TEAM-ADA Archives

Team Ada: Ada Programming Language Advocacy


Options: Use Classic View

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

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

Print Reply
"Robert C. Leif, Ph.D." <[log in to unmask]>
Thu, 18 Nov 1999 22:40:56 -0800
text/plain (113 lines)
From: Bob Leif
To: Nick Roberts et al.
I suggest that you look at Ada's (the language) grandfather's latest
creation: Niklaus Wirth and Jurg Gutknecht, "Project Oberon, The Design of
an Operating System and Compiler," Addison-Wesley-ACM Press, 1992 ISBN
0-201-54428-8. The overall structure appears to be sound and this was done
by a small group. Although, I have not actually used the Oberon operating
system, I suspect it is probably more reliable than Windows 98 etc.

I believe that a Linux with an Ada core, and an XML based GUI would be a
commercially viable product. PLEASE, if at all possible, absolutely minimize
the use of assembler. I would gladly trade-off performance for reliability
and compatibility with Ada real-time constructs. If Lockheed will ever
release Michael Card's object-relational database (FIRM) which is written in
Ada, you should be able to build a very advanced file system.

One of my goals for the SIGAda '99 Commercializing Ada Workshop was to
demonstrate the feasibility of creating a very advanced, Ada based operating
system. The use of XML-HTML is an effective way to break the Microsoft
monopoly. Printers and graphics boards could be programmed to directly
execute XML.

Amusingly, an Ada based operating system could be of significant use to
Ada's former patron, the US Defense Department. Besides, the often repeated
software engineering benefits of the separation of specification and body,
it should be possible for a customer to limit the encryption algorithms to
one or a few bodies. This way, even the operating system developers do not
have to know the security protection mechanisms.

-----Original Message-----
From: Team Ada: Ada Advocacy Issues (83 & 95)
[mailto:[log in to unmask]]On Behalf Of Nick Roberts
Sent: Thursday, November 18, 1999 9:49 AM
To: [log in to unmask]
Subject: "OS in Ada" project promoting Ada

Some of you may be aware that some volunteers and I have initiated a
project, called "AdaOS" at the moment, to build a complete operating
system written (mainly) in Ada. The URL is at the bottom of this

Please find below an extract from my latest post to the AdaOS project
'chat' page.


My original (highly ambitious!) idea for the OS was to create a secure,
distributed, object-oriented, general-purpose operating system, which
was well-designed, efficiently implemented, and written (mostly) in Ada.

The plan that I have in my mind as a first objective of the AdaOS
project is to implement just enough of the system software to be able to
support the execution of elf2/Linux (libc6+XWS) binaries, and to be able
to distribute the OS in a form that is reasonably easy to install on
ordinary PCs.

Being able to run programs such as StarOffice, Netscape, Word Perfect,
the GIMP, and the plethora of GNU/Linux/XWS software now available,
would, I think, enable us to attract an entirely larger scale of
audience to the project. This would, in turn, hopefully, bring a number
of new programmers into the project, and also really help to promote
Ada. I quite like these objectives.


Possibly this is a project that has the potential to help, in a small
way, to promote Ada. It started as a little pet hobby project of mine,
but it seems to have attracted quite a bit of attention (to my genuine
surprise); enough to make me feel that I can't treat it as just a
personal indulgence any more.

To this end, I hope to be able to adapt GNAT and GCC as the quickest way
to provide the project with an Ada compiler which can target the new
AdaOS microkernel (on the IA32). I am resigned to the task of delving
into the inner recesses of GNAT and GCC to do this, but I would be
heartily grateful for the assistance of anyone in a position to help. I
hope to get the kernel and basic support software up and running as
quickly as I can. (I think the kernel will have to be written in

The project currently has no funding, no backing, and no principal
target application. I am realistic, but can anyone suggest a way in
which the project could be adapted, extended, moved, conglomerated, or
affiliated so as to attract any kind of funding or backing, however

I certainly hope to eventually be able to write some fairly serious
application software in Ada for this OS, once it gets up and running, as
I believe it is a visible application that is likely to really attract
attention. But what application? Is there an existing or planned open
source or GPL Ada application project that I haven't heard of which
might fit the bill?

I personally have very little in the way of funds, and, of course,
limited time - I really wish I had more of either - to devote to this
project. It is strange, is it not, how it really does seem to be left to
impecunious volunteers such as myself to promote a language which, on
the face of it, ought (still) to be considered of major national
importance, by the US government, and by other national governments too?

I would appreciate advice about how best to make this project serve the
interests of the Ada community, and to promote the language, and, I
suppose, the philosophy of good software engineering that lies behind
the language.

Nick Roberts
Computer Consultant (UK)