make floppy images swappable
Brought to you by:
bartoldeman
Menu
▾
▴
I cannot get it to work here. I have $_vbootfloppy = "/home/some/symlink.img" and $_floppy_a = "threeinch:/home/some/symlink.img" in my .dosemurc, but when I let the symlink point to another image, DOSEMU still sees the first disk. (Same behavior as before.)
My patch doesn't touch $_vbootfloppy.
Is the problem specific to $_vbootfloppy?
Actually, I'm not really sure what's the difference between $_vbootfloppy anf $_floppy_a, but I'm trying to use it with the MS DOS 6.22 installation disks.
Comment out $_vbootfloppy and see.
(boot from hd then)
Then I get this when I try to access the floppy image:
ERROR: can't get floppy parameter of /home/julius/dosemu/dosinst/msdos622/dosemu-disk.img (Ongepaste ioctl() voor apparaat)
Not a good day to die!!!!!
Please see if this patch takes care of
that message.
If not, please attach an image.
Yep, now it works :)
Should I also test it with $_vbootfloppy already?
You know why there isn't just a boot from "floppy, hd, cdrom" instead of this strange $_vbootfloppy btw?
History. This is from 1994, when people wanted to be able to boot from a virtual floppy file but then have the real floppy drive available as A: (using bootoff.com in autoexec.bat).
(i'll test the patch soon, sorry for the delay!)
Yes, please test with $_vbootfloppy too,
but I think we should split the work. The
patch is already large enough, $_vbootfloppy
will go after.
Bart: Yes, that scenario may have existed, but
as well you can have A: as a boot image and
B: as a real drive, and then the problem is
solved: not even bootoff is then needed, or
it can just substitute A: with B:
Z_God: why are you using $_vbootfloppy at all?
Just use $_floppy_a always. Or is it unbootable??
My floppy is bootable, but I understood that DOSEMU would boot from C: if I wouldn't specify it.
I guess this option was used when it was still common to have two physical drives and you wanted to use both together with a separate boot image. I guess this usecase is not that common anymore...
So do you claim that $_floppy_a is never bootable?
Any reasons against declaring this a bug?
Or, maybe, there can be a special syntax,
for example, $_vbootfloppy="a:" may mean
boot from $_floppy_a...
It seems that there are indeed some issues, although I haven't tested much..
I just realized I overlooked the "-A" argument which makes DOSEMU boot from a floppy drive. With my current build (with this patch), I get a non-systemdisk error then though when I try to boot an MS-DOS installation disk.
Btw. a lot of disk images can be found here now:
http://archive.org/details/IBM_PC_Compatibles_TOSEC_2012_04_23
I just tested GIT without any patches, There is indeed a problem. Images that work with $_vbootfloppy are not working when set as $_floppy_a and together with the "-A" argument.
Should I report this as a separate bug or can it be solved together with this?
I tested on a system which has a real floppy drive now. On that system (with current GIT), I can properly boot from the real floppy drive when it's only set as $_floppy_a.
When I make an image of that same disk, it is not bootable when specified as $_floppy_a, but is bootable through $_vbootfloppy.
So it seems there is still a problem with booting from image files with this patch.
This patch is only making the image swappable.
Nothing more. If, however, with $vbootfloppy
and _without bootoff.com you get a non-swappable
image, then this is worth taking a look into.
Yes with $vbootfloppy I indeed get a non-swappable image _without running bootoff.com.
(After running bootoff.com I cannot read anything from drive A at all anymore.)
After running bootoff.com you are suppose
to read $_floppy_a
Yes, I have it set to the same image as $_vbootfloppy, so I would expect to see no changes. My configuration looks like this:
$_vbootfloppy = "/home/julius/dosemu/dosinst/msdos622/dosemu-disk.img"
$_floppy_a = "threeinch:/home/julius/dosemu/dosinst/msdos622/dosemu-disk.img"
I tried it with another image now, same result. It seems bootoff.com is completely broken (with images) with current GIT.
With my patch or without?
With your patch. Should I also test everything without?
Of course!
If the patch is broken that way then
it cannot be applied. But if it is an
old problem then I will apply the patch
regardless.
I just tested everything without your patch:
- bootoff with the same image
- bootoff with another image
- "dosemu -A" with an image file
It seems all the problems are indeed old and your patch didn't introduce any new problems. All of these are broken without it as well.
Should I report all these problems as separate bugs?
OK, thanks for testing.
As you may see, this is not a bug report,
but a patch ticket instead.
Since the patch is applied, the ticket is
closed.