KC,
Personally, I'm not sure if any of these points are valid. And I have
(semi) logical reasons to back this statement up (all IMHO).
First of all, the current JOS stage is in developing the kernel and JVM.
We've just really broken ground on going into the actual architecture. So
at this point there isn't much of an OS to speak of.
Since there isn't much of an OS, I feel that the project isn't as
interested in performance and usability, as it is with creating a new
paradigm shift in how people think and design OSs. With Java as our base
language, we have an unparalleled basis for extensibility and object
design. Unfortunately, we are needing to extend the ideas set forth in the
Java specifications, such as multiple processes support, and all its
related baggage.
Once a foundation has been established, then everything else can fall into
place. Thus, I feel that with most of issues you bring forth (security,
protection, resource allocation, etc), 80% of that depends upon the design
framework we're developing now.
But that doesn't mean that we won't be interested in performance and
usability.
When one first hears that we're developing an OS built on Java, many people
assume that it will be too slow. Especially since many OS kernels are
built in highly optimized native code. But, since we need to host our Java
Operating System on a machine in order for the architecture to work, we'll
have native code underneath everything. One of the primary goals of JOS is
to minimize that underlying native code, to make it as cross-platform
generic as possible. So, in some cases we can move things from the Java
layer down to the native kernel layer. But other optimization techniques
exist, such as pre-compiling the base OS Java code, and optimizing the Java
code itself. All in all, I don't think that performance will be a problem
if enough time is spent.
For security, if we keep to the current Java model, but allow for
per-process selection of the security model, as well as system-wide
resource security including per-user level scrutiny, we should have a good
foundation for making UNIX level protection or better.
User convenience is based upon how well our tools are designed. With a gui
and XML, we have the ability to do an excellent job supporting both the
UNIX / text lovers with the Windows / gui users.
"You too can help tomorrow become a better today!"
-Matt
KC
<[EMAIL PROTECTED] To: [EMAIL PROTECTED]
state.edu> cc:
Sent by: Subject: [JOS-Kernel] HAY I'm writing
a report
kernel-admin@jos.
org
03/22/2000 08:50
PM
Please respond to
kernel
Hay,
I'm writing a report about JOS and I was wondering if any of you
could possibly tell me (or guess, I understand that the operating system
isn't complete yet but any information would be helpful) how you would
rate (on a scale of 1 to 10) JOS on three criteria? I'm trying to
figure out how Java operating systems will compare to operating systems
now on the basis of efficiency (i.e. overall speed including efficient
resource allocation, efficient interrupt handling, and good response
time of various kernel and systems routines), security and protection
(for both the user and the operating system), and user convenience and
response time. Thanks!
-KC
_______________________________________________
Kernel maillist - [EMAIL PROTECTED]
http://jos.org/mailman/listinfo/kernel
_______________________________________________
Kernel maillist - [EMAIL PROTECTED]
http://jos.org/mailman/listinfo/kernel
