I don't know whether or not this is the right place for this, but I had some problems installing hamm on my system. As I do not yet want to drop my bo system, I decided to try it out on a looped filesystem. I have recompiled a 2.0.32 kernel with looproot support (patch from linuxhq.com). Installing bo to a looped filesystem is easy: mount it in tty2 on /target and press (twice) ESC in dinstall. Now pressing ESC twice in dinstall doesn't make dinstall re-examine the state of the installation. Executing a shell and then returning to dinstall seems to have the same effect, but then dinstall still doesn't see the looped fs on /target (bo installation had no problem with that). I then run losetup /dev/loop1 /hda6/loopfs.lin/hammloop.img, hoping dinstall might see the /dev/loop1 alongside with /dev/hda1,2,5,6 and 7 devices... It didn't. At last I got it installed, but not quite easy: first I tried to remove /dev/hda7 and ln /dev/loop1 /dev/hda7, but then dinstall tried to mount /dev/hda7 as a MSDOS filesystem (/dev/hda7 is an ext.DOS partition allright, but now /dev/hda7 was hardlinked to /dev/loop1). It seems that dinstall looks at the partition info of the hard disk to determine the type of the partition. As I do have an ext2 filesystem /dev/hda5, I chose that one to mount as a previously initialized partition, the executed a shell, unmounted /dev/hda5, removed /dev/hda5 and hardlinked it to /dev/loop1 and mounted it again. Exiting the shell resulted (finally) in being able to copy and install all the rest of the stuff.
Maybe someone can do something about it, so that installing to anything mounted on /target should be possible, like in bo. Greetings, Maarten Bezemer. -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
