Blog
Notes from building a file transfer product.
Engineering write-ups with the numbers we measured, protocol explainers, how to read a benchmark, and why the plans are shaped the way they are. Sourced, dated, no customer names.
2 min readFive questions to ask about any file transfer benchmark
Vendor benchmarks, ours included, are only useful if you can tell what was measured. Five questions that separate a result you can reason about from a number on a slide, and how our own published test answers them.
2 min readHow we store a 420,000-file browser upload as one object
An engineering note on bundles: why one file per uploaded file fell over at scale, how browser uploads now write straight into a single bundle at their final offsets, and what that did to finalize time and download speed on our test node.
2 min readWhat actually happens when you send a folder from a browser
A walk through the Web Edge upload path: reading the folder in Web Workers, batching small files into one request body, six batches in flight, the origin check on the node, and where the time goes on a home connection.
3 min readWhy a gigabit line delivers a fraction of that across an ocean
The arithmetic behind slow long-distance transfers: how latency and packet loss cap a single TCP connection regardless of link speed, with the formula, a worked example, and what a UDP transport changes.
2 min readWhy we do not charge for the people you send to
Per-recipient and per-gigabyte pricing punishes the teams with the most clients and the biggest files, which are the teams a transfer product exists for. How the plans are shaped instead, and what that means for an agency, a studio or a lab.