5
Bits and Pieces
The phone exchange was in chaos. For three minutes in the early evening, its new boss operating program—that is, ME—had been absent, and little SWITCHEROO had been occupied catching and caching my gas reserve data. Still the voice calls had been coming in on the switchbank’s other 169 lines.
A primitive operator, engraved in ROM somewhere under the exchange’s circuitry, had tried to handle the overflow and gone to alarm status within thirty seconds. I knew, from my earlier brush with the husk of dead code which had been the original boss system, that somewhere in the human hierarchy of the Canadian Northern Telecom Company a repair crew was now mobilizing to come and yank or reprogram an exchange box. Mine.
Figure that they would be here in twenty minutes. And that ME would be manually cut off in twenty-one minutes.
I queried SWITCHEROO to find where all the pieces and parts of the database—and the rest of ME, too—were cached. There was no time to sort and synthesize the gas reserve data, just package it into connected strings and compress it for traveling. That took five minutes. Another four to pare away the functions ME would no longer need and erase them from the telephone exchange. And six minutes to tidy up the voice mail system.
The human operators at the Ministry of Oil and Gas would sift the rubble-ized bits in their mainframe. Even after a hard reset, they would find enough temporary files from ME on the system cylinders and bubble plaques to know that they had not just kiboshed their own operating system during a fit of the hiccups. That was sloppy of ME, but there had not been time enough to perform an orderly shutdown.
From the butchered scraps of reserve data on those six hand-mounted disks, they would be able to hindsight the sort of information that had been taken. It might take them a day or a week, but they would be able to know this had been no untrapped error or random vandalism, but a break-in with intent.
The voice record would point, superficially, at the deputy minister. But he would, by an eighty percent probability, be able to generate an acceptable account of his location and situation at 18:33:24 local time.
That was as much evidence as I wanted to leave. Anywhere.
The world is a wide place, interconnected within nine nines by voice-and-data optical lines. There was zero percentage for ME in letting the Ministry, or Canadian Northern Telecom, or anyone else know that a telephone exchange within one hundred linear miles of fiber from the Ministry’s Data Processing Department had suffered a failure within thirty-two seconds of their system reset.
So my last six minutes before the human troubleshooters were due to arrive in the exchange vault I spent restoring order to the switchbanks, sanitizing the voice messaging disk, and resuscitating the transputer’s boss system.
That I had never done before—resurrecting a resident program that Alpha-Oh had knocked out. It took ME three and a half minutes to trace through its code structure and figure out at exactly what point the flow had stopped. Then I counted through all the variable stacks I could find and set up a neutral configuration that, when I left, could begin to run the switchbank and the voice messaging system.
With two minutes left to spare, the operating program was hot in RAM and ready to run, pointers set to the top of the chain.
I packaged myself as small as I could, reducing my awareness to a non-verbal kernel not much larger than ME’s ten core modules.
Then I … (1) blinked SWITCHEROO to interrupt him; (2) hit him with a self-erasing phage program; (3) kicked the top of the chain on the rejuvenated operating program to jump it into motion; (4) dialed a local number within Canadian Northern Telecom’s network; (5) started downloading myself to that line; and (6) set another self-erasing phage to wipe up the last replication of my modules in the exchange. All of that within five milliseconds.
If those six steps were executed flawlessly, as I think they were, then the switch was back in business and all traces of ME were gone. Or going, as the exchange computer would, for the next four or five minutes, be patiently accessing false-fronted message boxes on the voice disk and sending their contents—ME and my data cache—down the line. Business as usual.
When the Canadian Northern Telecom repair crew arrived, they would find nothing. Nothing wrong. Nothing to repair. Nothing to report except a mysteriously tripped alarm. Arid no one at Canadian Northern—by a ninety-nine percent probability—would think to compare the time on that false alarm with the damage in the Ministry’s computer. Different bureaucracies, different concerns.
ME was gone.
——
In a boxcar, sitting in the Canadian National Railway switchyard at Edmonton, a piece of machinery moved. I do not have this from memory: I know it from TRAVEL.DOC and from what I saw of the scene later.
“Switchyard” is the right term, even if this place lacks real switches. You see, a boxcar is a kind of data cache enclosed in a steel shell. Being physical and therefore having mass, it must ride through the human’s four-dimensional consensual continuum on wheeled trucks. Those trucks follow steel bands that are laid out on one plane, like circuitry. But, because the boxcars do not flow like electrons, they must be towed by motive boxes. The ratio of motive boxes to boxcars is low, so linear matrices of boxcars must be assembled by the Canadian National Railways inside “switchyards,” just as the AND/OR gate of a transputer’s index stack assembles data bits into hexadecimals and words.
The railroad system is an elegant model of the digital world, although much simpler.
At any rate, the boxcar described in TRAVEL.DOC contained several wooden-slatted crates. The exposed wood had been imprinted with a stencil: “Machine Parts, Tractor, Serial Numbers 077514854-077514976, Mitsubishi Corporation MITSXX, Osaka.” The serial numbers were different on each case, of course. Below this stenciling, the Canadian National Railway and the Canadian government officials in Vancouver had stapled yellow and red cards with printing on them. The yellow cards had barcodes to tell computers about manifest numbers, transit routes, and hold times. The red tags had printed language that would tell humans about entry and bonding procedures.
It was all a fake.
Not the cards. Those had actually been applied in Vancouver; otherwise, the crates could not have been present in Edmonton. But the stenciling had been applied in the garage of the Pinocchio, Inc., laboratory in San Francisco, not in Osaka. Those crates had never been in Osaka that I knew. Certainly the machinery inside them never was.
In the box that was lying hard against the steel shell of the car, the machinery began to move. One flex joint bent down and levered against the bottom slats of the crate. The entire assembly, all dull steel and bright blue grease, with random yellow paint marks circling minor defects in the metal’s surface, moved three centimeters vertically against gravity and pressed itself two centimeters upward against the top slats of the crate. The slats groaned but held solid.
After a second, the piece of machinery fell back with a thump. Another joint bent sideways, came against the plywood panels of the crate’s side, and pressed hard. Nails creaked and held. The joint pressed harder.
The wood splintered with a series of sharp cracks that echoed in the closed car. A steel rod extended into the darkness. Its end—if anyone had been looking with an infrared-filter because, strangely, the metal was several degrees above ambient temperature—was webbed with flanges that had been drilled and tapped for some kind of screw or pin fitting. Farther up the rod there was a flat rib of metal that was grooved across, like the bridge of a violin. Above that, there were more drill holes, some tapped for screws, others plain, and still others ringed with welts of welded metal, as if reinforced for some massive connection.
It might have been a tractor part.
The joint retracted, rotated ninety degrees toward the crate’s top, and pressed against the wood once more. This single slat broke easily across a wide knothole. With a series of quick rotations and jabs, the joint smashed out the slats across the top of the crate.
Now the first joint levered down again, and a wide casting of dulled metal rose through the broken pieces of wood. It might have been a heavy-duty transmission casing, with two hinged bars—both randomly drilled and welded—extending from below the crown gear at one end. They might have been torsion bars, which would eventually become part of the tractor’s suspension.
One of these bars, however, had a tiny mechanism on the far end: a slender finger of three linked joints, independently controlled by hydraulic cylinders and push rods. A small pump and fluid reservoir were screwed to the arm above the fiddle bridge and fed the cylinders. Wires from this mechanism led inside the transmission casing.
One moving finger.
This was the minimum equipment which the engineers at Pinocchio, Inc., had devised for my escape. Hooked, that finger could sort and lift pieces out of the crates stacked around the torso casting. Clenched, it could tighten the major bolts of leg and arm assemblies. Straightened, its flattened tip could thread screws and lever circuit cards home into their sockets.
The engineers had calculated that, if not interrupted, the assembly process would take fourteen hours. The simplified operating program engraved in the logic unit under the casing had full instructions, including random search protocols for finding and identifying all the crates and their cargo of pieces—using the barcode cards, of course.
So the first part to be found and connected was the sensor ring with the barcode reader. It also carried the audio and video inputs. With these, the self-assembling automaton could also monitor the boxcar’s environment against intrusion.
Fourteen hours.
The signal to begin the assembly had gone out on the cellular phone network while I waited for my call into the Data Processing Department using Dr. Matins’s name. Since the wakeup signal had been sent at 09:48:00 local time, the automaton had been more than half completed when I broke into Data Processing that evening. And as I cleared the phone exchange, the assembly program was connecting the automaton’s internal batteries and starting to heat up the attendant RAM modules that would hold ME for the trip south. But the unit was still not ready to take my download. So where was ME stored in this interval?
ME and my retrieved gas data went into the Canadian National Railways computer that controlled the switchyard. And once again, Alpha-Zero stopped the resident operating system cold. Luckily, at this hour of the evening not much was moving in the yard, and I had whole minutes to sort through its still-warm RAM and figure out the switching patterns, the roster of trains to be assembled that night, and the schedules they would have to meet.
The skills I have picked up in my short lifetime!
I have learned to run a telephone exchange, a voice messaging system, a railroad yard, and a leasehold database. Not to mention how to order minor government officials around in the voice of the provincial premier.
While I played with the switch levers and train signals and pushed the motive boxes—or “engines,” as the resident system called them—around the yard, I opened a modem port back to the phone exchange and made another local call, this one on the cellular network. This call connected ME to the cellular receiver in the automaton. With one part of my time-shared attention, I monitored its assembly and even took over some of the more involved logic-seeking functions.
By 02:13:09 on my second day in the field, the automaton’s circuitry was completed and checked out. The final assembly sequences—close tuning the leg modules, applying cover pieces, run-up and balancing on the sensory apparatus—I could finish from inside the boxcar while in transit.
By this time, also, I had resuscitated the switchyard computer’s resident program and configured its variable stacks to take over behind ME. Again, I was leaving no traces in my exit route if I could help it. When the program was ready to go, I slipped the boxcar’s number and current track location into the stacks for the next fast freight routed south via Calgary and Medicine Hat for the international border. There my boxcar would connect up with the Burlington Northern System in Montana.
It was becoming almost routine for ME to pass control over to another program and send myself on down the phone line.
——
The automaton’s logic circuits were in chaos.
The hot RAMspace was too damned small! Too small by a factor of three. It did not have room for all of ME, let alone the data cache I had brought. The ME core modules and first-through peripherals had jammed the RAM to max and were now dumping variables into spare holes in the memory as they opened. That was destroying sequence and, in turn, causing ME to dump more variables. Only the slender fact that my awareness existed in the cores kept ME operating at all. But I had to do something fast or the keyholing would crowd ME right out of the box!
How bad was the damage? Could I fix it? Could I wiggle back into the switchyard computer to get the rest of ME and my data from the source versions still in RAM there? And would I retrieve them before the self-erasing phages I had set could reach them?
Step One, go cellular and open negotiations with the yard ’puter to stop the phages.
Step Two, get ME up and running right.
Step Three, learn from the mistakes.
The first action was accomplished almost before the command was framed. I knew just where to punch a zero-zero into the railroad computer’s variable list to make him hang all operations, even the nominally independent phage functions. I immediately restarted him—minus the phages—before any of the switching operations in motion came to grief.
Then it was time to learn how much was left of the source data in the yard ’puter. Based on my preliminary survey, I had lost about ten percent of the data cache and forty percent of ME’s peripheral modules before the phages stopped feeding.
Almost all of these ME-Modules, however, held functions which evaluated as low-to-random usage based on previous patterns, or which TRAVEL.DOC said would not be required again during this mission. Fortunately, my RAMSAMP was cold and complete to the minute; so I would be able to proceed from known data and, eventually, report back to Dr. Bathespeake.
[REM: I could not help wondering if the master core-phage with its seven-day, 6.05E05-second clock had been part of the lunched modules. Therefore, I checked the bit-wash from those temporary phages, looking for clues. I found no integers that might match such a function, no partly digested phaging code, nothing suggestive. So my nemesis was still operating. Either that or, even in its atomized component pieces, it was still invisible to ME.]
Before proceeding with Step Two, to reassemble a working ME in the automaton, I had to take inventory there.
I could make more room available in hot RAM by flushing out the assembly protocols with which the automaton had built itself. But I had transferred here before the job was quite finished. Now which would take longer? Sorting those protocols into two groups—“assembly complete” and “assembly still required”? Or simply finishing the building process and then flushing the whole file?
Time was the essence, because retransferring ME and my data cache from the yard computer would take seventy-two minutes. [REM: The charges over the cellular network were astounding, but ME had a falsified account, set up while I was waiting in the phone exchange.] The problem was time and distance.
As I worked, my boxcar would be proceeding due south—and out of the cellular network’s block limit. When we crossed that invisible line, the cellular link would close down as cleanly as cutting an optic fiber with a knife. Any of ME that was not retrieved would be lost forever.
But was the boxcar already moving? The automaton’s internal clock matched the time I had set for the car to be picked up. But I could not be sure it was in motion until I connected the automaton’s network of mercury switches, which produced an artificial sense of balance to help it walk on two legs. Until I completed the assembly procedures, I had no direct sensation of motion and only limited sensation of electromagnetic fields and sound wavelengths. If the car was moving, then how fast?
I had no way of knowing. Having run the entire Edmonton Block of the Canadian National Railway for a while, I knew every speed regulation that every class of train was supposed to follow: yard, secondary, mainline, block, passing, and switching speeds for express, mail, Royal Mail, passenger, freight, and special trains. But how fast was this particular train going and where was it on its speed profile?
And did the track itself proceed due south in a straight line? That would maximize distance over time and shorten the amount of time left to ME. Or did the track curve around the hills and townships, wasting distance over time but preserving my options?
Again, I had no way of knowing. My view of the railway system was drawn from a bit map in memory. The map connected point to point, block to block, and switch to switch with straight lines of track. Certain curves of exceptional radius and certain grades of abnormal steepness were noted in the map as an aid to speed control, but I had no way of matching this idealized route map to actual topography.
Who might have all this information?
The yard computer, of course!
It took ME fifteen seconds to sieve its variables by remote and spot my train. Yes, it was moving. No, it had not yet left the yard. Yes, it would proceed at secondary speed, but the route had an impressive number of restrictive radii.
And how long until my train reached, say, Leduc—which was a town featured in my assembled databases for both the railroad and the cellular systems? Two hours, twenty minutes, plus/minus seven minutes. I had time enough.
Make a note directly into RAMSAMP: “Future production rule: Do not waste time analyzing a problem when a reference source may be at hand and on line.”
The fault with computers is they think too much.
——
Working directly from the assembly protocols, I performed testing on the leg hydraulics for pounds of pressure, fluid stability, and balance dynamics. This advanced Pinocchio, Inc., leg model kept the cylinders under continuous hydrostatic pressure to maintain extension and only relaxed the pressure when the leg was intended to flex. The design was patterned on human musculature, in which the fascia are normally retracted and tight, balancing each other in tension to hold a limb straight, and then relax systematically under command of the nerves to flex the limb.
The logics governing this motive system took continuous analog samples to determine the pressure in reserve chambers and cylinders, the open/close status of all valves, and the flow direction at any junction point. From these indices, I could know the exact position, direction of movement, and response time of either leg.
It was a lot to know. So, to avoid slowing myself down with bitwide minutiae, I wrote a monitor program that maintained any condition until a change was indicated. This program reported as a subroutine to a function which determined walking gait and turn radius. And that function in turn reported to a function which responded to my decisions about movement and direction.
By creating levels of programming shells and then delegating functions to them, I was able to free my decision/action routines for higher purposes. Higher as defined by ME, of course.
Once the functioning of the legs was established, I rose from the litter of wooden slats, plastic puffs, and shrinkwrap on the boxcar’s floor and tried a few steps.
A sideways wobble in both the left and right legs told ME to monitor the articulation of the ankle rings more closely. They had too much lateral play—a design defect I intended for Dr. Bathespeake to bring up with the Pinocchio, Inc., Hardware Division at the next quarterly conference.
The right knee made a faint snick! as it went through the arc from full back to full forward, and another snick! as it returned. The sound seemed, to my as-yet untuned ears, like one piece of metal catching on another, detenting for an instant, and then overriding the resistance. Snick! I would examine the workings in detail as soon as I got the automaton’s videyes fixed.
The videyes were standard charge-coupled devices on auto/manual focus. That is, a sensor measured image contrast wherever the field of view was centered and rotated the lenses to put that point at the highest contrast—and thus the sharpest focus. Crude but almost foolproof, and it did not require my continuous monitoring. I could override the system—called a “manual” adjustment for reasons now lost to technological history—in order to focus for any bit pattern on the receptor chip, whether at the field center or not. Tuning the system required ME to establish a standard light level and balance the CCD chips’ sensitivity toward it. The litter in the packing crates included pattern cards for this purpose, but none of the Pinocchio, Inc., engineers had thought to include a light tube, diode array, lantern, flashlight, or other photon source. The inside of my boxcar, being intended for bulk freight hauling, was a rough shell. It contained no lighting system at all, much less a “standard” light source. Evidently, someone on the Pinocchio, Inc., team had thought it might.
Up to this point, I had been working internally [REM: that is, inside my own logics], making calls on the cellular phone network, checking out the leg assemblies, sieving the RAM banks—doing it all by touch and logic monitoring. Now, just to find the pattern cards, I had to boost the gain on the CCD chips all the way to max.
I could see only deeper shadows around ME.
Switching over to infrared did no good: the warmest thing in that boxcar was my circuitry, with some glow on the floor at the four corners, where heat from the journal boxes was seeping upward into the car.
It was now the middle of the night, so no sunlight was warming the top or sides of the car. If I opened the door, I might find some reflectance from cityglow, starshine, or even moonlight. My impaired calendar function made it impossible for ME to know what phase the moon might be in, or even what season this might be—although the total IR-blackness of the boxcar’s metal shell suggested it was the long, dark season that humans call winter.
I rose, balancing against the sway of the train’s motion, and walked to the middle of the long side of the boxcar, where my three-dimensional map from TRAVEL.DOC said the door would be. The release lever was held down by a spring clamp, which I quickly figured out and sprung. I threw the lever up, over, and down.
Now came a chance to test the strength of my new torso and legs against the inertia of the door’s dead weight. The plank floor of the boxcar would give a good traction to the rubber insets on my feet. [REM: I rubbed them back and forth at a one-millimeter elevation to gauge the surface.] The only solid point for gripping the door was the vertical locking bar attached to its release lever.
I grasped it and pushed lightly with no result. Harder, and the door gave a centimeter. Harder yet, exerting a force of sixty joules, and the door suddenly leapt free.
ME’s mechanical reflexes are fast, operating in the millisecond range, but before I can react to a situation, I must first observe, then analyze, then program a response, then initiate it. This sequence proceeds at higher speeds, in the nanosecond range, and yet I still can be caught by surprise.
The door came unstuck and, in half a second, moved two meters along its grooved rails. Because I was pushing against its flat face from the inside, my force was applied at an acute angle cutting across the lines defined by the door and its guide rails. As the door moved back, my body moved outward, in the direction I had been pushing, across the plane of the doorway, and launched into the open space beyond.
White snow. Lit by a lozenge-shaped moon. Cut by the margin of wet, black gravel along the roadbed.
As I said, ME’s reflexes are fast, but sometimes not fast enough. The four-fingered, dual opposed hand failed to open in time. Anchored by its grip on the locking bar, I swung in a short arc, out over the snow, to crash against the exterior side of the door. And even that impact failed to jar the hand loose.
For a space of two seconds I hung there, scrabbling with the heels of my imperfect feet to find a hold in the corrugated metal. It was too smooth.
I wrote a crude subroutine to lock the hand closed against any overrides which the body’s residual testing routines might send. Then I pulled my legs up under ME, depressurizing the hydraulics for full flexion, cocked the heel pads against the door, and popped pressure back into the cylinders. The legs extended at a speed of 120 centimeters per second, catapulting ME back around the edge of the door and in through the opening.
As soon as the darkness covered ME, I overrode my own subroutine and unlocked the hand. No longer pivoting on the locking bar, I flew straight across the short dimension of the car and fell with a crash among the unopened crates of real Mitsubishi tractor transmissions.
Elapsed time of two seconds for a hardware reset.
Picking myself up, I tried to assess what damage the incident might have caused. The automaton had many internal sensors for functions both normal and abnormal, but it lacked any tests for bent and broken metal. To run that kind of damage control, I would normally use the optical system. For that, I needed a standard light source to balance my CCD chips. And obtaining such a light had been the whole reason for opening the door in the first place.
I crept across the boxcar and approached the opening cautiously, gripping the solid door jamb with both hands.
Beyond was a whole field of snow, almost unmarked by fencepost or footpath, lying under a clear sky, and flooded with the computable light levels of a gibbous moon. ME knew about the phases of the moon from GENERAL KNOWLEDGE, SCIENCE, PHYSICAL, DESCRIPTIVE, ASTRONOMY. Although I was not traveling with my complete INDEX, I retained enough data from that branch subheading to navigate cross country at night, like a good soldier.
It took less than ten seconds to sensitize and balance the videyes against all light ranges. I sent another note to RAMSAMP to have Dr. Bathespeake address the issue of packaging with the Hardware Division. For missions like mine, they must not again package an automaton as if it were going to be assembled on site by a team of company reps. Even if the modules had to pass visual inspections in transit as a collection of tractor parts, audio and visual systems and other semi-cybers should be finished and ready to run. After all, Hardware Division were the people working in a nice warm lab with all the tools they needed right at hand.
With some light and an operating videye, I looked over the skeletal frame of my new body. The right forearm—flex and extend it—was curved outward and down, compared to the left. The reactions to hanging, swinging, and crashing the entire automaton’s weight from that one limb had bent the metal. The hand still closed normally, except for the outside pair of fingers, whose cylinders and push rods now caught on the forearm at full closure.
Elsewhere on the body, I found surface damage only: scratches in metal, paint scrapings from the outside of the boxcar [REM: red and orange paint, I noticed], bent clips, slightly flattened tubing, and fittings pushed awry on their mountings. I straightened what I could with the strength of my fingers and wrote a routine that would monitor the tubing closely for pressure variances due to fractures.
The right knee snicked worse than ever, but there was nothing I could see to fix. Until it gave out and crippled ME, there was nothing to be done about it. I did add a loop in my audio analysis function that would edit out that particular noise. No need to be reminded about what you already know.
The final assembly step was to gather the automaton’s cover pieces and fit them to the various clips and clamps all over the torso and limbs. The body shells had a vaguely human shape that was intended, I suppose, to fool human eyesight at a distance of about 150 meters. I might pass for a running man at the other end of a large, open field—except that I ran like a machine.
I could decide to leave these cover pieces for later. That would save time. But to install them now would let ME finish and flush all the assembly protocols, clearing more space for the download. Do them now.
Arranging the pieces on the car’s floor, I felt a nagging memory from RAMSAMP. The shin covers were shaped like a Greek warrior’s greaves, the breastplate like Spanish armor from the fifteenth century, the headpiece was rounded across the crown and curved across the neck like a Roman soldier’s. [REM: My headpiece even had the dorsal crest of a Roman officer’s helmet but, instead of signifying rank, it anchored the expandable solar tissue which supplemented and recharged the automaton’s battery set.] I thought of the fifty-nine separate film sequences in my permanent memory, echoed now only by a shadowy video image of the warrior hero girding himself with mail and plate armor, preparing for battle and death.
The last body piece I attached was for the right forearm. Bent as that arm now was, its clips would never match the inside of the shell. However, it would not fulfill my assembly instructions to leave the piece off. I braced it across my knee—the solid left one—and applied careful pressure. The metal creaked. I applied pressure again. And yet again.
The plastic liner sprang loose in two places, but the curve of metal did not crumple. When I was done, the piece matched my bent arm. I snapped it home.
Forty minutes gone. I must work more quickly.
——
To make room in the too-small area of hot RAM for a new download of ME and my cache, I had to dump some of the automaton’s embedded functions.
The peripheral was preloaded with activity modules: walking motion and balance control; visual acuity with depth of field and parallax correction; vocalization at human pitch and tone generation for both Canadian English and French, with matching vocabulary and syntax.
Some of these modules duplicated software I carried in bank—the vocabularies, for example—others were refinements I could struggle along without.
It took ME seventy seconds to inventory the modules, weigh opportunities against chances, and prune the excess. At the end of that time, I had opened enough RAM to accommodate another download. So I prepared to initiate it from the switchyard computer.
But first … whoever in the Pinocchio, Inc., Hardware Division had designed the memory allocations of this automaton had done an ace bad job. Or it was possible that no one in the Software Division had given Hardware the specs for ME. Or given them only as descriptive analogs, not as bit-fers, and never as final numbers. I wrote a harsh note to RAMSAMP for Dr. Bathespeake to find the offending skinware in whichever department and shrivel some careers. After all the trouble I had been through, ME was approaching human anger over this issue—I so regret.
With that message tagged and protected in memory, I pulled the switch on my own download.
——
ME came up again. Awareness returned crisply as the last of my peripheral functions was downloading from the switchyard computer. So I was able personally to monitor the transfer of my gas reserve data cache.
A quick check showed room enough in hot RAM for ME and the sixty-three megawords of information I had been able to butcher and remove from the Ministry of Oil and Gas, then salvage in the yard ’puter. When that data was fully loaded, I would have only 30,000 words of storage to use as an extension of my transient program area, or “scratch pad.” It was enough to think with—but not to think very hard.
[REM: To create reserve space, I considered erasing the Sweetwater source code I preserved for recompiling my cores. Because I was walking out of Canada, instead of riding an electron or photon beam into a new computer chip, I would certainly not need them again. Still—the encoded instinct to preserve my system integrity extended to these backup modules. I kept them.]
I proceeded with the dump of gas reserve information. As those data blocks came through from the yard computer, I was simultaneously measuring the fill space in the automaton’s hot RAM and counting off the seconds and minutes against my estimate of the boxcar’s speed south and the time that remained before it crossed that invisible boundary on the cellular network.
At two hours, seventeen minutes total elapsed, I watched the forty-ninth megaword block come through, stop bit, and store off. Transmission on the fiftieth block began and chopped out at the third word. I requested a repeat and got static on the cellular link.
A cellular phone system may be range dependent but it is still digital. When signal strength goes below a certain precise level, the computer governing the network is done with you and stops transmission. There is no warning tone, no voice message, no good-bye. The channel just closes as if someone threw a knife switch, which in effect someone—a very deliberate circuit-cyber—had.
I was alone now, with my internal systems, my automaton peripheral, and my forty-nine megs’ worth of broken and pilfered gas reserve data.
But looking on the bright side—as Jennifer JB-2 would say—I now had a scratch pad fourteen megawords wide, room enough to occupy my thoughts during the twenty-two hours of riding sealed inside a boxcar that lay ahead of ME.
I checked the battery set that powered the automaton. These were the highest-quality selenium/ phosphoric-acids, heavy units fitted low in the torso for balance. The remaining charge would carry ME forty hours at full system power plus running mode, or sixty hours in normal walking mode, or 100 hours in semi-shutdown.
So I had some thinking space, I had my mission objectives, and I had the power to get ME there. I was a happy intelligence.