[NTLK] Lady Newt - the final episodes

[Aaron H]

Frank, 
Not only are you gifted in Newtonian Tenacity, you’re a heckuva story teller too. 
Thanks for writing that all up, and well done. 

Aaron. 

> On Feb 9, 2020, at 5:02 PM, NewtonTalk <newtontalk at pda-soft.de> wrote:
> 
> Since you explicitly asked for it, here are the remaining episodes of the
> little-known series "Please, oh please don't leave me!", starring (blush)
> yours truly. Most of you know what happened so far, but I'll add a short
> summary of the earlier episodes just in case your memory's as poor as mine.
> 
> Episode I
> 
> Her Royal Highness, also known as Christian's Lady Newt, has ended up on
> Frank's workbench because she's no longer willing to cooperate peacefully.
> She's comatose and doesn't start up, regardless of what kind of reset Frank
> treats her to. However, until recently she has been very reliable. Ever
> since Frank sold her to Christian more than 13 years ago, she hasn't missed
> a beat. So her owner, like far too many Newton owners out there, was more or
> less convinced that since his little green friend had never acted up on him
> in the past, it wasn't going to do so in the future. As a result, the most
> recent backup available was... well... not all that recent.
> Without any kind of technical documentation like e. g. a schematic diagram,
> there's little in the way of systematic troubleshooting. However, an amazing
> amount of defects are actually visible to the naked eye: Burnt components,
> tiny tin balls left over from manufacturing and jammed between two closely
> spaced pins of an integrated circuit, poor soldering, tiny copper traces
> etched away by leaking batteries or other fluids that don't belong in a
> Newton, you name it. So Frank scrutinisingly checks Lady Newt's mainboard
> (henceforth referred to as MLB_1) under his microscope. He finds some stray
> tin balls at places where stray tin balls shouldn't be, but none of them
> creates a short, so removing them doesn't change a thing. Near the end of
> this episode you can see Frank, being out of ideas, replacing the ROM board,
> which doesn't fix things, either. Episode I ends with a close-up of a
> tear-streaked Frank declaring MLB_1 officially dead. Followed, of course, by
> the obvious cliffhanger: Frank straightening himself and slowly walking
> towards the box labeled "2x00 mainboards".
> 
> Episode II
> 
> Frank has found a 2100 mainboard (from now on referred to as MLB_2) labeled
> "Untested". He temporarily puts it into one of the Newtons from his
> Frankenstein family and makes sure that it powers up OK. Then he puts it
> into Lady Newt.  This gives him the opportunity to find out two annoying
> issues: Her display shows a white vertical line, and her backlight is almost
> nonexistent. So he takes Lady Newt apart again, replaces the LCD unit and
> the backlight, and powers her up again. The audience is beginning to think
> that the happy ending is near. Unfortunately, there's one issue left that
> must be checked: Does any of the software Christian needs use a serial
> number based on the unique Newton ID? If it does, it wouldn't be possible to
> re-register the software. The Newton ID is hard-coded in an integrated
> circuit on the mainboard, and said mainboard has just been replaced. So
> Frank would need to transplant this chip from MLB_1 to MLB_2, which he'd
> rather not do unless it can't be avoided. At the end of episode II Frank
> sends a list of Christian's software to NewtonTalk, asking if any of it will
> base its registration key on the Newton ID.
> 
> Episode III
> 
> It seems that Frank is lucky for a change. None of the software requires a
> registration key based on the Newton ID, so Frank installs and registers all
> of the software Christian has meanwhile dug out and sent him per E-Mail.
> Frank has to deal with a couple of software-related obstacles during that
> process, but before long everything seems to be ready for action. He then
> runs Lady Newt through his usual thorough test sequence. When she has
> successfully passed the last test, the audience begins to think that the
> happy ending is near. Their hope is brutally shattered when Frank tries to
> turn her off. She always turns back on immediately after being turned off.
> Even when she's powering down on her own due to her sleep preference
> setting, which is kind of spooky. The only way of getting rid of this
> problem is a brainwipe. This is definitely not the expected behavior of a
> healthy 2x00 mainboard. Episode III ends with a close-up of a tear-streaked
> Frank declaring MLB_2 officially weird.
> 
> Episode IV
> 
> Frank starts a prize competition at NewtonTalk. Everyone who comes up with
> the correct explanation for Lady Newt's weird behavior is promised a brand
> spanking new 2x00 pen shipped at Frank's expense. As expected, replies are
> scarce, so Frank needs to use the little brain he's equipped with to figure
> out his own solution. Hours of meditation eventually yield a result based on
> a miniscule amount of intelligence and a huge amount of luck. The problem is
> caused by Lady Newt having been powerless for more than a minute. The
> problem only occurs when Eckhart's 2010 operating system patch is installed.
> MLB_2 will work just fine with Apple's original OS version, even if left
> powerless for hours.
> Frank briefly contemplates talking Christian into setting Lady Newt's date a
> couple of decades back, thus removing the necessity of Eckhart's patch, but
> decides that this, if accidentally made public, would damage his reputation.
> Since damaged reputations are very hard to fix at Frank's age, episode IV
> ends with a tear-streaked Frank, desperately shaking his head and flipping
> the bird at MLB_2. Followed, of course, by the obvious cliffhanger: Frank
> straightening himself and slowly walking towards the box labeled "2x00
> mainboards".
> 
> Episode V
> 
> Frank has found a 2100 mainboard (from now on referred to as MLB_3) labeled
> "Powers up, otherwise untested". He temporarily puts it into one of the
> Newtons from his Frankenstein family and makes sure that it still powers up
> OK.
> Then he fleetingly caresses Lady Newt, who's still equipped with MLB_2, and
> backs her up to a card because he doesn't want to install and register
> Christian's software again and again. Being paranoid, he engages the card's
> write protect switch. Finally he replaces MLB_2 with MLB_3 and restores the
> backup he just made.
> The result is somewhat boring: Everything seems to be working OK on the
> first attempt. Despite Eckhart's 2010 patch being installed, Lady Newt
> powers off just fine every single time, regardless of the time she has been
> without power. She has also passed Frank's thorough test sequence with
> flying colors. The audience begins to think that the happy ending is near.
> Episode V ends with Frank looking dreamily at his hot-air soldering station,
> followed by no cliffhanger at all. The audience begins to wonder if this
> stupid series is ever going to end.
> 
> Episode VI
> 
> Lady Newt is ready and eager to be sent back. But somehow Frank is reluctant
> to take this step just yet. Quite obviously, Lady Newt and her owner have an
> exceptionally good relationship. Last year Christian publicly confirmed at
> NewtonTalk that he'd give her filet steak and caviar if she wanted it. Frank
> is happy that she didn't take him up on that promise, since file steak and
> caviar would have made troubleshooting much more difficult. However, the
> mere offer speaks for itself. In one of his recent E-Mails Christian even
> expressed his intention of treating Lady Newt to a romantic weekend to
> welcome her back. Although she was still a bit under the weather when Frank
> read that to her, he could have sworn he heard a faint, but happy sigh
> coming from her speaker.
> 
> Of course it's cool to send a dead Newton in and get a working Newton back,
> equipped with all the software you regularly use. But somehow this would be
> as if your wife woke up from a coma, fully recovered bodily, but without any
> knowledge about your common past. Admittedly this might be exactly what many
> a partner secretly wishes for, but Frank is sure that it's different with
> those two. So he's contemplating doing something he has never done before.
> Something that might have never been done at all. Something that gives him
> the creeps.
> 
> Episode VII
> 
> Lady Newt's long-term memory lives in two Flash RAM chips. Of course, these
> chips might have been erased or even destroyed the moment MLB_1 died. But
> maybe they haven't. There might be a fully functional brain slumbering in
> MLB_1, harboring all the common memories of Christian and his little green
> friend. Of course, ripping Lady Newt apart yet again isn't an option. Frank
> is not going to compromise her current health. But if he can figure out a
> way of transplanting the Flash RAM chips from MLB_1 to one of his
> Frankenstein Newtons, he might be able to back Frankenstein up to a card and
> restore said backup to Lady Newt. That way not only her programs would be
> back, but all her data. For all intents and purposes she would be what she
> was immediately prior to her untimely death. Given the fact that the
> backlight is now much brighter, the display no longer has an annoying white
> line and the bottom part of the case is no longer cracked, she'd be even
> better than that.
> 
> However, the prospect of soldering these chips is terrifying. Really, really
> terrifying. They come in a 56 lead TSOP pinout, meaning they have two rows
> of 28 pins each that need to coexist peacefully in less than 14 millimeters.
> Pins are but 0.15 millimeters wide and 0.3 millimeters away from their
> neighbors. Even a slight tremble of your hand would appear like a severe
> case of Parkinson at the magnification required for this kind of soldering.
> So Frank decides to practice with organ donor boards before deciding to
> attempt the real thing.
> 
> It being impossible to heat up 56 pins at the same time with a mere
> soldering iron, Frank buys a tweezers soldering iron kit on eBay. The
> description says that it'll allow him to easily heat up all the pins at the
> same time and then, since it's basically a heated-up pair of tweezers, just
> as easily lift the chip up. Unfortunately, things turn out to be much
> different in the field. It takes ages to heat the chip up, and it is
> impossible to lift it without bending at least half of the pins into
> oblivion and shorting the rest out.
> 
> So next it's the hot air station's turn. This is when Frank comes across an
> additional problem. Removing such a chip with a hot air station requires a
> lot of heat to be present at the same time at every single pin. Ideally
> you'd use a nozzle that's about as wide as the chip you want to remove. This
> much heat, however, will also unsolder and often move many of the tiny
> components nearby. This isn't much of a problem as long as it happens on the
> organ donor board. Most likely you'll never want to get that board back to
> full working conditions anyway, so it doesn't matter if components move or
> disappear. All you want is get the chips off the board fast and without bent
> pins, using as little heat as possible to make sure they won't die from
> excessive heat. After practicing a bit, Frank successfully removes the Flash
> RAM chips from MLB_1 and puts them in a safe place.  
> 
> The organ recipient board, however, is another cup of tea. You don't want
> components to dislodge, disappear, or fly all over your board. After all,
> you need to keep that board in full working condition. You can try wrapping
> it in aluminum foil and cutting out the area around the chips, but you still
> can't be 100% sure that everything you wanted to stay in place has actually
> done so. So Frank decides on a different approach. He'll cut the pins, one
> by one,  off the chip's body, remove the pin-less body, unsolder the pins
> separately from their soldering pads by means of a very tiny soldering iron,
> and finally remove the tin from the soldering pads with desoldering wick.
> 
> Without further ado Frank grabs yet another untested mainboard from his box
> and puts it in Frankenstein. After all, he needs to be dead sure that the
> board will power up and write to at least one of the card slots after the
> transplant. Frankenstein powers up just fine and willingly obliges to
> produce a test backup. Frank begins attacking the pins with his scalpel.
> Unfortunately, since the chips' height is less than a millimeter, he needs
> to cut close to the board's surface. Had Frank taken that into account, he
> might well have succeeded at the first try. But he has recently been exposed
> to a momentary lapse of reason, so he does not only cut the pins, but also a
> fair amount of the traces connected to the soldering pads. Which, of course,
> is only visible after cutting the last pin and removing the now pin-less
> bodies.
> 
> Episode VIII
> 
> Surprisingly Frank, who usually follows Monty Python's excellent advice to
> always look at the bright side of life, isn't angry that he has just killed
> a perfectly healthy Newton 2100 mainboard. After all, you can never have too
> many organ donor boards, can you? With a smile that almost looks genuine he
> grabs another board labeled "Untested" from his box and puts it into
> Frankenstein.
> 
> Capacitor C59 kindly waits until Frankenstein has powered up and the test
> backup is in full swing. Then it explodes spectacularly, emitting smoke and
> smell that only Watts and Volts can produce. Frank thanks his guardian angel
> for making the developers at Apple place this component under the ROM board.
> Had it been located elsewhere, it might have blown up into his face. Frank
> opens all windows within reach although it's cold and windy outside, removes
> the exploded capacitor and replaces it with that from the mainboard he
> killed a moment ago. You can never have too many organ donor boards, can
> you?
> 
> Just to be on the safe side, Frank puts Frankenstein below a sturdy
> cast-iron grill pan, plugs its AC adapter in a power outlet and lets
> Frankenstein simmer for the next 30 minutes. Then he tries making his test
> backup for the second time. This time it ends successfully. Frank cuts the
> pins again, this time making sure the scalpel won't touch the board. He then
> removes the pins separately from their pads just as planned, cleans the
> solder pads using desoldering wick, and finally applies a generous amount of
> expensive flux to both the pads and to the pins of the chips he has removed
> from MLB_1. Then he aligns each chip on its pads, puts his index fingers on
> it to keep it in place, and solders its top left and bottom right pins so
> that it won't move while soldering the other 54 pins.
> 
> The latter turns out to be a tad on the difficult side because the diameter
> of Frank's thinnest soldering tin is larger that the distance between two
> pins. After removing the occasional bridge between adjacent pins, though,
> each pin is soldered, cleaned with alcohol, and generally looking good. So,
> quite satisfied with himself, Frank turns Frankenstein on. Unfortunately,
> Frankenstein's way of saying hello is a flashing backlight, which is his
> kind and considerate way of letting Frank know that he (Frank) didn't quite
> succeed while messing with his (Frankenstein's) Flash memory.
> 
> Episode IX  
> 
> At this stage of a project you basically have two options. You can
> reluctantly admit that you neither have the equipment nor the skills or the
> balls to meet such a challenge successfully, and then proceed to your garden
> to pull up weeds. The other option is clandestinely flipping the bird at the
> mainboard, dreaming of a huge helping of decent single malt whisky, deciding
> against it because you might still need to solder tiny pins and even tinier
> wires under your microscope, and finally whip out your multimeter. Frank
> decides for the latter.
> 
> 112 measurements later he knows that every pin is securely connected to its
> respective soldering pad. So the most likely reason for Frankenstein's
> unruly welcome is a short between two adjacent pins. This gets confirmed
> after another 108 measurements. One of the two chips has a short between pin
> 44, a data line, and pin 43, the positive supply voltage. Unfortunately,
> this short isn't visible, so it must be below the chip. Since Frank isn't
> all that eager to unsolder and resolder the chip a second time, he simply
> unsolders pin 44, bends it up so that it no longer touches the soldering
> pad, cuts the trace between the pad and a via, removes the via's solder
> resist coating, and finally connects the bent-up pin and the via with a very
> thin wire. Then he switches Frankenstein on again.
> 
> Episode X
> 
> Frankenstein powers up just fine. After a while he politely informs Frank
> that his internal store's signature has been altered and that a hard reset
> might correct the problem. Since that would be the worst mistake ever,
> Frank, without further ado, makes the backup it took him so long to prepare.
> 
> 
> The rest is history. Frank restores the backup to Lady Newt. Since he
> doesn't want to snoop around in Christian's private data, he only checks the
> owner info, which lives in the same chips as the private data. The owner
> info shows Christians full name, so things must have worked out as planned.
> This episode ends with Frank, carrying a tiny parcel, waiting at the post
> office. When it's his turn, he seems reluctant to hand it over, but he
> finally does. Then he takes out his Newton, opens Notes, and jots down the
> following:
> 
> Defying the impossible can be a lot of fun. Especially if you win.
> (attributed to Frankfucius, anno 2020)
> 
> ----------------------------------------------------------------------
> 
> http://newtontalk.net/
> http://twitter.com/newtontalk