First, I just wanted to apologize for any bounces anyone may have received
related to my email address this weekend.  A power failure Friday morning
apparently caused a failure of one of the Humanities building's 'smart'
hubs, which happened to be the one that globegate was on.  After trying
unsuccessfully to get the actual problem fixed, and after determining that
faculty office connections were working, we rolled globegate (running on
its UPS) down the hall to the Modern Foreign Languages main office.  There
will be a similar, more temporary outage when we roll it back, after the
CC gets the network problem fixed....

Second, the funny part....  I finally fixed the hardware problem that had
been plaguing our PB145 since the last repair or so....  It was crashing,
primarily when used on one's lap, and remarkably stable on tabletops, etc. 
Every so often, it would fail to boot with the chimes of death, but this
started getting more and more frequent, until yesterday, it never even got
the video display initialized to the dotted background before it died.

Well, a few days before, I had partially figured out the problem when I
adjusted the LCD angle while running NetBSD and got a vm fault.  After
repeating this several times and getting the same results, I opened the
unit.  Expecting some board to have dirty contacts, I opened the unit and
cleaned those... put it back together, same problem.  I also looked for
cold solder joints and couldn't see any.

I then tried booting with the upper half of the case separated... booted
perfectly, and I was unable to make it crash with the case apart, no
matter how much I flexed, twisted, hit, etc. the bottom half of the case. 
Same thing for the top half, obviously, since there doesn't seem to be a
lot in it...  video and sound and PRAM battery....

After careful inspection, I noted that the metal plate on the bottom of
the keyboard had warped over time (sagging between the three pairs of
screws), so I straightened it with a small amount of upward pressure in
the center of each half, and haven't gotten a vm fault since.  It seems
that not only was that warped plate causing sporadic missed keystrokes,
but it was also pressing down on, and apparently shorting against the RAM
chips attached on the small board atop the processor board....


Later,
David

-----BEGIN GEEK CODE BLOCK-----
Version: 3.1
GCS/CC/FA/H/L/MC/M/MU/PA/TW d-@ s:>- a-- C++++>$ UBLAS*++++>$
P+?>$ L+++>$ !E--- W+++>$ N++(+++)>+++$ !o? K-? !w--- !O
M++>$ !V-- PS+>$ !PE- Y+>$ PGP+>$ t+++>$ 5+>++++$ !X- !R tv+>$
b++>$ !DI !D- G++(+++)>$ e>++++ h--! r--- !y-
------END GEEK CODE BLOCK------