Fix: Old Version Label v1.0 in Comparison Tables

If you’ve been using vibe-coding-universal for benchmarking or cross-environment comparisons, you may have noticed a discrepancy in how version labels appeared in generated tables. The latest update addresses this: comparison tables now correctly display "v1.0" instead of stale old version labels. This might seem minor, but for teams relying on accurate version metadata to interpret results, it’s a necessary correction.

The Context

vibe-coding-universal is a lightweight framework for running and comparing code snippets across different runtimes, interpreters, or hardware configurations. It outputs structured comparison tables that pair metrics—like execution time, memory footprint, or throughput—with their respective versions. These tables are meant to provide clarity at a glance, especially when iterating on performance optimizations or testing new runtime releases.

However, before this fix, the version label in these tables could fall out of sync after an update. If you built the tool from source or applied a patch, the comparison tables might still reference an older version string from the configuration or cache. This led to confusion: developers saw "0.9.3" or similar in the label while the actual codebase was already at v1.0.

What Was Fixed

The changelog entry reads: fix: old version label → v1.0 in comparison tables. In practice, this means the label generation logic now pulls the version directly from a canonical source—the project's __version__ attribute or a version-manifest file—instead of relying on a cached or hardcoded fallback. The fix ensures that every time you run vibe-coding-universal to produce a comparison table, the version column header reads "v1.0" if that’s the current release.

Changes were isolated to the table rendering module. Previously, version label assignment happened during initialization and was never refreshed after a version bump. Now, it’s dynamic: the version string is retrieved at render time from the installed package metadata.

The Code Change

Here’s a simplified version of the core change that went into the table rendering logic:

# before (cached version)
class ComparisonTable:
    def __init__(self, metrics):
        self.version = get_cached_version()  # could be stale
        self.metrics = metrics

    def render(self):
        return Table(
            header=["Metric", f"Version {self.version}"],
            rows=[(m.name, m.value) for m in self.metrics]
        )

# after (dynamic version)
class ComparisonTable:
    def __init__(self, metrics):
        self.metrics = metrics

    def render(self):
        # always fetch current version at render time
        current_version = importlib.metadata.version("vibe-coding-universal")
        return Table(
            header=["Metric", f"Version {current_version}"],
            rows=[(m.name, m.value) for m in self.metrics]
        )

The key difference is moving from a cached version attribute to a fetch on each render. importlib.metadata.version reads directly from the packaging metadata, so it reflects the actual installed version, not an intermediary label. This eliminates stale references entirely for standard use cases.

Why This Matters for Developers

• Reproducibility: If you share a comparison table from vibe-coding-universal, the version label now unequivocally identifies the environment. No more guessing whether the label matches the actual codebase.
• CI/CD accuracy: In automated benchmarking pipelines, version labels are often used to tag results or trigger downstream processes. A wrong label breaks that chain. This fix keeps those integrations reliable.
• Minimal overhead: The change has negligible performance impact—metadata lookup is cheap compared to the benchmarks themselves. The only trade-off is that the table now requires the package to be installed in a standard way (pip, wheel, etc.), which is already the expected deployment method.

This is the kind of fix that experienced developers appreciate: it removes a subtle, non-obvious source of error without adding complexity.

The Bigger Picture

vibe-coding-universal is built for developers who need transparency in cross-version comparisons. This update aligns the tool’s output with that goal. If you rely on comparison tables for performance reports or regression detection, pull the latest release and rebuild any tables that depend on version labels. The "v1.0" tag is now the source of truth, not an artifact of stale caching.

Beyond this fix, the project continues to emphasize minimal configuration and predictable output. Keep an eye on future updates for additional rendering options and metric plugins.