I just looked in to this, and it's a few issues.
And that's it - it should work. The whole pipe overflowing thing is a pain, but I don't see a nice way around it unless the pipe is allowed to ignore the chunksize argument when it sees that it has a bunch of data available for it - but that could cause problems for endpoints that really only do want small amounts of data.
© Espruino, powered by microcosm.
Report a problem