[thelist] GET requests and sent content

[Richard Harb]

G'd evening : )
cc: "thelist at lists.evolt.org" <thelist at lists.evolt.org>
X-BeenThere: thelist at lists.evolt.org
X-Mailman-Version: 2.1b5
Precedence: list
Reply-To: Richard Harb <rharb at earthling.net>, "thelist at lists.evolt.org"
	<thelist at lists.evolt.org>
List-Id: thelist at lists.evolt.org <thelist.lists.evolt.org>
List-Help: <mailto:thelist-request at lists.evolt.org?subject=help>
List-Post: <mailto:thelist at lists.evolt.org>
List-Subscribe: <http://lists.evolt.org/mailman/listinfo/thelist>,
	<mailto:thelist-request at lists.evolt.org?subject=subscribe>
List-Archive: <http://lists.evolt.org/pipermail/thelist>
List-Unsubscribe: <http://lists.evolt.org/mailman/listinfo/thelist>,
	<mailto:thelist-request at lists.evolt.org?subject=unsubscribe>
X-List-Received-Date: Mon, 11 Oct 2004 20:33:28 -0000

Never heard that one before - sounds very 'urban legend'.

Anyway:
When using the GET method the User Agent constructs an URL that is sent to the server containing 
	all enabled variables of the form in which a submit was triggered
	in the form of variable name [or id if name is not present]
	equals 
	urlencoded value of the input/select/textarea/
	delimited by ampersand(s)
nothing more (nothing less).

The POST essentially does the same, but the submitted variable names and their values will not show up in the location bar.

@see http://www.rfc-editor.org/rfc/rfc1866.txt page 46 ff

Maybe I'm missing something?

<opinion>
OTOH client side checking of data before it is submitted is a good addition to server side checking as it may save a couple roundtrips if the data filled into the form is incomplete or inconsistent. Those couple of bytes sent once in the form of javascript may essentially save some bandwidth in the long run when dealing with more complicated forms where something can go wrong ...
</opinion>

HTH

Richard Harb



-----Original Message-----
From: Paul Bennett
Sent: Monday, October 11, 2004, 9:48:21 PM
> Morning all,
> I have recently begun a new job and am involved in building some  web
> applications which will and do get a substantial amount of traffic. The
> applications have been coded so far to use all sorts of client-side JS
> to package and post form requests (some via get, some via post, some via
> the xmlhttp object). I have been told that the focus is on minimizing
> the size of the data flowing between the client and server.

> Now I am definiteley one to avoid javascript when there is a built in
> browser function which does the same thing by default. I have been told
> by the very knowledgable lead developer that normal form get and post
> requests also send back the entire html of the page they are on and 
> hence they use JS to minimize the size of the sent data.

>  This to me seems ludicrous. My understanding is that the browser 
> packages the request and simply sends back the form data to the server.
> If this is true then my job becomes simpler and the web applications can
> instantly become more accessible as currently the js stuff locks out
> some browsers (I hate this) which also doesn't make sense to me in the
> age of standards support and openess that web development is focussing on.

> Can anyone via actual reference confirm or deny my suspicions about the
> browser GET / POST system???

> Many thanks,
> Paul B