Re: XFree86 4.0.1 works!
On Thu, Oct 26, 2000 at 06:55:43PM +0200, Marcus Brinkmann wrote:
> On Tue, Oct 24, 2000 at 03:49:40AM -0700, Steve Bowman wrote:
> > Diamond Viper V770 using nv driver.
> > Umm. This driver has at least one really annoying bug. Since this is
> > X-related instead of hurd-related, I'll probably work it last unless it
> > just bugs me too much. Before starting X, I have white on black, after
> > X exits the first time, it's a very faint white on black, after the 2nd
> > and subsequent exits, it's black on black so I have to type blind or work
> > from a remote login. Something's not getting saved or restored right.
> I see now something similar wioth the vga driver in current CVS X.
> There are always funny terminal colors when I exit (pink on black, bright
> green on black, dark blue on black etc).
> > On debian-x, there was some discussion about using nvidia's non-free
> > driver. I didn't really follow the discussion very closely since I
> > hadn't moved up to X4 on linux, but I think I'll go check the archives
> > to see if this is a widely-known bug with the xfree86 nv driver or what.
> Maybe it's a broader problem?
I've partially localized where this could be - can't remember off
the top of my head. Time and motivation permitting, I may look at
it some more (need also to check debian-x archives first).
Before that, I'll probably go back to trying to debug the "mad
translators" as Mr. Bihlmeyer likes to call them. I think I can
reproduce this on the mouse translator. Otherwise, I told you
killing pflocal is nasty but didn't mention why. kbd and term both
blew up at once. Killing pflocal while X is running may be an even
more "reliable" way of reproducing this at will (if anyone else
wants to get into gdb :-)). The down side is it happens very fast
so you'll have to be quick. Even if I don't catch it for awhile,
I'll learn something about hurd debugging :-).
> > Anyway, as long as I stay in X all is well (sort of :-)).
> Hehe. I was actually not yet able to start it properly. The server starts
> up, but when using startx, twm is also runs, but the menus are black on
> black. But I have a completely messed up install, because I have Debian X
> 3.3.6 packages, X 401 manual install and X 4.0.1 phase2v21 Debian packages
> all in the same root fs tree, installed over each other. I'll have to clean
> up and reinstall from scratch to test it properly, but only after Branden
> released new deb source with Hurd support in it.
I have a mess too, but it's OK unless I want to go back to 3.3.6
since 4.0.1 is a full install I can't hit any 3.3.6 pieces.
Reinstalling 3.3.6 leaves the risk of hitting 4.0.1 pieces that
3.3.6 debs aren't installed on top of. Fortunately, I don't have
two flavors of 4.x stepping on each other in my mess :-).
> I will probably upload the 4.0.1 debs I produced saturday, although because
> of the above bugs I am not sure how useful this is.
Will test if I get a chance. Hmmm. I've got one test patch in
hurd to back out and one in X for opening the mouse with O_RDONLY.
Do you have your planned fix in hurd for that (mouse RW support)?
Oh, another thing on the kbd business - has anyone looked at the
3.3.6 kbd open code? That would be a useful thing to do and I have't
done it. May be faster than patching in your kbd test program....
Steve Bowman <firstname.lastname@example.org> (preferred)
Buckeye, AZ <email@example.com> <firstname.lastname@example.org>
Powered by Debian GNU/Linux and GNU/Hurd <http://www.debian.org>