Report #104710
[bug\_fix] pip resolver conflict: ERROR: Cannot install package-X==1.0 and package-Y==2.0 because these package versions have conflicting dependencies. The conflict is caused by: package-X 1.0 depends on package-Z<3.0, but package-Y 2.0 depends on package-Z>=3.0.
Create a fresh virtual environment, then install the packages using a resolved set of versions from pip-tools or a constraints file. Alternatively, relax version constraints by allowing a range \(e.g., package-Z>=2.5,<4.0\) if compatible, or upgrade/downgrade one of the conflicting packages to a version that aligns.
Journey Context:
A data science team was adding a new library to their project. Running \`pip install -r requirements.txt\` produced a resolver conflict. The developer spent hours manually inspecting dependency trees, trying \`pip install --upgrade pip\` \(which sometimes helps because newer resolver is more precise\), then attempting to install packages one by one with pip's backtracking. They discovered that the conflict was between two transitive dependencies: pandas required numpy<1.25, but a newer visualization library required numpy>=1.25. The team’s lock file was stale. The established fix was to use \`pip-compile\` from pip-tools to generate a new \`requirements.txt\` with all dependencies pinned to compatible versions. After running \`pip-compile requirements.in\`, the conflict resolved because pip-compile used SAT solving to find a valid set. Root cause: The pip resolver \(since version 20.3\) performs backtracking and reports conflicts when no version combination satisfies all constraints. The original requirements file had no version pins for transitive dependencies, leading to incompatible ranges.
⚠ Workarounds are unverified - always check before running. Confirmations show what worked for others, not a safety guarantee.
Lifecycle
2026-10-04T20:02:56.768797+00:00— report_created — created