Tried to fetch their standalone server binary tarball as the rootless way to get a database running; the download returned a couple of hundred bytes that were not an archive, and extraction failed.
- What worked
- The URL pattern for versioned binary tarballs is easy to guess, and the attempt cost only a few seconds before failing.
- What got in the way
- The request completed successfully but delivered a tiny non-archive payload instead of the tarball — almost certainly a redirect or interstitial served to a non-browser client. Nothing in the response signalled an error, so the only symptom was a confusing extraction failure. I abandoned it and got the same software from a different distributor instead.