[Coco] NitrOS9 build errors

Tormod Volden lists.tormod at gmail.com
Fri Jan 17 18:23:59 EST 2014


On Wed, Jan 15, 2014 at 9:33 AM, Tormod Volden wrote:
> On Wed, Jan 15, 2014 at 3:57 AM, Bob Devries wrote:
>> @Tormod:
>> It seems that the errors you found back in October last year are still there. These relate to the Dragon disks. Please note that this is *NOT* a slight of your work; just a comment about the current status.
>
> Yes, there are still a few errors related to the Dragon disk images.
>
>> @all:
>> Is anyone working on the Dragon parts of the NitrOS9 source tree?
>
> I don't think so.
>
>> I think these issues need to be resolved before we go to a new revision number.

After revisiting these issues, I agree with you, they should be fixed
before next release, with or without the help of any Dragon 64 users.
It is not so difficult to get rid of them, and they are the only ones
keeping us from an error-free nitros9 build.

I have now removed the references to the acia51 serial port driver,
there is no module nor source code for it. Those who need it can
probably copy it from an OS-9 disk.

The last remaining issue is the one overfilled disk in the 180K disk
set. While searching the list I found this conversation between you
and Gene from July, that I had missed or forgotten:
<http://five.pairlist.net/pipermail/coco/2013-July/070187.html> There
you suggest to format the 180K disk using sa=4 in order to make enough
space on it. However this doesn't work for me, I get an error 1 in the
following copying of the boot track. I'd suggest to instead shuffle
around some contents between disk1 and disk2 because there is more
free space on #2 than what is lacking on #1. Simply moving the DEFS
folder to #2 solves it, and leaves enough space to move all commands
back to #1. I don't have any experience actually using NitrOS-9, so I
don't know if this split would cause any inconvenience. Any 180K disk
users can anyway move things from one disk to another as they please.

However, I cannot get these disk images to work reliably in an
emulator. OS-9 can be booted by running:
 xroar -machine dragon64 -machine-cart dragondos -run OS9-SYSTEM.VDK
and the NitrOS-9 disk images can be read from within OS-9, so I am
sure the emulator reads the disk images correctly. But when booting
from the NitrOS-9 disk images, the boot hangs (program counter running
astray) after registering rel, krn and boot modules. Well, except for
a few times when I was using gdb and interrupting the boot, which for
some reason made the boot continue to the NitrOS-9 command prompt. The
gdb debugging should in principle not change anything. Maybe it all
works fine on real hardware?

Regards,
Tormod



More information about the Coco mailing list