Hi Riccardo,

> one of the things I would like to see fixed is "xdm working out of the
> box"... that is make xdm=yes and a normal user added with useradd
> should work. Maybe barebones, maybe not a tuned Xsession, but at least
> work. Turning XDM on and then not even root can login is surely scary!
> if I had no telnet I couldn't have rebooted the machine nicely. That
> might scare off someone .)

The docs clearly recommend creating a non-root user, logging on with
that user and then su-ing to root, so any condition that we create that
prevents this from occuring should be considered a show-stopper. I got
tripped on this not too long ago when I botched an update from RC3 to
RC4 and found myself locked out of my serial console, since I'd blown
away my users file and root is not allowed to log on by default. Since I
agree with the principle of disabling root logins from terminals and
screens, we better make sure we don't lock out users that can su to
root. 

> Yes there is a lot of progress. But I think the works need to be
> tested a lot. Although macallan18 stuff works on my 9500 for X, my
> 4400 seems"reluctant"...

What we want to do is get to a code freeze point, then we start doing
QA. There needs to be an established testing criteria, like Michael
Wolfson being able to build all 700 packages. Since RC5 was already
tagged, we really need to get bug fixes in as soon as possible and then
try to have something we can point to as significantly better than RC5
before RC6 is tagged.

> If someone has a good idea on how to dual-boot Os-X and NetBSD savring
> and replacing each time the OF variables for example (remembering all
> 1.05 oddities...= I could try to find some old hard drive and at least
> test the current stuff on that box. But it is every-day workstation so
> i can't destroy it.

The biggest obstacle I've found with OF 1.0.5 is where it locates OF
when real-base is -1 (0xffffffff). You can actually work around this and
leave the OF variables alone if you link the kernel to 0x500000, instead
of 0x100000. This is in the Makefile:

#TEXTADDR?=      0x100000
TEXTADDR?=      0x500000

You should be able to use XPostFacto to set the boot device and input
and output. I haven't used the OS X version, but I have used it to
dual-boot MacOS 9 and OS X, and played with it enough to feel it can
assist with dual-booting NetBSD and OS X.

> As my new 4400 test shows the X on exit proble doesn't always show up!
> also the behavior is often very different... 

Your earlier report was encouraging. Michael and I are coming to common
ground on both the X exit and the I/O issue, so hopefully I will be able
to assist him in some manner in getting them resolved (or at least I
won't get in his way anymore :-).

tim