Archive-Date: Thu, 16 May 1996 15:55:52 MET-1MET DST
Sender: owner-free-vms@lp.se
Message-ID: <199605161334.WAA16035@kekux.kek.jp>
Date: Thu, 16 May 96 22:55:25 GMT+9:00
From: "Makoto Shimojima, Univ of Tsukuba/CDF" <mako%utkbp.decnet@kekux.kek.jp>
Reply-To: Free-VMS@lp.se
To: free-vms@lp.se
Subject: RE: Emerald project

> I have not read it, though; I have just asked a University librarian
> to get a copy of it. Now, if the authors had an online copy...

The hardcopy of the USENIX paper came in yesterday. Since I have just
began reading it, I will not comment on it. The good news is that it
had E-mail addresses of the four authors in it, so I sent them a message
if they had any other papers that I could read. One of the authors,
Bill Davenport, Jr., promptly replied to me and sent me the original
USENIX paper as well as two other (unpublished) papers in PS format.

Now, is anyone interested in reading these articles? Or perhaps I
should send them to the list owner so that he could make them available
in his web?
					mako
					(mako@hep.px.tsukuba.ac.jp)

PS: Bill got interested in this ML and I believe is joining us any
    minute, so we will soon be able to hear first-hand experience!
================================================================================
Archive-Date: Thu, 16 May 1996 19:18:19 MET-1MET DST
Sender: owner-free-vms@lp.se
Date: Thu, 16 May 1996 19:18:16 MET-1MET DST
From: levitte@lp.se
Reply-To: Free-VMS@lp.se
To: Free-VMS@lp.se
Message-ID: <009A26E4.E5AA8E40.1@lp.se>
Subject: RE: Emerald project

>From: "Makoto Shimojima, Univ of Tsukuba/CDF"
>      <mako%utkbp.decnet@kekux.kek.jp>

>Now, is anyone interested in reading these articles? Or perhaps I
>should send them to the list owner so that he could make them available
>in his web?

Please send them on to me (address below)

>PS: Bill got interested in this ML and I believe is joining us any
>    minute, so we will soon be able to hear first-hand experience!

Great!  I bet he has some knowledge to share with us.

-- 
R Levitte, Levitte Programming; Spannvägen 38, I; S-161 43  Bromma; SWEDEN
Tel: +46-8-26 52 47, (via nation.se) +46-8-21 24 90;  No fax right now
PGP key fingerprint = A6 96 C0 34 3A 96 AA 6C  B0 D5 9A DF D2 E9 9C 65
Check http://www.lp.se/~levitte for my public key.
================================================================================
Archive-Date: Thu, 23 May 1996 23:42:41 +0200
Sender: owner-free-vms@lp.se
Message-ID: <9605232131.AA06417@us2rmc.zko.dec.com>
Date: Thu, 23 May 96 17:33:34 EDT
From: Bill Davenport - OpenVMS Security Group <davenport@tabsco.ENET.dec.com>
Reply-To: Free-VMS@lp.se
To: free-vms@lp.se
CC: davenport@tabsco.ENET.dec.com
Subject: Hi

Just dropping a short note into the mailing list.  Chris Kaler and I are both 
interested in what's going on here, and are quite willing to share advice and 
experiences we had in building our prototypes documented in the VMS-on-Mach 
(USENIX) paper and the later Intrepid paper (one of the unpublished papers).

Feel free to post any/all of the papers on the web site.

Chris and I have often times wanted to port our existing code to a new Mach 
base and run it on our PCs.  Given some more work (actually quite a bit more) 
the system would be able to run many standard programs.

Bill

================================================================================
Archive-Date: Fri, 24 May 1996 02:33:56 +0200
Sender: owner-free-vms@lp.se
From: John_David_Galt@cup.portal.com
Reply-To: Free-VMS@lp.se
To: free-vms@lp.se
Subject: Re: Hi
Date: Thu, 23 May 96 17:32:39 PDT
Message-ID: <9605231732.1.17565@cup.portal.com>

I would like to see what has been worked out, especially as far as the
mapping of VMS internals to the Mach kernel.  I have a fair knowledge of VMS
internals, and would like to take on the task of rewriting all the library
routines (system services, LIB$xxx, RMS, etc.), which are my favorite thing
about OpenVMS.

I am new to this list and hope I'm not "butting in" too much, but there
doesn't seem to be a lot of traffic (or action) here, and I'd like to help
make this thing happen.  I have never worked for DEC, so there are no
non-disclosure issues that would keep me from participating.  If I have a
weakness it's that I don't know a lot of details about the Mach kernel; just
what was in the trade press about the time that OSF/1 came out.

John David Galt
================================================================================
Archive-Date: Fri, 24 May 1996 06:18:55 +0200
Sender: owner-free-vms@lp.se
Date: Fri, 24 May 1996 14:18:57 +1000
From: Paul Nankervis <CCPN@LURE.LATROBE.EDU.AU>
Reply-To: Free-VMS@lp.se
Subject: Re: Hi
To: Free-VMS@lp.se
Message-ID: <01I533Q7RVBM8Y4YAQ@LURE.LATROBE.EDU.AU>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; CHARSET=US-ASCII
Content-Transfer-Encoding: 7BIT

Unfortunately traffic has been a little light on this list for a while.

One reason is that there have been some problems with the list. I know
that some of my messages never managed to get through... :-(

Another is the nature of the task being contemplated - it is huge! You
don't have to think about it for long to realize how much work it
would take. However it is also true that the individual pieces are
relatively manageable. It might be possible to make it happen if a
sufficient number of volunteers could be coordinated properly. But
without a concrete starting point I doubt that too many volunteers
will come along.

Another difficulty may be that some parts are not very well documented.
For example I have never seen adequate documentation about the
internal structure of RMS indexed files? (On the other hand the file
system is very well documented?)

Now I have a particular question for Bill Davenport: Is any of the code
from VMS-on-Mach or Intrepid available to build on? I assume that the
bulk of the VMS-On-Mach specific code in C but that it relies on a
lot of existing VMS code for things like File-system, RMS, Utilities, etc

Does anyone else have any code which could be used as a first step?

Paul Nankervis
================================================================================
Archive-Date: Fri, 24 May 1996 15:34:01 +0200
Sender: owner-free-vms@lp.se
Message-ID: <9605241321.AA21694@us2rmc.zko.dec.com>
Date: Fri, 24 May 96 09:21:49 EDT
From: Bill Davenport - OpenVMS Security Group <davenport@tabsco.ENET.dec.com>
Reply-To: Free-VMS@lp.se
To: free-vms@lp.se
CC: davenport@tabsco.ENET.dec.com
Subject: FWD: Re: Hi

>>Now I have a particular question for Bill Davenport: Is any of the code
>>from VMS-on-Mach or Intrepid available to build on? I assume that the
>>bulk of the VMS-On-Mach specific code in C but that it relies on a
>>lot of existing VMS code for things like File-system, RMS, Utilities, etc

I'm not sure as to whether I can make any of the code we developed available 
due to legal reasons, but will ask.  The vast majority of the code base we had 
was code written in C to perform logic that we were looking at in listings 
written in Macro and Bliss.  We did some cheating as to the file system in that 
we actually had a very minimal RMS implementation layered on top of the 
existing BSD server present on our Mach system.  There was nothing about the 
model that precluded doing a real file system, but with only four engineers you 
can only do so much.

There's also a quite large working design document we wrote while we were 
looking at how to approach builing our system.  The document is currently 
company confidential - I'll see if we can relax the classification such that it 
can be made available.  The paper is usefull in that it goes into all of the 
pieces of VMS 5.4 and how they could be re-implemented in a multi-server 
microkernel system.

Bear in mind that our work was performed on a VAX 3100 workstation using a Mach 
3 kernel.  This was specifically chosen to help minimize the amount of effort 
that we would need to expend to produce a demonstration of the concepts.  We 
unfortunately never had any time to attempt porting the work to a different 
architecture (at the time we were thinking either a MIPS box or a 486).

Bill

================================================================================
Archive-Date: Fri, 24 May 1996 16:40:52 +0200
Sender: owner-free-vms@lp.se
Date: Sat, 25 May 1996 00:01:46 +1000
From: Paul Nankervis <CCPN@LURE.LATROBE.EDU.AU>
Reply-To: Free-VMS@lp.se
Subject: Re: FWD: Re: Hi
To: Free-VMS@lp.se
Message-ID: <01I53P7JFF6G8Y4YQ5@LURE.LATROBE.EDU.AU>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; CHARSET=US-ASCII
Content-Transfer-Encoding: 7BIT

Hi,

>          We did some cheating as to the file system in that
>we actually had a very minimal RMS implementation layered on top of the
>existing BSD server present on our Mach system.  There was nothing about the
>model that precluded doing a real file system, but with only four engineers you
>can only do so much.

I would be interested in learning more about your pseudo file system. How
were file attributes and the like stored? I imagine that it might be possible
to store a pretend file header as the first block of a native file, or
maybe to always open files with the attributes expected by the calling
program? Or assume all files are STREAM LF?

If you do manage to get permission to release documentation and/or code
it would certainly give this group some direction!

Thanks for your messages!

Paul Nankervis
================================================================================
Archive-Date: Fri, 24 May 1996 19:41:58 +0200
Sender: owner-free-vms@lp.se
Message-ID: <9605241729.AA08414@us2rmc.zko.dec.com>
Date: Fri, 24 May 96 13:29:19 EDT
From: Bill Davenport - OpenVMS Security Group <davenport@tabsco.ENET.dec.com>
Reply-To: Free-VMS@lp.se
To: free-vms@lp.se
CC: davenport@tabsco.ENET.dec.com
Subject: FWD: Re: FWD: Re: Hi

>>          We did some cheating as to the file system in that
>>we actually had a very minimal RMS implementation layered on top of the
>>existing BSD server present on our Mach system.  There was nothing about the
>>model that precluded doing a real file system, but with only four engineers 
>you
>>can only do so much.
>
>I would be interested in learning more about your pseudo file system. How
>were file attributes and the like stored? I imagine that it might be possible
>to store a pretend file header as the first block of a native file, or
>maybe to always open files with the attributes expected by the calling
>program? Or assume all files are STREAM LF?

This is one place where looking under-the-hood to see how it runs highlights 
one of the biggest hacks we had.  Through the RMS interface we provided support 
for only variable length ascii records - this made that logic very simple.

For other access to files (e.g. image activation and paging) we directly made 
access by page through the I/O server running on the system.

Just as an aside, Chris and I just spent the lunch today trying to get a saved 
copy of the system up and running.  We're having some problems with the file 
system (that fsck doesn't seem to be repairing correctly), but are hoping to 
try and get our code base up and running again in one of our offices.

Bill

================================================================================
Archive-Date: Tue, 28 May 1996 14:44:49 +0200
Sender: owner-free-vms@lp.se
Message-ID: <9605281234.AA07539@us2rmc.zko.dec.com>
Date: Tue, 28 May 96 08:34:32 EDT
From: Bill Davenport - OpenVMS Security Group <davenport@tabsco.ENET.dec.com>
Reply-To: Free-VMS@lp.se
To: freevms@tabsco.ENET.dec.com
CC: davenport@tabsco.ENET.dec.com
Subject: Chris Kaler will be at DECUS in Atlanta

Drop by and say Hi if you're attending.

Bill

From:	STAR::kaler "24-May-1996 1655"   24-MAY-1996 16:55:01.66
To:	'Bill Davenport' <davenport@star.zko.dec.com>
CC:	
Subj:	

Didn't have time to post this to the FreeVMS
so feel free to tell them that I will be at DECUS
if anybody wants to talk...

Chris


================== RFC 822 Headers ==================
Return-Path: kaler@star.zko.dec.com
Received: by galaxy.zko.dec.com (UCX V4.0-10B, OpenVMS V6.2 VAX);
	Fri, 24 May 1996 16:55:31 -0400
Received: from cgk2.zko.dec.com by us2rmc.zko.dec.com (5.65/rmc-22feb94)
	id AA21388; Fri, 24 May 96 16:46:17 -0400
Received: by cgk2.zko.dec.com with Microsoft Mail
	id <01BB4991.4764E220@cgk2.zko.dec.com>; Fri, 24 May 1996 16:51:39 
-0400
Message-Id: <01BB4991.4764E220@cgk2.zko.dec.com>
From: Christopher Kaler <kaler@star.zko.dec.com>
To: 'Bill Davenport' <davenport@star.zko.dec.com>
Date: Fri, 24 May 1996 16:51:38 -0400
Encoding: 6 TEXT

================================================================================
Archive-Date: Thu, 30 May 1996 07:48:31 +0200
Sender: owner-free-vms@lp.se
Date: Thu, 30 May 1996 15:48:05 +1000
From: Paul Nankervis <CCPN@LURE.LATROBE.EDU.AU>
Reply-To: Free-VMS@lp.se
Subject: Re: Chris Kaler will be at DECUS in Atlanta
To: Free-VMS@lp.se
Message-ID: <01I5BLTWV0HE8Y57BA@LURE.LATROBE.EDU.AU>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; CHARSET=US-ASCII
Content-Transfer-Encoding: 7BIT

> Drop by and say Hi if you're attending.

This would be nice - unfortunately Atlanta is a loooong way from home  :-(

Any news on the possible release of VMS-On-MACH material?

Paul Nankervis

================================================================================
Archive-Date: Thu, 30 May 1996 10:22:45 +0200
Sender: owner-free-vms@lp.se
Date: Thu, 30 May 1996 10:22:42 +0200
From: levitte@lp.se
Reply-To: Free-VMS@lp.se
To: Free-VMS@lp.se
Message-ID: <009A319A.66503C60.4@lp.se>
Subject: Re: Chris Kaler will be at DECUS in Atlanta

>From: Paul Nankervis <CCPN@LURE.LATROBE.EDU.AU>

>Any news on the possible release of VMS-On-MACH material?

I have them, and will put them on-line this evening.

-- 
R Levitte, Levitte Programming; Spannvägen 38, I; S-161 43  Bromma; SWEDEN
Tel: +46-8-26 52 47, (via nation.se) +46-8-21 24 90;  No fax right now
PGP key fingerprint = A6 96 C0 34 3A 96 AA 6C  B0 D5 9A DF D2 E9 9C 65
Check http://www.lp.se/~levitte for my public key.
================================================================================
Archive-Date: Thu, 30 May 1996 14:37:51 +0200
Sender: owner-free-vms@lp.se
Message-ID: <9605301230.AA08645@us2rmc.zko.dec.com>
Date: Thu, 30 May 96 08:30:35 EDT
From: Bill Davenport - OpenVMS Security Group <davenport@tabsco.ENET.dec.com>
Reply-To: Free-VMS@lp.se
To: freevms@tabsco.ENET.dec.com
CC: davenport@tabsco.ENET.dec.com
Subject: FWD: Re: Chris Kaler will be at DECUS in Atlanta

>Any news on the possible release of VMS-On-MACH material?

Chris and I have asked the question of management about releasing the source 
code that we developed - but haven't heard back yet...

The papers that we wrote on VMS-on-Mach and the Intrepid followon work have 
been released - hopefully we'll see them online shortly.

Bill

================================================================================
Archive-Date: Thu, 30 May 1996 18:09:33 +0200
Sender: owner-free-vms@lp.se
Date: Thu, 30 May 1996 18:09:31 +0200
From: levitte@lp.se
Reply-To: Free-VMS@lp.se
To: free-vms@lp.se
Message-ID: <009A31DB.9CC695A0.3@lp.se>
Subject: WWW pages updated

I've remade the pages at http://www.lp.se/free-vms/ a whole lot, among
others by trying to indentify things that need to be done NOW, and by
inserting links to the documents I have received so far.

Please browse those pages, give me comments (on this list or privatelly),
help to build a good task list to begin with...

-- 
R Levitte, Levitte Programming; Spannvägen 38, I; S-161 43  Bromma; SWEDEN
Tel: +46-8-26 52 47, (via nation.se) +46-8-21 24 90;  No fax right now
PGP key fingerprint = A6 96 C0 34 3A 96 AA 6C  B0 D5 9A DF D2 E9 9C 65
Check http://www.lp.se/~levitte for my public key.