On Mon, 13 Apr 1998, Ken Nakata wrote:

> On Sun, 12 Apr 1998 22:37:50 -0500 (CDT),
> "David A. Gatwood" <marsmail@globegate.utm.edu> wrote:
> > On Mon, 13 Apr 1998, Ken Nakata wrote:
> > It's in the Mach Kernel part, which doesn't actually use a miniroot.  The
> > Booter INIT runs the Mach Kernel, which gets data on the hardware and then
> > loads the default_pager and the vmlinux server from the root partition as
> > specified in the lilo.conf file (the lilo.conf file is read by the Booter
> > while still in MacOS).  Normally that's a MkLinux partition, but when
> > using the RedHat installer (pre-DR3), that root partition is an HFS
> > partition.  The vmlinux *server* (the installer itself) has a miniroot,
> > but by this time, the Mach Kernel has already loaded the vmlinux server
> > from an HFS partition.
> 
> Ah, OK, so MK has to have a minimal HFS support built in, just like it
> had to have a minimal UFS support back in CMU Mach days.
> 
> If that's the case, it sure looks like a case of license infringement.
> IIRC, the CMU license terms are very similar to those of UCB, of which
> covered code and its derivatives cannot be later GPL'ed (OSF MK is
> still under CMU license as well as OSF license, isn't it?)

Then let's look very closely at the terms on the files. :-)

The real test is to look in the kernel config stuff and see if the
hfsutils stuff is being used all the time. ??

Take care,

Bill