🚀 go-pugleaf

In article <Mc92MDA9ABSzEw1Q@objmedia.demon.co.uk>
           snail@objmedia.demon.co.uk "Snail" writes:

> Well,
> 
> after all your very good advice, I thought I'd have a shot at a POP3
> client. Thing is I never get a chance at any negotiation with the server
> as the password phase never seems to get any further. In addition I've
> also run a similar set of tests using turnpike to retreive email via
> POP3 - this also seems to fail to get the email.
> 
> I've just run a test of a POP3 client on pop3.demon.co.uk.
> Prior to this I've sent a message to myself (not using turnpike so that
> turnpike can't shortcut it and avoid sending it). I know this email
> sending program (which is not Turnpike) works, since if I turn SMTP
> receive email on in Turnpike, my email gets delivered.
> 
> Here's the login transcript (I haven't added or removed any 
> characters):-
> 
> Connecting to pop3.demon.co.uk...
> Sent: USER objmedia
> Received: +OK demon POP3 server ready
> Sent: PASS <mypassword>
> 
> ...and thats it, I never get a reply. My password is correct. I'm
> connecting as usual (Win 95 DUN). Can anyone shed any light on this. I
> can't use the Turnpike winsock as the Java runtime can't set the
> connection if that one is used. If I use the standard windows winsock,
> the java runtime can see the connection (indeed! where is that
> documented - nowhere).

You appear to be "speaking before you're spoken too".  As the Mad Hatter
told Alice, this is not always productive.  Adjust your client to *wait*
for the OK welcome *before* sending the USER command.

Moreover, are you sending the correct line termination sequences, as
defined in the RFC?  ISTR that some people have had problems using
certain telnet clients performing this action manually, because of telnet
option negotiation confusing the POP3 server, so ensure that you're not
sending anything like this over the socket.

-- 
Brian {Hamilton Kelly}                                         bhk@dsl.co.uk
    "We've heard that a million monkeys at a million keyboards could produce
     the Complete Works of Shakespeare; now, thanks to the Internet, we know
     this is not true."            Robert Wilensky, University of California