[Info-vax] OS design, was: Re: [OT] Wirth style languages, was: Re: Obscure Ada compiler vendors?

[Keith Cayemberg]

On Tuesday, April 9, 2013 7:35:54 PM UTC+2, Simon Clubley wrote:
> On 2013-04-09, David Froble  wrote:
> 
> >
> 
> > Of more concern to me is VMS (as far as I know) is more of a monolith, 
> 
> > with lots of complexity and inter-relationships.  Is this reality?
> 
> Yes. There are inter-relationships in VMS which simply do not exist in,
> 
> say, Linux, even though both are monolithic kernel designs.

IMO that is one of the many myths that have been perpetuated about OpenVMS architecture, also by principals that should know better (Wikipedia also has it completely wrong). OpenVMS has a "Layered OS Design" as described by Andrew Tanenbaum in his "Operating Systems Design and Implementation"(*). Please see sections 1.5 Operating System Structure, 1.5.1 Monolithic Systems and 1.5.2 Layered Systems. This is a definitive organizational improvement over the monolithic Linix, DOS and CP/M kernel designs. Even if the layers may be tightly integrated and coupled in OpenVMS, it doesn't change the fact that the layers exist and do provide for an organizational division of functionality useful in building, changing, enhancing and protecting kernel subsystems and structures. Examples in which the layered and modular subsystem design of OpenVMS was used to introduce extensive functional enhancements include the later introduction of SMP, ACLs, Clusters/Distributed Lock Manager, ODS-5, and Kernel Threads not conceived in the original OpenVMS OS Architecture.

For more independent testimony of the layered nature of OpenVMS OS Design please see the following resources...

Operating Systems Concepts, Second Edition by David Donald Miller, Butterworth-Heinemann, 1997, see pages 65 - 67.

VAX/VMS Internals and Data Structures, Version 5.2 by Ruth E. Goldberg and Lawrence J. Kenah, Digital Equipment Corporation, 1991, see pages 9 - 14.

OpenVMS Software Overview, Digital Equipment Corporation, 1995, see section 3.1.1 "OpenVMS Kernel", page 3-2.
http://h71000.www7.hp.com/doc/73final/documentation/pdf/OVMS_SW_OVERVIEW.pdf

HP OpenVMS Programming Concepts Manual, Hewlett-Packard Development Company, 2005, see section 20.2 Preserving System Integrity.
http://h71000.www7.hp.com/doc/82final/5841/5841pro_057.html#ssr_sys_integrity

Operating System Structure by David Duggan (Sandia), 2011, see slides 4 - 13

See also...
https://sites.google.com/site/assignmentssolved/mca/semester2/mc0070/2

(*)
Operating Systems Design and Implementation, Third Edition
By Andrew S. Tanenbaum - Vrije Universiteit Amsterdam, The Netherlands, Albert S. Woodhull
- Amherst, Massachusetts
Publisher: Prentice Hall
Pub Date: January 04, 2006
Print ISBN-10: 0-13-142938-8
http://passthrough.fw-notify.net/download/161326/http://it.tdt.edu.vn/~tttin/giangday/HDH/Operating%20Systems%20Design%20&%20Implementation%203rd%20Edition%281%29.pdf

There are certainly more sources, but this should already suffice for academic argumentation and peer review. Those wanting to argue that Layered Kernel is not a class of kernel, should first explain what class of kernel MULTICS exhibits, and also explain the multitude academic sources that do indeed include Layered Kernel as a class. It appears that in many cases, the dropping of "Layered Kernel" from the OS Topologies in some sources may have been due to an unconsciously ingrained, self-serving, and biased evaluation of technologies. At least for a time, monolithic OS design was actively used as a derogatory term (like Legacy) which has been a marketing mechanism to deride any technological competition someone wanted to label as old and out-of-date.

Cheers!

Keith Cayemberg