RenderCache Ships Its First Alpha, a Self-Hosted Cache That Serves Crawlers Rendered Pages

RenderCache, the studio’s tenth project, is a pre-render cache for public JavaScript sites that runs on your own server. It renders public pages in Chrome, keeps the finished HTML and hands that copy to crawlers, while visitors get the normal app. The first alpha, 0.1.0-alpha.1, came out on 16 October for Linux on x86-64: one Go executable, with Chrome running beside it. The code is on GitHub under Apache 2.0.

The hard request is the first one, for a page with no saved copy yet. RenderCache doesn’t keep a crawler waiting on Chrome: a cold page comes from the origin, maybe still an empty shell, and renders in the background. So it warms pages from the sitemap ahead of time, refreshes them in the background and keeps the older copy if the origin fails for a moment. For each request from a recognized crawler it also writes a receipt with the status, the cache outcome and the bytes written. It’s a record of what left the server. What the crawler did with the page, it can’t say.

It’s an evaluation release, and the project makes no claim that it’s ready for production. It supports one process per data directory and keeps its queue in memory, and so far there’s only the Linux build. What comes next depends on what real deployments turn up. The roadmap names a lighter browser engine, Lightpanda, and a tie-in with BareProxy among possible directions, with no dates.