python_targets_python3_9(-),python_targets_python3_10(-)] required by (app-admin/ansible-base-2.13.0-1:0/0::gentoo, installed) USE="-test" ABI_X86="64" PYTHON_TARGETS="python3_10 python3_8 python3_9" ^ ^^^^^
what do you want me to do about it? https://github.com/ansible/ansible/blob/stable-2.13/requirements.txt#L13
Perhaps you could start by testing it with the newer version. Or at least reporting the problem upstream.
searching the ansible bug list shows that they have work ongoing to update resolvelib https://github.com/ansible/ansible/issues/74569
This now became more important, as new version of dev-python/pdm requires new resolvelib.
Looks like it's finally happening, just this last week. https://github.com/ansible/ansible/pull/84218 I asked if there were other changes but I don't think so. At the very least the next version should have this change I think.
Unfortunately it looks like that didn't make it into the latest release and no update on if relaxing the limit alone is ok (as per the merge) or if it needed other work.
The bug has been closed via the following commit(s): https://gitweb.gentoo.org/repo/gentoo.git/commit/?id=b2210100652a0bd0162bcfb633121215f13457c6 commit b2210100652a0bd0162bcfb633121215f13457c6 Author: idealseal <realidealseal@protonmail.com> AuthorDate: 2025-03-24 13:50:29 +0000 Commit: Sam James <sam@gentoo.org> CommitDate: 2025-03-25 04:30:17 +0000 app-admin/ansible-core: drop 2.16.6 Closes: https://bugs.gentoo.org/847604 Closes: https://bugs.gentoo.org/848486 Closes: https://bugs.gentoo.org/921642 Signed-off-by: idealseal <realidealseal@protonmail.com> Signed-off-by: Sam James <sam@gentoo.org> app-admin/ansible-core/Manifest | 1 - app-admin/ansible-core/ansible-core-2.16.6.ebuild | 53 ----------------------- 2 files changed, 54 deletions(-)
Request to reopen bug. Dropping 2.16.6 does not resolve the issue: (dev-python/resolvelib-1.1.0:0/0::gentoo, ebuild scheduled for merge) USE="-test" PYTHON_TARGETS="python3_12 (-pypy3) (-pypy3_11) -python3_10 -python3_11 -python3_13" conflicts with <dev-python/resolvelib-1.1.0[python_targets_python3_12(-)] required by (app-admin/ansible-core-2.18.1-r1:0/0::gentoo, installed) USE="" PYTHON_TARGETS="python3_12 -python3_11 -python3_13" ^ ^^^^^ # ack resolvelib /usr/portage/app-admin/ansible-core/ /usr/portage/app-admin/ansible-core/ansible-core-2.18.1-r1.ebuild 40: >=dev-python/resolvelib-0.5.3[${PYTHON_USEDEP}] 41: <dev-python/resolvelib-1.1.0[${PYTHON_USEDEP}] /usr/portage/app-admin/ansible-core/ansible-core-2.18.3.ebuild 40: >=dev-python/resolvelib-0.5.3[${PYTHON_USEDEP}] 41: <dev-python/resolvelib-1.1.0[${PYTHON_USEDEP}] /usr/portage/app-admin/ansible-core/ansible-core-9999.ebuild 36: >=dev-python/resolvelib-0.5.3[${PYTHON_USEDEP}] 37: <dev-python/resolvelib-1.1.0[${PYTHON_USEDEP}]
Those are different versions. This bug isn't (I don't think) about the general issue of ansible pinning aggressively.
My interpretation of comment 4 is that ansible's blocking of resolvelib-1.1.0 is causing grief; my interpretation of comment 5 is that a "next version" fixes this. Unfortunately, AFAICT commit 771f7ad29ca is not tagged in any released versions yet (up to and including v2.18.4) but I may have missed something. Anyhow, I would think that this bug needs to remain open until such time as ansible release X unblocks resolvelib < 1.1.0 and the corresponding changes make their way into an ebuild?
https://bugs.gentoo.org/848486#c0 showed pinning to <0.6.0, though. But you're right that https://bugs.gentoo.org/848486#c4 onwards is discussing the same problem again with a newer version.
