- How do I install nemrutco/nova-filterable-metrics?
- Install it with Composer using composer require nemrutco/nova-filterable-metrics. There is no service provider registration, asset publishing, config file, or database migration to run. After installing, chain the filters() method on a Nova Trend, Value, or Partition metric card. Because the package is archived, pin the version in composer.json and confirm the constraint resolves against your Nova and PHP versions.
- Which Laravel Nova versions does this package support?
- The last release, v1.3.0 from 2023-03-31, added Nova 4 support, so Nova 4 is the target. Nova 5 and later are not covered, and since the repository is archived there will be no upstream patch for them. If you are planning an admin-stack upgrade, test the Composer constraint on a branch before committing.
- How do I apply Nova filters to a dashboard metric card?
- Chain ->filters() on the card and pass filters that are already registered on that dashboard. The package renders them as native Nova dropdowns and date pickers above the metric. The whole benefit comes from reuse, so the card can accept the same status or date filter a resource table on that dashboard already uses. With no existing filters in place there is nothing to attach it to.
- Does it work with Trend, Value, and Partition metrics?
- Yes, the fluent filters() call is available on Trend, Value, and Partition cards. Other metric types are not part of the documented scope. Validate each card individually against a known query, since aggregation logic differs between these metric types.
- Is nemrutco/nova-filterable-metrics still maintained?
- No. The repository is archived and the last release was v1.3.0 on 2023-03-31, with 24 stars and zero dependents. Adopting it means adopting and stewarding a small piece of code rather than simply adding a dependency. The realistic mitigation is an MIT-permitted fork with a named internal owner.
- Does Nova support metric filtering natively now?
- Check the current Nova documentation first, because if Nova has shipped native filter support for metric cards, this package adds nothing and the decision collapses to zero. The package only bridges two existing Nova extension points, filters and metric cards, so a native equivalent makes it redundant. That check should be the first step of any evaluation.
- Will Nova filters applied to a metric card produce correct numbers?
- Filters are passed into the metric's query, so correctness depends on how each card's aggregation interprets them. A filter that behaves well on a resource table does not automatically translate into a meaningful aggregate. A branch with one card wired to one filter plus a smoke test against a known query is the cheapest way to catch silent mis-filtering that produces plausible but wrong numbers.
- Can I use nemrutco/nova-filterable-metrics on Laravel 11 or 12?
- Not out of the box. The last release targeted Nova 4 and the Laravel 9/10 era with PHP 8.0+, and no upstream updates will arrive because the repository is archived. Whether the Composer constraint resolves depends on your full dependency graph, so test it before an upgrade. If it blocks, a fork with widened constraints is the usual fix.
- How do I test that filters are applied to a Nova metric?
- Wire one existing filter to one card on a branch and compare the metric output against the same filtered query run manually. Also compare it with the filtered resource table on the same dashboard, since a mismatch is exactly the problem the package exists to solve. Keep that smoke test in CI so a Nova or PHP upgrade surfaces a break.
- What is the effort to reimplement filterable Nova metrics myself?
- The package is a thin adapter that adds a fluent filters() call to Trend, Value, and Partition cards, so a reimplementation is measured in days rather than weeks for a Nova-experienced developer. That in-house cost is the real comparison against adopting an archived package, because a fork is an indefinite internal maintenance line. If you have no Nova dependency or your metrics live in a separate BI tool, neither option is relevant.