HP3000-L Archives

January 1996, Week 3

HP3000-L@RAVEN.UTC.EDU

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Eric Schubert <[log in to unmask]>
Reply To:
Eric Schubert <[log in to unmask]>
Date:
Fri, 19 Jan 1996 17:08:07 -0500
Content-Type:
text/plain
Parts/Attachments:
text/plain (87 lines)
From: Eric Schubert <[log in to unmask]>
>
>Date:    Wed, 17 Jan 1996 01:23:13 GMT
>From:    Eero Laurila <[log in to unmask]>
>Subject: Re: My public comments to a private message
>
>Eric J Schubert ([log in to unmask]) wrote:
>
>: With the release of MPE 4.0 (1993), allowed direct IP connections from the
>: campus network.  We experimented with a few prototypes and we implemented
>
>- Well this is something that I've always been wondering why...  Why is it
>  that people think that TCP/IP stack came to the hp3000 only recently???
<snip>
>
Sorry,
should have said "allowed the development of TCP/Internet Protocol services
directly"
 
(vt3k Chris - feel free to jump in...)
 
One major stumbling block was the lack of DNS support before MPE 4.0.
 
Internet applications and source code of the day (telnet smtp email and
gopher) were strongly based in open style Unix systems (Sun for example)
that supported an Internet suite of protocols.  MPE was busy being proprietary.
 
Before MPE/Posix, it was worthless to try to port any Internet Unix code to
MPE.  The effort was too great.  May I note that other vendors provided
Internet services (inbound telnet) bundled into their OS for free what HP
charged big money for (TAC is really bad).  May I note that there were
abundant free Internet telnet clients for Mac's, DOS, Unix desktops.  But HP
and partners had a nice proprietary cash cow on the 3k terminal market.
 
That was mid 1992.  Yes, the MPE culture has drastically changed since then
(thank goodness!).
 
I can recant a few milestones of MPE moving toward more open solutions
(others may have more details than this like ALLBASE, etc....):
 
() External DNS was supported in MPE 4.0.  3k DNS still is not.
 
() Berkeley sockets interface was constructed in 4.0.  This was missing the
fork() call so not much could be ported yet.  Finger servers appeared.
 
() Posix started to appear in 4.5.  hello HFS. Hello fork().
 
() ARPA services bundled.
 
() NS/open became available.  Solved some of the expensive client problems.
 
() MPE 5.0 Posix up to speed.
 
() DCE core services/RPC appeared 5.2. But hey, they charge you heavy for
it! Must be the former group that handled ARPA services.  They're back!
 
() jazz.external.hp.com appeared.  Proof of openness had arrived.
 
() today - 3000 vendors actually providing WWW Internet software solutions!
 
() Inbound telnet ... errr, soon.
 
...and others.
 
Ok, Eero.  Start to see a pattern here???
 
A milestone of MPE being "open" was bundled ARPA services .  Why?  It was
more a management decision of corporate direction than anything.
 
The most significant was Internet software BSD ports to the 3000 (after
Posix of course) and that happened only by the extra effort of a few
dedicated HP employees in late 1994 - thanks Mike Belshe!
 
So yes, technically, from your view of things TCP has been around on the
3000 doing all that wonderful proprietary, pricy, ftp, VT, NS/3000, 3000 to
3000 peer stuff...
--
 
The short answer - my meaning was focused on Internet services and without
DNS support, most network engineers wouldn't consider managing it.  All
happened around MPE 4.0 release.
 
----------------------------------------------------------------
Eric J. Schubert                    Senior Data Base Analyst
Office of Information Technologies  Univ of Notre Dame, IN USA
(219) 631-7306                      http://www.nd.edu/~eschuber

ATOM RSS1 RSS2