In a message dated 01-07-16 23:17:20 EDT, you write: > >> MSIE is choking due to a '0' byte happening in the middle of the > >> response being sent back. Ah... okay... so you mean something is 'truncating' the data and adding the terminator chunk length too early? The last data buffer is actually 'missing' or something? That shouldn't really bother most browsers unless it's a graphic in which case the presentation layer stuff culd blow trying to do the GIF LZ77 decode or the JPEG unwind. > Also, I don't think we update Content-Length: in the chunked trailers > yet. > Chuck Does anyone? I've looked at a LOT of real-time captures from all over the world and I don't recall ever seeing anyone add 'Content-Length:' back to the Entity Headers at the end of a chunked transmission. Actually... to tell the truth... I don't think I've ever seen anyone anywhere use that part of the HTTP spec whereby you can add/update header fields at the end of the transmission. I also don't know of any clients that would have the faintest idea how to handle that ( Headers fields coming AFTER all the data arrives ). It's usually way too late for that part of the RFC to have any effect at all on what's happening unless the Entity header pertains to what's supposed to happen AFTER the data arrives ( Like changing 'Connection: Keep-Alive' to 'Close' somewhere during the chunked transfer ). Kevin Kiley
