On Sun, 2005-01-23 at 08:43 -0500, shoppa_classiccmp(a)trailing-edge.com
wrote:
I'm trying
to get xearth to compile on this 'ere Tek XD88, which only
has X11 release 3 - xearth uses a few odds and ends which only showed up
in R4.
I believe that they are fundamental to xearth. Before X11R4 it was
impossible to do xearth (that is, a circular window on the underlying
background).
Believe it or not, I just got it running, at least in the root
window :-) Hacked it to only ever work at 8bpp, and I had to ditch the
XAllocSizeHints / XSetWMNormalHints code - plus a few minor other
changes (strftime is ascftime on this machine etc.)
Next stop, XPlanet... (yeah, right ;)
XFractint might be a possibility though...
Did you ever settle the partitioning scheme used on
your XD88? While
others were offering suggestions for reading various styles of
partition maps, I was of the opinion that the partition map was part
of the kernel...
Nope, I haven't found anything yet, other than getting df's idea of ffs
partition sizes (and working out the number of blocks allocated for
swap). I haven't had a chance to see if those numbers match up with
anything in the first sector of the raw disk image yet.
The boot process is a little strange; in manual mode the system always
asks for a boot device (disk / tape / network) - selecting disk produces
some disk activity and then the system drops to dinex, which seems to be
some sort of monitor / system loader. 'boot' from here actually boots
the kernel (with option to boot other than the default one).
Dinex is present on the filesystem under /boot when running though, so
it appears that it's loaded from the disk itself rather than being
something in ROM. The option to boot various kernels implies that dinex
knows all about ffs filesystems and what partitions are present though,
prior to the desired kernel even being loaded.
cheers
Jules