OSDev.org

Blah blah blah blah.

Every project goes through the same cycle.

Author says "I wanna make an OS!!"
The author learns ASM...
He implements a boot loader...
He learns the details of p-mode...
he writes a somewhat functional P-mode kernel...
He implements hello world and perhaps
even a window manager and possibly a filesystem.
If he's REALLY 31337 like Richard A. Burgess he
publishes a book about his work and becomes
somewhat renownd untill his publisher cancels
the printing six months later... (Developing your own
32 bit operating system, 1995)

More likley If he has gotten even this far he
runs out of energy or his design hits limitations
and he quits. Regardless the project never really
goes anywhere. (Same with me but I never even got
to the point of getting a kernel out the door. =\)

Its the same old bullshit again and again.

DUDES; LISTEN TO ME!

This endless vishious cycle must end!

You (we) people are failing because we keep going
around doing the same bloody things again and again.
Sure we're getting a great education out of each attempt
but it isn't getting anyone anywhere. If that isn't enough
for you then you're like me and you want to push your project
to the next level but you won't get there by
re-inventing the wheel for the hundredth time.

This isn't another call to "UNITE TO MY PROJECT, ITS THE BEST!!!"
No, its something better. Its a project to help *YOU* (and me),
all of us actually, to break free of this damn cycle and emerge
once and for all on the next level!

The product I want you to help to develop will be called "bridge" (common noun).

On the left will be "MetaOS" It will be a free and standardized microkernel that
will run on DOS and essentially use DOS/config.sys to do all the little things that
waste our time. You may not think its manly to run on DOS, you should boot the system yourself!
No, that's not the point. Developing your own x86 kernel is a total waste of time!
The point is to get BEYOND DOS, not to re-implement it again ad nausium. Let DOS do what it
was designed to do. Let it WORK for you. *USE* it. EXPLOIT it. It can take you anywhere you want to
go from the battle for azeroth to the throne of Orion. (I over indulge in those trips. =\ )

Meta OS will be a development kit including a state of the art DPMI kernel including full SMP support
with controll program interface. It will provide drivers, libraries, compilers, linkers, manuals...
Everything.

On the other end of the bridge will be WHATEVER YOU MAKE!!!
I want to implement what I call Sphere OS and then its successor
which is in a much earlier stage of development. =\

The *Real* problems in OS design are how to manage resources, objects, and functions
on a high level. *ABSTRACTION*. This can mean whatever you want it to. But this is what
you need to focus on. The linux kernel is nothing without its *abstractions*. The principal
abstraction linux provides are the rather crude notions of files and processes but from there
you are on the road... As I hinted, the concepts of files and processes are obsolescant
I would like to replace them, but that's up to me, and you.

What I am offering you is an on-ramp to the fast lane to the big time level of OS development.

I know what I am talking about. I have been working on my projects since '94. Though I didn't get
to the point where I was a reasonably good programmer much less an OS specialist till '97...

DOS is not just my inspriation its also the ANSWER!

DOS is down not because it sucks but because of POLITICS.
DPMI was a threat to Windoze so Microsoft killed it.
No matter what host you use, you will be using DPMI 0.90 on DOS
or if you are running OS/2 you will get a version of 0.95...
I am going to make DPMI 1.0 with full SMP support a reality!

Then I am going to make a 2.0 version that provides a guest-OS interface
too.

DPMI can be a state-of-the art microkernel.

REMEMBER: A microkernel is not an OS!!!

Its the beginning of an OS.

Quake, ZSNES, and a bunch of others I don't know
about use DPMI to take DOS to the next level.

You can use it/VESA 2, to do increbile things!

DOS will even load your code for you. =)

I hope you will see that I have a plan here and
that it can benefit you as much as me, to the
completion of MetaOS at least. Then we will be
able to work on the Real Problems...

>On 2001-03-12 21:26:12, AlonzoTG mailto:[email protected] wrote:
[ .... basically calling to extend the project ....]

That was a very nice speech to "Extend DOS"
However, there is no thing that many people did not, however,
look in as a "goal" that their project is to stick
with Ken Thompson's idea of a multi-user and environmental resources
sharing OS that is "reusable," in the foundation of the
Free Software Foundation, where people [a lot of people] "learn"
how things operate just like they look into the nutrition chart
to see what are they eating.
To what you've said that many of the things such as DPMI and other
great ultilities, but no one ever said officially they are
published under the lincese of GPL (meaning it's not going
to be a free-for-all game unless you code your own DPMI server).

Windows 32-bits (9x, NT, and ME) are the extention of DOS,
for there backward support on the things in DOS,
such as games in real mode only require a reboot to DOS.

There are other DOS vendors yet not many "normal users" would be
spending time trying to figure out how to install a sytem
that is not going to be compatible (sometimes) with their Voodoo 5
and lots of device drivers.

To extend the projects to the next level, I personally think,
information [only] should be freely disturbuted (under
complance with "copyright respects") like the textbooks
in school campuses. And each individual learns most out
of their "own" project --> it's like a personal oddesy.

While others that are constantly needing a good system
for gaming and so forth will just find the existing systems
in the market good-for-this-purpose. I personally
play SNES games in Zsnes, and I believe they only need
to publish their source code, it could be modified to
any version available. And they've also added versions
that supports 32-bit environments.

What I try to say is that many people is still not yet
ready for a guest-OS that loads a self-made kernel for
they are still too advanced somehow in a place and time
that not many of the hackers even have a "full picture"
of the entire route in the oddessy of OS programming.

Dispite some weird thoughts in me, ... something like
a guest-OS could be helpful. Can I get a copy of your
guest-OS anywhere?

>While others that are constantly needing a good system
>for gaming and so forth will just find the existing systems
>in the market good-for-this-purpose. I personally
>play SNES games in Zsnes, and I believe they only need
>to publish their source code, it could be modified to
>any version available. And they've also added versions
>that supports 32-bit environments.

Bah!
ZSNES under DOS *IS* 32 bit.
QUAKE under DOS *IS* 32 bit.

Heck, Everything using dos4gw is 32 bit under DOS,
though I can't say wheather or not it uses the DPMI host
or a more primitive service.

>Dispite some weird thoughts in me, ... something like
>a guest-OS could be helpful. Can I get a copy of your
>guest-OS anywhere?

No, I just proposed it. =\
I just ordered the official DPMI 1.0 spec and today
I'm going to look into liberating Qdpmi from Symantec,
(or just throwing a sourcerer at it.)

>On 2001-03-13 11:43:42, Anonymous wrote:
>>While others that are constantly needing a good system
>>for gaming and so forth will just find the existing systems
>>in the market good-for-this-purpose. I personally
>>play SNES games in Zsnes, and I believe they only need
>>to publish their source code, it could be modified to
>>any version available. And they've also added versions
>>that supports 32-bit environments.
>
>Bah!
>ZSNES under DOS *IS* 32 bit.
>QUAKE under DOS *IS* 32 bit.
>
>Heck, Everything using dos4gw is 32 bit under DOS,
>though I can't say wheather or not it uses the DPMI host
>or a more primitive service.
>
>>Dispite some weird thoughts in me, ... something like
>>a guest-OS could be helpful. Can I get a copy of your
>>guest-OS anywhere?
>
>No, I just proposed it. =\
>I just ordered the official DPMI 1.0 spec and today
>I'm going to look into liberating Qdpmi from Symantec,
>(or just throwing a sourcerer at it.)

Check out RDOS :-
http://www.rdos.net/rdos/index.htm

Yours Sincerely,
Aaron Gray

FreeDOS (www.freedos.org) isn't that bad either!
>On 2001-03-18 21:41:15, Aaron Gray wrote:
>>On 2001-03-13 11:43:42, Anonymous wrote:
>>>While others that are constantly needing a good system
>>>for gaming and so forth will just find the existing systems
>>>in the market good-for-this-purpose. I personally
>>>play SNES games in Zsnes, and I believe they only need
>>>to publish their source code, it could be modified to
>>>any version available. And they've also added versions
>>>that supports 32-bit environments.
>>
>>Bah!
>>ZSNES under DOS *IS* 32 bit.
>>QUAKE under DOS *IS* 32 bit.
>>
>>Heck, Everything using dos4gw is 32 bit under DOS,
>>though I can't say wheather or not it uses the DPMI host
>>or a more primitive service.
>>
>>>Dispite some weird thoughts in me, ... something like
>>>a guest-OS could be helpful. Can I get a copy of your
>>>guest-OS anywhere?
>>
>>No, I just proposed it. =\
>>I just ordered the official DPMI 1.0 spec and today
>>I'm going to look into liberating Qdpmi from Symantec,
>>(or just throwing a sourcerer at it.)
>
>Check out RDOS :-
> http://www.rdos.net/rdos/index.htm
>
>Yours Sincerely,
> Aaron Gray
>

Zsnes's source code was released awhile ago hehe