In describing the web client/server data cycle, we make a point of
distinguishing between the web server and a data-handling
application. Often, this distinction is implicit -- the client
always has to make an HTTP request in order to send data to the
data-handling application, so it naturally follows that a web server
is involved. In Flash form development, however, we must remain aware
of the invisible handoff between the web server and the data-handling
application. Data moves from Flash to the server either on the end of
a URL (using GET) or in a stream of variable names and values (using
POST). When a web server error is encountered, Flash does not display
the HTTP error messages that the server sends (as a browser would).
For example, if the web server can't find a CGI script, it
sends a "404 Not Found" message, but Flash doesn't
display it. Similarly, if a CGI script's permissions
aren't set correctly, we don't see any execution-failure
error message. In order to isolate client/server problems when
working with Flash, it's useful to monitor the web
server's HTTP error log while attempting to run scripts. You
may find that the web server is trying to tell you something that
Flash can't express.