W3C home > Mailing lists > Public > www-lib@w3.org > July to September 1999

Re: HTTP Post Response end of transmission determination

From: Henrik Frystyk Nielsen <frystyk@w3.org>
Date: Mon, 12 Jul 1999 06:51:11 -0400
Message-Id: <3.0.5.32.19990712065111.0364e620@localhost>
To: Wilbur Streett <WStreett@mail.Monmouth.com>, www-lib@w3.org
At 15:55 11/07/1999 -0400, Wilbur Streett wrote:
>I have some clients that have decided to make a financial system based on
>the HTTP protocol.  I discussed the issues associated with HTTP end to end
>acknowledgement a few years ago, but I don't recall exactly were or when,
>so I'm forced to ask here for a reference or direct answers to the issues.
>
>I recall that HTTP responses without a content-length field simply stream
>the response back to the client side.  But I am wondering about how the
>transmission is actually closed, in particular in the case of a persistent
>HTTP 1.1 connection between the client and the server.  How does the client
>know that the response is complete?  If there is a transmission failure, is
>there any specific response back to the server?

The notion of a valid message delimiter is much more important in HTTP/1.1
with persistent connections than in HTTP/1.0 with TCP close at EOF. The
most common way of delimiting messages in 1.1 is by using the content
length header or chunked transfer encoding. Libwww of course supports both.

Henrik
--
Henrik Frystyk Nielsen,
World Wide Web Consortium
http://www.w3.org/People/Frystyk
Received on Monday, 12 July 1999 06:51:43 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 5 February 2014 07:15:16 UTC