I'm trying to write a test server to work with google chrome's native messaging. all messages are passed as a unsigned 32-bit number followed by that many bytes of utf-8 encodes json.
Ok so I read 4 bytes, and unpack my length, read length bytes and decode using utf8::upgrade and a json parser (tbd). And I take the response I generate, encode it as json determine length, and pack.
But the devil is in the details. perldoc -f pack is unclear as to the size of 'i'. Some places it appears this may vary by processor, and could be 32 or 64 bits. other places imply that it is 32 bits only. Also I am uncertain as to the best way to encode the response string should I use utf8::encode, pack, or something in Encode::
Your advice is appreciated.