[NTLK] Lady Newt problem guessing lottery: And the winner is...

[NewtonTalk]

...nobody. Sorry about that. Not that I expected any different outcome.

Just in case you've forgotten what Lady Newt's problem was: She refused to
be turned off. Like all self-confident women do. Instead, she immediately
turned on again after being turned off. The only way to fix this was by
brainwiping her. Which, when women are concerned, I always find a tad on the
embarrassing side.

This is what list members guessed:

1) A problem with the Interconnect Port.

2) A short across the reset pins.

3) The sleep preference being set to "Never".

4) Espy Enabler being installed.

5) Characteristics inherent in the circuit.

Although I kind of liked number five, the amateur lawyer in me decided to
disqualify this entry from taking part. Easily explaining pretty much every
Newton problem not visible to the naked eye, it would have unfairly placed
the more specific guessers at a disadvantage.

For a long time I didn't have a clue how to provoke Lady Newt's little
indisposition. I had seen it twice, and both times I was able to get rid of
it by brainwiping her. I never managed to figure out what to do to get it
back AFTER the brainwipe.

Until, that is, when I decided to check the temperature resistor's contacts
in her battery compartment because I had the impression that she was having
a hard time distinguishing rechargeable battery packs from battery holders.
As expected, the contacts were a tad on the dirty side, so I cleaned them,
bent them down a bit and put Lady Newt back together. Imagine my surprise
when the problem was back when I switched her on again.

Having the problem back gave me the opportunity to test whether it only
occurred when switching Lady Newt off by means of the switch, or whether it
ALWAYS occurred when she was told to power down. So I set the sleep
preference to one minute, made sure she wouldn't stay awake when the AC
adapter was connected, and waited. She turned off on her own after about 90
seconds and immediately turned on again. This was downright spooky...

Generally, my solution-finding processes aren't based on intelligence. Let
alone on competence or smartness. Why should they! Most times, stubbornness
and luck are all you need. So I brainwiped her again because, other than
giving up, this was the only thing that came to my mind. After that I
started thinking.

"Well", thought I, "what has happened during the last couple of minutes that
has made this darn problem reappear?"

Obviously, I had cleaned the temperature resistor contacts in her battery
compartment. But I strongly doubted that merely cleaning these contacts had
brought her problem back. Especially since I had done the last couple of
tests without a battery, so there hadn't been anything those contacts
could've possibly contacted.

Quite often, when I'm at a total loss (especially when I've yet again
mislaid my car or front door keys), I close my eyes and try to recall what
exactly, and in what order, I have done since I last saw the missing item.
Throwing in the occasional "Ommmm" often helps tremendously. So does sitting
cross-legged, although it hurts. So I saw myself...

a) Unplug the AC adapter

b) Remove the battery

c) Ommmm

d) Remove the back part of the case

e) Unhook the bottom half of the battery compartment

f) Ommmm

g) Clean the temperature resistor contacts

h) Do a, b, d and e in opposite order

i) Ommmm... 

k) What, other that a bad soldering joint or some other type of flaky
contact, could have brought her problem back?

l) Ommmm... 

m) Ommmm... 

n) Ommmm...

o) Now wait a second!

DO NOT READ ON. NOW IS YOUR CHANCE TO FIGURE OUT ON YOUR OWN WHAT THE
PROBLEM WAS. Just for the sake of self-respect, of course. The lottery's
closed, sorry :-)

-----------------------------------

During your contemplation it might have occurred to you that both the AC
adapter and the battery had been disconnected while Lady Newt was under
surgery. She had been completely without power!

I found hope again and removed both the AC adapter and the battery for about
30 seconds. Then I powered her up again. Unfortunately, the problem hadn't
come back. She could still be switched off just fine.

:-(

Although getting older has a lot of drawbacks, it also has the occasional
benefit. This is because the older you get, the wiser you get. And my
accumulated wisdom now made me realize that when one's luck appears to be
exhausted, it's time for stubbornness to enter the game. So I simply
repeated what I just did, this time powering Lady Newt up after about two
minutes of powerlessness.

Lo and behold, she refused to turn off. Since I had recently brainwiped her
and not yet reinstalled the backup containing Christian's software, this
ruled out any issues with said software. This problem was obviously caused
by Lady Newt being powerless for more than a minute or so. Not bad for
investing five seconds of stubbornness, is it?

The question was: Why! If this was a hardware problem, it would always be
there. It wouldn't care whether Lady Newt had or hadn't recently been
brainwiped. So it had to be a software problem. But what software is left
after removing any third-party software by brainwiping the patient?

Yes, exactly! The operating system, or an operating system upgrade.

The operating system lives in the ROM board. The upgrades live in the Flash
RAM chips on the mainboard. So the obvious next step was to replace the ROM
board. Which, unfortunately, didn't change a thing. So maybe an operating
system upgrade was the culprit?

Unfortunately, in a 2x00 Newton OS upgrades are normally a one-way road.
Once installed, you can't get rid of them. Unless...

Unless, a long time ago, you accidently put an eMate ROM board in a Newton
2100. It'll take quite a while, during which the Newton appears to be dead
as a doornail, but eventually it will notice that someone's tampering with
its operating system. Then it'll get rid of any installed operating system
upgrades and start doing what the ROM board tells it to do.

A Newton will power up just fine with an eMate ROM board. Of course, the
display will only show gibberish because Newtons and eMates have different
display parameters. However, switching the Newton off using the switch will
normally work just fine. And it worked just fine with Lady Newt. Even though
she had been powerless for at least 10 minutes before booting with the eMate
ROM.

I then put the original ROM board back in, waited ten minutes and powered
Lady Newt up again. The OS version had fallen back to (717006)-0. The
problem was gone. It didn't reappear by leaving Lady Newt powerless.
Regardless of how often I tried.

Just to be on the safe side, I then reinstalled Eckhart's Y2k patch (which
had been installed when I noticed the problem for the first time). This
brought the problem back. Reliably so. 

This couldn't be more puzzling. The problem never appears with the factory
OS, and it always appears with Eckhart's patch. But these days pretty much
every Newton 2x00 owner out there has installed this patch. So this can't be
a systematic error, or we'd all know about it. I've installed this patch on
at least 100 Newtons before running them through my standard test sequence,
and I've never had this problem one single time.

An educated guess would be that Lady Newt's mainboard has a hardware problem
that will only surface when Eckhart's patch is installed. Given the fact
that she must be powerless for more than about a minute for the problem to
come back, I'd assume the hardware problem to be located somewhere close to
a capacitor. But somehow I find all this hard to believe. Even Newtons that
have been powerless for a year do not act like this.

Be it as it may: I've just dug out another 2100 mainboard that will
hopefully bring Lady Newt back to her former glorious self. I'll keep you
posted. Thanks, Christian, for your patience.

Cheers

Frank

-- Newton software and hardware at http://www.pda-soft.de