342 open issues clogged the repository’s tracker within 48 hours of the v4.0.0 release in January 2026, marking a 140% increase in bug reports compared to the previous stable branch. According to Jagat Review, the transition forced a 15% spike in CPU utilization across legacy clusters that were previously idling at 22%. This was not a clean migration; it was an infrastructure overhaul that many teams were unprepared for when the update hit production environments. The changelog omitted the 9.8 CVE severity rating found in the core authentication module, a vulnerability that existed for twelve days before a patch was discussed in private maintainer channels. By the time the fix arrived, unauthorized API access affected 2,400 enterprise endpoints.
The financial weight of version debt
Moving from v3.x to v4.0.0 required rewriting 42% of existing configuration files due to the sudden deprecation of the YAML parser. Production telemetry from mid-February 2026 showed that 1 in 5 deployments failed during the initial rollout phase because of silent failures in the environment variable mapping. While marketing materials focused on speed, the reality for on-call engineers involved 18.4% higher memory overhead in containerized environments. These numbers reflect a shift in architecture that prioritizes feature density over operational stability, leaving infrastructure teams to manage the fallout of unoptimized garbage collection cycles. Engineers spent 72 man-hours debugging a memory leak that the documentation claimed was resolved during the beta phase.
Unmerged PRs and latency realities
Data from late February stress tests indicates a 200ms latency floor added to the API gateway, a metric completely absent from the official release notes. We see 84 critical pull requests remain unmerged as of March 08, 2026, despite the software reaching “General Availability” status weeks ago. The disparity between the promised performance gains and the 12% decrease in throughput observed in high-traffic nodes suggests that the internal testing suites failed to account for real-world load. This gap resulted in 64 hours of cumulative downtime for early adopters who trusted the initial versioning logic. The migration cost is not just a number on a spreadsheet; it is the $40,000 in monthly egress fees generated by unoptimized logging calls hardcoded into the new kernel, which triggered 1.2 million unnecessary requests per hour during peak traffic.
Gemini 3 Pro is no longer available. Please switch to Gemini 3.1 Pro in the latest version of Antigravity.
Gemini 3 Pro is no longer available. Please switch to Gemini 3.1 Pro in the latest version of Antigravity.