[thelist] File delivery system?

[Bill Moseley]

On Sat, Mar 24, 2007 at 03:03:39PM +0000, kasimir-k wrote:
> I would have thought that if the connection is still open after the file
> has been output to the client, then I could be pretty sure that the
> client has received the file.

No, you can't assume that.

> > Are you sure there isn't a proxy sucking up the file and closing the
> > connection before the download completes?
> 
> And the proxy would keep the file to itself and not pass it on to the
> client? Is this really possible?

The proxy could close the connection to your server but then the
download could still be aborted between the person downloading and the
Proxy.  A proxy's (and reverse proxy's) role is often to buffer between
slow clients and resource intensive servers to free up the servers to
serve other clients.  Think about AOL, for example.


> And after reading
> http://www.w3.org/Protocols/rfc2616/rfc2616-sec14.html#sec14.44
> I'm under the impression that the server sends this to the client, and a
> proxy shouldn't add this on its own.

Right, shouldn't is the appropriate word.  There's a lot of stuff that
IE shouldn't do but does do.  Just pointing out that things happen
beyond your control.  It's the Internet, after all. ;)


> But if the idea is to limit the number of downloads, then both of these
> approaches fail - the user could download the file many times over the
> given period, and reply "no" to the question.

And that's different than them downloading it once and copying it multiple
times on their own machines?

-- 
Bill Moseley
moseley at hank.org