Final release and maintenance status

ResearchPlot maintenance ended on 2026-10-01. 2.0.1 is the final release of researchplot-venues. The published distributions, documentation, signed release tag, and MIT-licensed source remain available for existing users and forks.

No further feature, venue-profile, dependency, or security updates are planned. Upstream issues and pull requests may not receive a response. There is no support or security-response commitment. See the security policy.

What the final release changes

Version 2.0.1 closes two project-verdict gaps: missing required deliverables or referenced evidence cannot produce a passing project result, and a complete Project.bundle() checks the coverage-aware project result before committing the submission directory. The bundle gate includes staged copied or generated artifacts and supplied live figures. Required failures use CLI exit code 1; missing required evidence uses 3. Input and capability errors continue to use 2.

The release does not complete deferred 2.0 roadmap work. The limitations guide documents the narrow submission-manifest bridge, conservative manuscript matching, advisory visual diagnostics, and optional registry setup. A report establishes only the encoded rules for the selected historical profile, not editorial acceptance or the completeness of current venue guidance.

Venue profiles are historical evidence

The bundled profiles retain their exact coordinates, digests, source URLs, and verification dates. They are not being revalidated after maintenance ends. Compare those records with the venue's current official author instructions before a new submission.

Bare conference aliases such as cvpr continue to select an installed 2026 profile. They do not advance to the current or next conference year. Year-pinned conference profiles remain immutable and do not receive publisher-style age warnings. Publisher freshness warnings are a prompt for review, not proof that a source is still correct.

The optional registry client requires a user-configured trusted root and repository. No maintained public ResearchPlot registry or future profile update service is promised. Normal audit, check, export, bundle, and local-web operations remain offline.

Install and preserve a working environment

Use a virtual environment with a tested Python version. The final release was designed for Python 3.11–3.14; compatibility with future Python or dependency releases is not maintained.

python -m venv .venv

Activate it (.venv\Scripts\Activate.ps1 in PowerShell or source .venv/bin/activate on macOS/Linux), then install the exact release:

python -m pip install researchplot-venues==2.0.1
researchplot --version
researchplot profile list
researchplot doctor --profile nature@2026.08.0 --json

For the local workspace, use python -m pip install "researchplot-venues[web]==2.0.1". Select other extras explicitly before recording the environment. Run your actual audit/export workflow and inspect its verdict and unresolved checks before treating the environment as usable.

The package pin alone leaves dependency versions open. Capture the environment and download its install artifacts while they are available:

python -m pip freeze > requirements-lock.txt
python -m pip download -r requirements-lock.txt -d wheelhouse

Store the lock, wheelhouse, Python version, operating system, project configuration, profile lock, source data, and reports with the research project. To reinstall the captured packages into another compatible virtual environment without PyPI:

python -m pip install --no-index --find-links wheelhouse -r requirements-lock.txt

A freeze file and wheelhouse apply to the Python/platform environment in which they were prepared; they do not guarantee installation on a different operating system or Python version. Source distributions may also need build tools or dependencies. Use python -m pip download --only-binary=:all: -r requirements-lock.txt -d wheelhouse when a fully wheel-based capture is needed and all selected packages provide compatible wheels. Check the resulting offline install before relying on it.

Archive the Matplotlib backend, fonts, LaTeX state if used, and exported files as well. Matching package versions alone cannot guarantee identical rendering. Frozen dependencies can contain vulnerabilities: review upstream advisories and assess any updates in an independently maintained environment. Upstream ResearchPlot will not provide compatibility or security fixes.

Continue development in a fork

Fork the MIT-licensed repository, preserve license and attribution notices, and give your fork an explicit maintenance and security policy. Follow the fork development guide for setup, tests, profile governance, and packaging.

A released profile coordinate is immutable. Revalidated or corrected evidence must have a new revision and digest with traceable official sources. Keep uncertain rules unspecified and preserve unresolved checks in reports. Choose new package/publication identities for a separately maintained distribution; upstream PyPI ownership and signing identities do not transfer with a fork.

Before enabling publication or scheduled workflows in a fork, configure its own GitHub environments, Trusted Publisher identities, signing keys, Pages settings, and dependency-update policy. The archived upstream workflows and documentation are reference material, not an ongoing service.