HTTP, like most of the original Internet protocols is based on transmissions of text messages. This was partly because the original nodes of the Internet had different internal binary representations, and quite a few of them had different character codesets (ASCII or EBCDIC). It's much easier to translate codesets than binary fields, especially when you're bouncing a message through a dozen intermediate computers or more.
There's another fringe benefit - it's easier to debug the protocol if you can directly print eveything going up and down the wire - or even drive it manually via a telnet to port 80.
You can write Internet apps all day long, but you'll never truly be an Internet Guru until you know what the format and sequence of those messages are. They're documented in a set of RFC (Request For Comment) documents, and the one for HTTP v1.1 is RFC2068:
http://www.w3.org/Protocols/rfc2068/rfc2068 The short answer tou your question, though, is that if you have a hyperlink or a form whose action is "GET", a message in the following form is sent:
GET
http://www.myserver.com/mypage.cgi?arg1=foo&arg2=bar If a form has action POST, it will look more like:
POST
http://myserver.com/mypage.cgi arg1
foo
arg2
bar
There may be some inaccuracies here, and I haven't bothered to show the headers, but that's the core of it. The RFC is the authority.
[ January 11, 2002: Message edited by: Tim Holloway ]
