From: ovek AT arcticnet DOT no (Ove Kaaven)
Newsgroups: comp.os.msdos.djgpp
Subject: Re: Any one writing OS in DJGPP?
Date: Sat, 25 Jan 1997 21:16:04 GMT
Organization: Vplan Programvare AS
Lines: 58
Message-ID: <5cg5t4$4sk$1@troll.powertech.no>
References: <32E14858 DOT 4CF6 AT concentric DOT net> <32E24EA0 DOT 7BEE AT blackmagic DOT tait DOT co DOT nz> <5bumbf$tov$1 AT troll DOT powertech DOT no> <32E488FA DOT 3744 AT blackmagic DOT tait DOT co DOT nz> <5c9j4o$61s$1 AT troll DOT powertech DOT no> <32EA933B DOT 1B89 AT concentric DOT net>
NNTP-Posting-Host: alwayscold.darkness.arcticnet.no
To: djgpp AT delorie DOT com
DJ-Gateway: from newsgroup comp.os.msdos.djgpp

Mark & Candice <mhewii AT concentric DOT net> wrote:

>Ove Kaaven wrote:
>> >A million monkeys...
>> 
>> I could probably make public my source to aid the 999 999 other
>> monkeys, if they would have any interest in it, especially if they
>> would like to work on it too.
>> 
>> Hmm, on the other hand... what would happen if all the monkeys put
>> together their pieces to make the ultimate djgpp-OS? ("Windows NT, go

>Posting your source would be great!

Posting 300 K (over 10 000 lines) of C code? On this newsgroup
(mailing list)? I thought people were already getting enough trash in
their mailbox.

>You might be suprized to find how quickly the other 999,999 could
> get together and make a 'djgpp OS'.

>There are many of us out here making OS's, if we all share a little
> this could work quite well.

Allright, since I'm such a "nice" person (uh?), you can find the full
sources so far on http://www.arcticnet.no/~ovek/files/os.html, but I'd
appreciate it if these sources get a "moral" treatment, so to speak...
in other words, at least have the courtesy to let me know if you steal
them and sell them to Microsoft, IBM, Intel, Netscape, or all of the
aforementioned, or something similarly unscrupulous.
(Though naturally, I hope they're going to be used for what they were
meant to be used for.)

Implemented (well, at least working) components of OS:
-Exceptions, IRQs, and software interrupts
-I/O port "manager", supports port I/O virtualization, with which a
virtual 8259 interrupt controller (for the BIOS VM) has been
implemented
-Primitive integrated disassembly output, helpful when exceptions
occur (but no stack trace)
-BIOS VM machine, for disk access, that can run concurrently with
other tasks (die Win95)
-"Universal" filesystem driver (incomplete) with FAT16 fs implemented
-Page mapping (not actually used for very much yet)
-Multitasking (must be a bug somewhere, can't make more than 2 tasks
work together)
-Message posting
-Console driver
-Keyboard driver
-Simple shell
-Preliminary libc

and maybe a few more goodies you might discover by taking a look.
(and a lot of goodies someone might implement someday.)
Yes, I know the source could have been made a little better...
but this isn't any of those "the check is in the mail", so...