Subject: Re: Misc Statements/Problems/Suggestions [long!]
To: None <port-arm32@NetBSD.ORG>
From: Markus Baeurle <emw4maba@gp.fht-esslingen.de>
List: port-arm32
Date: 11/17/1996 16:57:42
Hello Stefan!

Although this comes a bit late, I still want to make some comments.

In message <199611061216.NAA05153@cip56.cscip.uni-sb.de>
          Stefan Schlesinger <stscl@cscip.uni-sb.de> wrote:

> First of all, I have to thank everyone contributing to the development
> of RiscBSD over the last months (years now, in fact) for the very good
> work they've done. I've been reinstalling RiscBSD some five or six times

!!!

> data transfer from non-german domains has gotten _awfully_ slow recently
> (~2 Byte/sec at daytime), I had to spend several nights to get those files
> and have them installed, though not without problems again.

I can not at all confirm this. I've had very good rates recently where most
probably our ISDN link was the limiting factor (more than 6kB/s from
ftp.ph.kcl.ac.uk through a 64kBit/s line).

> - X still doesn't work for me, after setting the relevant paths and
>   changeing the access rights for /dev/vidc??? (correct filename forgotten, 
>   sorry) and /dev/quadmouse, using startx just gives the NetBSD intro-screen,

Well, Xarm-27 as it comes out of the set doesn't have the setuid-Bit set and
belongs to bin:bin instead of root:root, that's the real cause for the problem.

>   a short glimpse of the standard background and then a shut down to the
>   command line with the message
>     Returning from InitInput()
>   again. This has been checked with fwvm, twm and without any Window manager
>   at all, always the same result. In fact, fvwm additionally crashed with a

Hmm, I didn't see that on the two machines where I've helped to install RiscBSD.

>   at all, always the same result. In fact, fvwm additionally crashed with a
>   'Data abort: Permission error (page)' and produced a core dump.
>   Remember, I got all the sets present on ftp.ph.kcl.ac.uk at October 30th.

That's true, putting fvwm (and often twm has the same problem) into the xinitrc
almost always leads to a core dump, while starting it from an xterm usually
works.
Anybody else who has problems with fvwm on his StrongARM machine?

> Apart from X RiscBSD seems to work stable now, as far as I could see.
> This has been important, since one of our new customers is an Internet
> Provider who has been very impressed by the speed of the StrongARM RiscPC and
> is planing to use one under RiscBSD as their main server, as soon as the

Hohum, I don't think RiscBSD is stable enough for such a job yet. It should not
crash, but RiscBSD still does that now and then. It most probably will in a few
months, but still lacks a bit of stability now IMHO.

> Therefore, Mark, if you're willing to have another mirror in Germany
> (Stuttgart seems to be updated not very often anyway), it'd be somewhere
> on ftp.rz.uni-sb.de or ftp.cs.uni-sb.de. I'd feel honoured to do the
> administrative work for that.

That's good to have, but how often is it updated? When I tried to fetch the new
sets Mark had announced the day before, they were not present at uni-sb yet.

Cheers, Markus