Subject: Re: dead computer
To: None <port-hpcmips@netbsd.org>
From: Bernd Sieker <bsieker@freenet.de>
List: port-hpcmips
Date: 04/19/2002 01:41:05
On 18.04.02, 13:27:26, M. Warner Losh wrote:
> In message: <20020418125503.D10508@mt.sri.com>
>             Chris Jones <chris@cjones.org> writes:
> : On Wed, Apr 17, 2002 at 10:06:46AM -0600, M. Warner Losh wrote:
> : 
> : > Remove all the batteries you can find, including the 2032 coin sized
> : > ones if your box has them, wait 10-20 minutes, reconnect the batteries
> : > and then plug it into the wall for 15 minutes.  Then, and only then,
> : > try to turn it on.
> : 
> : Yep, that worked; thanks!
> : 
> : Now I've reproduced the cause of the hang:  "shutdown -p" from NetBSD
> : (fairly recent -current).  So now I get to take the thing apart
> : again...
> : 
> : I'll file a pr about this.

I don't know if anybody else reported this, but this is also what
happens on my z50. It worked fine with 1.5.3_RC1, but now I run 1.5ZC
(I wanted the audio driver), and shutting down the machine locks it
hard, so that I also have to unplug the main battery. (I do not use
backup batteries, they are used up too fast anyway, and I don't need
to keep WinCE data in RAM.) On z50 I only have to unplug the battery
quite short (10 seconds is enough), and then I can turn the machine on
again.

Other than that I'm _really_ happy with the machine. If I don't need
floating point it's actually pretty fast. Handles things like ssh
etc. nicely.

Time to save money for a 512MB (or 1GB?) CF Card. And in the latest
c't (renowned German computer magazine) a test showed that there are
_huge_ differences in CF-card energy consumption; standby powers range
from 0(!) to 80 mW, depending on the brand. I think Kingston and
Toshiba were at 0, some others actually actually drew as much current
as IBM's MicroDrive.


> Warner
> 

-- 
Bernd Sieker

NetBSD: Rock solid!
		-- Lars-Johan Liman