Path: utzoo!utgpu!jarvis.csri.toronto.edu!rutgers!tut.cis.ohio-state.edu!killer!wnp
From: wnp@killer.DALLAS.TX.US (Wolf Paul)
Newsgroups: comp.sources.wanted
Subject: Re: Need dial(3C) compatible with HDB UUCP
Message-ID: <8283@killer.DALLAS.TX.US>
Date: 5 Jun 89 22:49:55 GMT
References: <4522@ncrcae.Columbia.NCR.COM> <8244@killer.DALLAS.TX.US> <13695@ncoast.ORG>
Reply-To: wnp@killer.Dallas.TX.US (Wolf Paul)
Distribution: na
Organization: The Unix(R) Connection BBS, Dallas, Tx
Lines: 71

In article <13695@ncoast.ORG> allbery@ncoast.UUCP (Brandon S. Allbery) writes:
>As quoted from <8244@killer.DALLAS.TX.US> by wnp@killer.DALLAS.TX.US (Wolf Paul):
>+---------------
>| In article <4522@ncrcae.Columbia.NCR.COM> heath@ncrcae.Columbia.NCR.COM (Robert Heath) writes:
>| >Does anyone have a dial(3C) library that is compatible with HoneyDanBer UUCP ?
>| 
>| Yes. I recently was looking at HDB source at a client's with a source license,
>| and there is a comment in there to the effect that HDB does not use dial(3).

This quote is misleading -- I was not saying, "Yes, I have a HDB-compatible
dial(3) library" but rather, "Yes, AT&T abandoned dial(3), HDB does not use it."


>+---------------
>
>"Curiouser and curiouser."  The system I use supports an HDB dial(3) -- but
>a few versions behind the rest of the system.  For example, it doesn't know
>about the ",M" subfield in Devices, although it knows about Dialers et al.
>
>+---------------
>| However, if all you need is superficial compatibility, primarily with
>| the HDB lockfile scheme, it would probably be not too difficult to
>| modify the dialinfo library by Gene Olson, which can be found in Volume 6
>| of comp.sources.unix, at your nearest friendly archive site.
>+---------------
>
>Yes, but what of us who want to be able to use dial(3) to access HDB services
>and don't have HDB source?  (I've written a terminal program which keeps its
>noise in the status line, rather than scribbling on the screen the way cu
>does, and invokes rz/sz to do file transfers.  Unfortunately, not having
>dial(3) use HDB facilities means that I'm going to have to do a lot of extra
>work to make it a reasonable replacement for "cu" -- boo, hiss, AT&T has a
>lot of gall to make it etremely difficult to improve on that ugly beast cu!)

I suspect that with a complete set of man pages (which describe the contents
of the HDB data files) one could come up with a more than superficially 
compatible modification to dial(3).

Basically, as I understand it, it works like this:

1. You can have multiple sets of files, a different one for each program,
   i.e. cu, uucico, as well as your own. $LIBDIR/Sysfiles lists these.
   So the first thing is to look up your program name in $LIBDIR/Sysfiles.

2. If a system name is specified, you next look into the appropriate Systems
   file and find the first entry for that system. Grab all the info from it.

3. Next look in the Devices file for an entry which matches the dev and 
   baud fields from the Systems entry. Grab its info, which should include
   the name of the dialer.

4. Go look into the Dialers file to find th chat script for your modem.

5. Now, check the lockfile for the device, and if it does not exist (
   if the PID it contains is dead) create your own and write your PID
   into it in the form "%10s\n\0".

6. And finally, using the info from all of the above data files, 
   talk to your modem and place the call. If it fails go back to step
   two and grab the next entry for that system, etc.


7. If no system is specified, but a phone number and speed, just skip step
   two and go directly to step three.

I am sure this is not complete, but it is a starting point. My problem is
that I don't currently have time to do this.
-- 
Wolf N. Paul * 3387 Sam Rayburn Run * Carrollton TX 75007 * (214) 306-9101
UUCP:   {texbell, killer, dalsqnt}!dcs!wnp
DOMAIN: wnp@killer.dallas.tx.us or wnp%dcs@texbell.swbt.com