Hello to folks interested in either yahns and kcar. For those of you unfamiliar with either, kcar is currently an HTTP response parser which converts arbitrary bytestreams to Rack-compatible response status and headers (and optionally body chunk filtering).
yahns currently uses an HTTP request parser from a well-known HTTP server; but kcar for the optional reverse HTTP proxying capability. kcar is derived from the HTTP request parser used by yahns; so there is enough common code between HTTP 1.x request and response parsing to stick them into the same object. Using kcar for request parsing will transparently reduce icache footprint and allow wider object recycling (should yahns go down that path); as well as reducing installation time and footprint. Since neither project is particularly popular at the moment, there'll be more room to experiment with potential performance improvements and memory usage reductions. https://yhbt.net/yahns/README mail archives: https://yhbt.net/yahns-public/ git clone git://yhbt.net/yahns https://bogomips.org/kcar/ mail archives: https://bogomips.org/kcar-public/ git clone git://bogomips.org/kcar Fwiw, yahns has been hosting both sites since late-2013. -- unsubscribe: yahns-public+unsubscr...@yhbt.net archive: https://yhbt.net/yahns-public/
