Path: utzoo!attcan!uunet!mcvax!hp4nl!hpuamsa!frank
From: frank@hpuamsa.UUCP (Frank Slootweg)
Newsgroups: comp.sys.hp
Subject: Re: IENQAK on HP-UX 2.10 ???
Message-ID: <7810012@hpuamsa.UUCP>
Date: 1 Mar 89 15:19:09 GMT
References: <16@pttdis.UUCP>
Organization: HP NL
Lines: 23

Bob Niland (rjn@hpfcdc.HP.COM) wrote:

> Even if the 800 supported IENQAK, it wouldn't help you, since the HP-UX
> (500) implementation was host-mode only, i.e. the host driver would
> issue an ENQ and wait for an ACK before sending a block of 80 characters.

  I missed that point (which was made by Mr. van Houten).

  Mr. van Houten called me back. He had expected "ienqak" to be a input
handshake. I guess its name was chosen based on "ixon" ("What do you
want? UNIX *or* sensible option names?" :-)).

> All is not lost, however, since ENQ/ACK is a half-duplex protocol, it is
> relatively easy to implement in user code.  Have your application perform
> input in single-character or raw mode, and look for ENQs.  ACK them.

  The customers application (actually a server) can indeed do the
handshake (i.e. it does not *have* to be done in the driver). I advised
him to set EOL/VEOL (see "termio(7)") to ENQ. A "line" will then be
terminated by either a newline or an ENQ and single-character or raw
mode is not needed.

Frank "Xoff" Slootweg, Hewlett-Packard, Dutch Country Response Center.