158: Chapter 158 The Universe Doesn't Accept Almost

August twenty-seventh, evening.

The sun had not completely set yet, merely sliced into a dull red patch of light by the cluster of teaching buildings to the west. Slanting across the edge of the sky, it singed the edges of the clouds into a layer of fuzzy gold.

The wind picked up—that characteristic northern evening wind carrying the scent of dust and earthy grass. Sweeping straight down the main avenue, it gradually lifted the dry heat accumulated on the asphalt road during the day.

The shouted drill commands on the sports field had ceased, leaving only a few scattered square formations wrapping up.

Loose laughter and chatter of freshmen echoed from the direction of the Zijing Living Area.

That laughter was unpolished, imbued with the exhaustion and relaxation unique to the seventh day of military training.

Having survived the hardest first few days, the shell of unfamiliarity between people had been softened by sweat. Now there were arms thrown over shoulders, mutual nicknames, and boisterous banter calling out names across dozens of meters.

Jiang Lin stood beneath a plane tree by the roadside, his camouflage jacket slung over his arm, his collar revealing a neck marked by a faint red tan line.

He was looking down at his phone.

A publicly available NASA flowchart was open on the screen.

LOX.

LH2.

Engine Bleed.

Engine Chilldown.

RS-25.

Zhao Chengyu passed by holding a water cup. He had originally intended to call him to the cafeteria, but when his eyes caught the screen, his steps paused mid-stride.

"Artemis I?"

He leaned in to take a look, his gaze swiftly sweeping over those lines of English abbreviations.

"LOX, LH2, Engine Bleed, Engine Chilldown... this is the cryogenic propellant procedure before an SLS launch, right?"

Jiang Lin looked up at him.

"It's the first step in humanity's return to the Moon. The whole internet has been buzzing about it these past few days."

Zhao Chengyu shrugged and said.

"Though I'm just watching for fun. Liquid hydrogen chilldown, engine temperature, launch window—I can roughly get those. But if you really break down RS-25's ground support system, that's a whole another world."

Jiang Lin tilted the screen slightly toward him and explained, "Before RS-25 ignites, it relies on the bleed and chilldown procedures to condition the engine and related piping into a state where cryogenic propellant can enter stably."

Zhao Chengyu nodded in agreement, "Yeah, once liquid hydrogen enters, the temperature difference is too massive. Metal contraction, seals, valves, sensor readings—everything will glitch. I understand the logic, but why are you looking at this?"

"I was thinking about cooling issues," Jiang Lin said.

Engine chilldown wasn't meant to make the rocket look more sophisticated, nor was it engineers looking for trouble out of boredom.

It existed because the vast temperature difference between metals, seals, valves, sensors, and cryogenic propellants had to be squeezed into a controllable range before ignition.

Even a single degree off was unacceptable.

In fact, what was truly dangerous wasn't the cold itself.

It was the shrinkage, leakage, reading distortion, and localized stress caused by sudden cryogenic temperature changes.

It was the tiny deviation quietly accumulating in milliseconds at an unmonitored interface, capable of causing the entire system to collapse.

Jiang Lin was not unfamiliar with this logic at all.

The multilayer high-pressure cooling lines on the PMCU-17 wreckage had once given him the exact same sense of oppression.

When a high-energy system collapses, the core doesn't necessarily explode first.

A more common beginning is an inaccurate reading in a cooling bypass, an uncrossed phase-change window, or a temperature curve failing to drop as expected.

Just a tiny fraction, slowly and silently pushing the entire system toward the point of no return.

By the time someone realized it, the extreme heat had long since burnt through.

Zhao Chengyu waited on the side for two seconds.

He watched as Jiang Lin's eyes slowly drifted away, floating to somewhere he couldn't see or even imagine.

That absentmindedness was unique; it wasn't day-dreaming or spacing out, but rather like a person's consciousness had suddenly been pulled into another time and space, leaving behind merely a shell standing beneath the plane tree.

He didn't continue speaking, instead waving a hand in front of Jiang Lin's eyes.

"Hey—Earth to Jiang Lin."

Jiang Lin's eyes refocused.

"Alright," Zhao Chengyu chuckled, showing a row of white teeth. "Then I won't disturb your research into Engine No. 3's body temperature. I'll go research the doneness of cafeteria chicken legs first. Between these two topics, I think the latter holds greater practical significance for me."

Holding his water cup, he took two steps away before turning back around with a half-joking seriousness on his face.

"Hey, if the day ever comes when you're going to the moon, remember to tell me in advance. I can be responsible for standing beside you and asking useless yet particularly sharp questions. Like, 'Are you sure this screw is tightened?', 'What if it can't launch?', or 'Did you bring enough dry rations?'"

Jiang Lin smiled and said, "Competition for that role will probably be pretty fierce."

"Then I'm signing up right now—first come, first served." Zhao Chengyu raised his water cup with utmost gravity as if taking an oath, then turned around and sauntered off toward the cafeteria.

Zhao Chengyu walked off with his water cup.

Jiang Lin watched his figure vanish under the streetlights of Zijing, put away his phone, and looked up at the sky.

An evening in Beijing was quite ordinary.

The streak of red at the horizon was fading, gradually giving way to twilight blue.

There were white vapor trails left by airplanes, birds returning to their nests, and lights turning on one after another in distant high-rises.

There was no crimson aurora.

No anomalous star flashes.

No curved band of light traversing the celestial dome and slicing the entire sky in two.

Yet certain lines had already connected in his mind.

Moon.

Cooling.

Launch window.

PMCU-17.

TM-7.

Seventh Canopy Station.

These words were like nodes scattered in the dark, now strung together one by one by an invisible thread.

Jiang Lin stood under the plane tree for a long time—until the streetlights fully illuminated, until the last square formation dismissed, until the laughter in the evening breeze gradually faded away.

August twenty-seventh, 11:40 PM at night.

The fourth-floor corridor of Zijing Building 17 grew quiet; the motion-sensor light at the end flickered on occasionally, only to go out again after dozens of seconds.

The curtains were drawn in Room 402.

The desk lamp was turned to its lowest brightness, casting cool white light onto a supply list.

Jiang Lin sat at the desk, pressing down on a pencil with his finger.

This list was no longer a gap-filling checklist like during previous trips to the Wasteland.

For the eleventh trip to the Wasteland, the objective had already moved beyond survival and single-point verification.

Northeast Seventy-Three Abnormal Collapse Zone.

PMCU-17.

TM-7 Regional Underground Calculation Array.

Outer Ring Service Layer.

Seventh Canopy Station.

FDSO-2076A.

Dark Dust String Hypothesis.

These terms pressed onto the paper one after another, each weighing heavily.

Jiang Lin wrote at the top of the list:

[Northeast Seventy-Three Second Expedition]

Below was another line of notes.

[Infer TM-7 outer ring thermal management network and inner ring seal boundary from PMCU-17 exposed wreckage point.]

He drew the first horizontal line.

1. Expedition Platform.

G-Explorer-B had already proven itself in the previous round in the Wasteland.

It could travel long distances, cross unstable scree slopes prone to landslides, and autonomously return under low visibility and weak communication environments relying on a logic system that wasn't particularly smart, but reliable enough.

But that was still not enough.

The Northeast Seventy-Three Abnormal Collapse Zone was far more troublesome than ordinary wilderness.

That region had thin-shelled terrain, anomalous thermal inertia bodies, deep engineering wreckage, residual electromagnetic interference, and underground structural echoes that remained unexplained to this day.

Sending G-Explorer-B again as it was would be wasting the opportunity.

[G-Explorer-C Modification Target: Low-speed, high redundancy, disposable, relay-capable, capable of bringing back minimal evidence.]

Then came the detailed items, written one by one with extreme restraint.

Foot protection layer: Two spare sets of POM-CF, plus a replaceable sacrificial layer.

Joint seals: Add fine-dust barrier rings.

Main Control: Dual-storage synchronous write, forced cache flushing to disk before power-off.

Communication: Short-range high-bandwidth link not used as primary chain; low-speed steady-state telemetry prioritized.

Lighting: White light and near-infrared dual channels.

Sensing: Thermal infrared, magnetic field, low-frequency vibration, residual gas, weak radiation counter.

When writing weak radiation counter, his pencil paused for a moment.

The PMCU-17 unit itself did not necessarily carry radiation risks.

However, what kind of power supply structures, shielding structures, and sealing mechanisms a regional underground calculation array had undergone during the Cataclysm era was something nobody knew.

The Wasteland wouldn't spare him a hazard just because he brought one fewer piece of equipment.

Jiang Lin retained this item.

2. Surface Re-measurement.

This time, he needed to find out whether there were still un-zeroed remnants of engineering states underground.

If that record in PMCU-17's offline cache wasn't an artifact, and if TM-7's inner ring had once entered a frozen-state keep-alive process, there might not be an obvious heat source deep in the collapse zone.

It might only leave behind a localized thermal hysteresis.

Just like long after a person has died, certain deeply buried tissues still retain a trace temperature difference that defies environmental laws.

Or perhaps periodic low-frequency vibrations—a faint pulsation so regular it didn't resemble a natural phenomenon.

Faint heat exchange brought about by residual cooling paths.

Or anomalous heat capacity boundaries left behind by certain cooling lines, sealing layers, or inner ring isolation structures.

Jiang Lin wrote on the list:

[Drone Payload: Thermal infrared priority higher than HD aerial photography.]

[Imaging Window: Once after dusk, once before dawn.]

[Target: Re-verify whether thermal anomalies possess new localized source terms.]

[Focus: Search for faint thermal hysteresis matching inner ring frozen states, residual cooling paths, or low-power keep-alive operations.]

3. Underground Approach.

He paused even longer under this entry.

When PMCU-17 was dug out back then, the greatest danger came from the surrounding strata.

Thin shells.

Cavities.

Collapses.

Scree slippage.

Low-frequency resonance attenuation anomalies.

Any misstep could cause the ground surface to collapse, swallowing both human and equipment into the darkness thirty meters below.

Jiang Lin could certainly send G-Explorer-C in first.

However, if the next objective was to confirm whether field-corresponding evidence still existed for the inner ring frozen-state keep-alive record, surface and fissure-edge telemetry alone might not suffice.

Deployable nodes were needed.

Relays were needed.

Winches were needed.

Severable cables were needed.

It was necessary to abandon equipment rather than human life under the worst-case scenario.

He wrote down:

One set of mechanical winch.

Two spools of high-strength aramid rope.

Traction pulley block.

Six deployable relay nodes.

Four disposable environmental recorder boxes.

Enclosure requirements: Drop-resistant, dust-proof, moisture-proof, able to withstand short-term electromagnetic interference.

The recorder box functions were simple.

Temperature, humidity, barometric pressure, magnetic field, vibration, timestamp.

Write once every ten seconds after landing.

No networking required.

No real-time transmission required.

As long as it could be hauled back next time, it would leave behind the facts of that darkness.

If it couldn't be hauled back, that was just the way it went.

Some equipment was born to be sacrificed.

4. Cooling.

This entry shouldn't have been a separate section.

From a supply perspective, cooling was merely part of underground detection and thermal inertia observation. Thermocouples, infrared temperature measurement modules, insulation materials, phase-change samples, and temperature loggers could all be categorized into Class 3 or Class 4 crates.

Yet Jiang Lin stared at the word "Cooling" for a long time.

One of the most abnormal parts on the PMCU-17 wreckage was its cooling interface.

This specification exceeded the requirements of a small maintenance unit.

Multilayer high-pressure piping.

Crystallization on the inner wall of the fracture.

Interface redundancy.

Localized ceramicized ablation marks.

The pipe diameter and wall thickness far exceeded the heat dissipation demands of an peripheral maintenance calculation unit itself.

In the winter of the nineteenth year, Jiang Lin had stood before the emergency wake-up interface, withholding pressing the power button for ten full seconds.

He had wanted to know if electrical signals still lingered inside that piece of black metal.

Even if it was just the most primitive beep when the motherboard powered on.

Even if only a meaningless garbled character flashed across the screen.

In the end, however, he withdrew his hand.

The internal circulation piping of PMCU-17 had completely severed and drained dry, and the phase-change coolant had leaked and evaporated long ago. The silicon-based transistors inside the remaining core packaging—or some computational medium he couldn't name yet—might be in a high-risk thermal-sensitive state.

Without a reliable thermal conduction path, even low-voltage small currents could form uncontrollable hotspots at unknown short-circuit points, localized breakdown areas, or residual high-impedance channels, burning out the only remaining data core.

Later, spending nearly three years using mining station hydraulic pumps, industrial robot hydraulic servo valves, and silicon-based synthetic fluid, he pieced together an ugly and heavy external cooling bypass.

When that rig operated, it was terribly noisy.

Yet it truly built a firewall for the first emergency diagnostic layer wake-up of PMCU-17.

And it was precisely that external cooling bypass that allowed Jiang Lin to read a few lines of system feedback in the winter of the twenty-third year.

[PMCU-17 Emergency Diagnostic Layer Activated]

[Core Logic Unit Access: Denied]

[Local Log Offline Cache: Sector Partially Readable]

[Environmental Thermal Cycle System: Severely Failed]

[Internal Maintenance Bus Status: Degraded Operation]

[Main Array Optical Communication Link: Physically Lost]

[Last Network Sync Time: November 3, 2096, 04:17:22]

Therefore, the importance of cooling required no reminder from Artemis I.

He had long since learned this lesson the hard way.

The launch flowchart of Artemis I merely pushed another matter right in front of his eyes.

The external cooling bypass he pieced together previously was merely a temporary life-saving device.

Its purpose was to keep the PMCU-17 wreckage from burning out under low-voltage diagnostic conditions.

The severed multilayer high-pressure cooling lines on the original PMCU-17 wreckage pointed to something far larger.

That was a severed remnant of the outer ring thermal management network belonging to the TM-7 Regional Underground Calculation Array.

Jiang Lin flipped to another page of notes and rewrote the keywords he had previously extracted from the PMCU-17 cache.

[THERMAL_PHASE_BYPASS]

[RING-3 COOLANT STATE]

[CAPILLARY PRESSURE DROP]

[OUTER LOOP OVERLOAD]

[PHASE-CHANGE ROUTING FAILURE]

Phase-change cooling bypass.

Ring-3 coolant state.

Capillary pressure drop.

Outer loop overload.

Phase-change routing failure.

These terms all pointed to something.

They pointed to a far more complex thermal management system.

If TM-7 was indeed a regional underground calculation array, its computing power consumption could not possibly be low.

Deeply buried underground, its heat dissipation environment was even worse than on the surface.

If it was to continuously process data during the Star Tide Cataclysm from canopy stations, lunar far-side observation arrays, Earth-Moon peak shaving windows, and underground sealing protocols, it had to operate under extreme thermal loads long-term.

Ordinary air cooling was not enough.

Ordinary liquid cooling might not be enough either.

It needed a cooling network capable of protecting the core region even amidst localized pipe failures, pump group degradation, uneven thermal loads, coolant phase-change fluctuations, and outer loop bypass overloads.

A factory-grade high-pressure phase-change cooling network.

[part:gemini-3.5-flash-lite]

[Reverse Engineering Targets for the TM-7 Outer-Ring High-Pressure Phase-Change Cooling Network Connected to PMCU-17]

Five questions are listed below.

First, the original coolant type or phase-change temperature zone.

Second, the bypass trigger conditions.

Third, the outer-ring pipeline redundancy topology.

Fourth, the coupling method between the core heat source area and the outer-ring heat dissipation area.

Fifth, whether there are records of overheat protection, frequency reduction, isolation, or frozen state in the logs upon failure.

After writing the fifth item, he stopped.

Frozen state.

Once this word was connected to the low-confidence records in the PMCU-17 offline cache, its weight changed.

For a large-scale underground computing array that suffered a collapse of its outer-ring cooling system during a catastrophe, the most reasonable strategy was not necessarily a complete shutdown.

It might layer its load reduction.

Shut down peripheral services.

Cut off non-critical links.

Freeze high-heat zones.

Press the most core indexes, protocols, and civilization backup directories into a low-power state.

Wait for a maintenance window that nobody could guarantee.

But this was merely an inference.

The main array optical communication links of PMCU-17 had long been physically lost.

He had no real-time signals in his hands.

No handshakes from the TM-7 inner ring.

No direct evidence proving that the underground was still operating.

All he possessed was a state fragment in the local offline cache with low confidence, broken context, abnormal timestamps, and partial verification passed.

Jiang Lin copied that record out and gave it a more accurate name.

[INNER_RING_FREEZE_KEEPALIVE_RECORD]

Mark them item by item below.

[Source: PMCU-17 local offline cache, non-real-time link.]

[Explanation: Before the physical loss of the main array optical communication links, the outer-ring service layer suspectedly recorded the inner-ring frozen-state keep-alive tag.]

[Risk: Cache misalignment, clock drift, status code misread, or outer-ring failure artifacts cannot be ruled out.]

[Conclusion: Does not prove that the TM-7 inner ring is currently still running; it only serves as a secondary expedition trigger condition.]

The formulation could only be considered accurate up to this point.

Whether the underground was still alive was something he could not write.

The underground had once left behind a non-zeroed engineering status record.

This was already enough.

Jiang Lin added a line of red notes beside it.

[Prohibit attempting to awaken the TM-7 inner ring.]

[Prohibit writing to unknown main array links.]

[PMCU-17 is restricted to offline diagnostics and read-only cache review.]

[Read-only, isolate, bring back evidence.]

When the pen tip landed on the last word, the news feed on the right side of the computer refreshed with a message.

Over the past few days, the most prominent news in the global aerospace circle had all been about the same name.

Artemis I.

Artemis I.

The first giant rocket of NASA's new generation return-to-the-Moon program, the Space Launch System, carrying the Orion spacecraft, was about to make its first uncrewed circumlunar flight attempt at the Kennedy Space Center.

News headlines rolled by one after another.

[NASA prepares for historic Artemis I launch]

[SLS rocket on pad 39B]

[Return to the Moon]

[First step toward sustained lunar presence]

Humanity returns to the Moon.

Sustained lunar presence.

Deep space exploration.

Mars outpost.

These words were very romantic in the eyes of many people.

Rockets.

The Moon.

Astronauts.

The blue Earth rises from the grayish-white lunar surface.

Jiang Lin looked at the title with a calm expression.

What he was looking at was the launch pad.

Cryogenic propellants, liquid hydrogen, liquid oxygen, engine pre-cooling, ground systems, countdown procedures, and window management.

And how a massive engineering system simultaneously maintained correctness across countless tiny links.

Humanity's journey to the Moon never relied on a mere sentence of flying over there.

It was first and foremost a series of engineering accounts that did not allow for lies.

For a rocket to leave the launch pad, the ground, rocket body, fuel, valves, pipelines, sensors, software, weather, window, and personnel procedures must all meet the conditions within the same period of time.

This matter was very close to proving.

But it was messier than proving.

If there was an error in a proof, one could stop, rewrite, introduce a lemma, and add a boundary description.

Rockets would not do that.

When the countdown reached that point, the window was right there.

If the cooling was not completed, it was simply not completed.

If hydrogen leaked, it was simply a leak.

If the engine temperature did not enter the range, it simply could not ignite.

Once the launch window was exhausted, it could only stop.

The universe does not accept close enough.

The Moon does not accept close enough either.

Jiang Lin picked up his pen again and wrote another line below "cooling".

[The first bottleneck of high-energy systems is often not in computation, not in propulsion, but in heat.]

After writing this sentence, he paused for a moment.

This sentence pointed in three directions simultaneously.

Artemis I.

TM-7.

Computing centers in reality.

The predicament of modern GPU clusters was also not purely a shortage of cards.

Cards could be bought.

Racks could be piled up.

Models could be split up.

If the heat could not be dissipated, all computing power was just a bookkeeping number.

If the power per unit rack could not go up, the computing power density could not go up.

If the temperature difference of hot spots could not be suppressed, long-term full load would be unstable.

If the energy consumption of the cooling system could not be brought down, the computing center would be dragged to death by electricity bills and heat dissipation combined.

The walls of data centers were often heat walls.

Jiang Lin lowered his head to look at the keywords of the PMCU-17 cooling bypass.

If future humans had once developed a mature high-pressure phase-change cooling network on an underground regional computing array like TM-7, then what it could change first in reality might not be rockets, nor necessarily lunar bases.

It was more likely to be large-scale computing centers.

GPU clusters.

High-density racks.

Long-term full-load training.

Low PUE.

Higher computing power per unit volume.

Lower cooling energy consumption.

More stable heat flow management.

This thought popped up very quickly.

Jiang Lin did not run along with it.

There was no technology yet.

There were only wrecks.

A few incomplete words.

Ablated pipeline fractures.

Several incomplete status records in the PMCU-17 cache.

From the Wasteland relics to real-chip cooling architectures, there were materials, fluids, pressure, sealing, pumps, control, manufacturing processes, reliability verification, and engineering safety in between.

Any single link was enough to let a beautiful idea die on the test bench.

Jiang Lin wrote on the paper:

[Reality Application Direction: Large-scale GPU Cluster Thermal Management.]

Three words were added behind:

[Temporarily Sealed.]

Now, for the eleventh Wasteland mission, it was not yet time to develop chip cooling technology.

There was only one priority: to confirm what function that original-factory-grade high-pressure phase-change cooling network on the PMCU-17 wreckage actually undertook in the TM-7 outer-ring thermal management system.

What was its coolant?

How did it shunt?

How did it trigger the bypass?

How did it protect the inner ring when the outer ring was overloaded?

How did it fail?

After failure, what was left behind?

Only when these questions had answers would reality be qualified to take a step forward.

Jiang Lin flipped the paper to the next page.

Written in the header:

[Long-term Lunar Construction and Star-Tide Preparation Project: Reality Trigger Notes]

This title was too broad.

Based on the evidence he currently held, it could not be regarded as a real plan.

Therefore, he first wrote two sets of parentheses below the title.

[Used only as a Wasteland evidence retrofitting framework]

[Must not be used as reality conclusions]

Only then did he begin to break it down.

First, why is the Moon important?

Romance, flags, and propaganda slogans all came afterward.

On certain scales, the Moon was the second observation base easiest for Earth civilization to establish, a deep-space early-warning node, an electromagnetic silence observation platform facing away from Earth, a low-gravity engineering test field, and an external redundancy in the catastrophe era.

If the Dark Dust String came from the direction of the outer heliosphere, surface observations on Earth would always be passive.

Earth was enveloped by the atmosphere, disturbed by weather, blocked by day and night, and cut up by political boundaries.

The lunar farside observation array repeatedly appearing in the Wasteland cache could not just be a coincidence.

It was the outer eye.

Second, where lies the difficulty in long-term lunar construction?

Sending people up was only the first step.

Keeping equipment alive for the long term was the starting point of long-term construction.

Thermal control, power supply, dust, radiation, communication latency, thermal cycle fatigue, seal aging, material behavior under low gravity, and the continuous erosion of mechanical joints, heat sinks, optical windows, and sealing surfaces by lunar dust.

When writing about lunar dust, Jiang Lin tapped the paper surface with the tail of his pen.

The Wasteland red sand was already annoying enough.

Lunar dust was finer, sharper, drier, and carried static electricity, presenting a long-term abrasive erosion.

If future humans really wanted to build a long-term observation array on the Moon, what collapsed first might not be the computing system; sealing rings, heat sinks, connectors, bearings, and solar panel surfaces might all fail to hold out first.

Third, the underlying logic of the star-tide preparation project.

Jiang Lin wrote very slowly.

[Discovering anomalies in the outer heliosphere does not mean Earth takes immediate action.]

[The early stage must be observation, debate, misjudgment, and review.]

[Only then will there be preparation projects.]

[First stage of preparation projects: Confirmation.]

Confirming the particle size spectrum.

Confirming the flux.

Confirming the direction.

Confirming the relative velocity.

Confirming the heliosphere boundary compression response.

Confirming the high-risk windows of the Earth-Moon system.

Confirming whether the particulate impact anomalies of existing spacecraft are increasing.

Confirming whether external particle flow characteristics exist in the background noise of solar activity.

Fourth, why cut in from long-term lunar construction?

Because Earth is too close.

When high-risk particle flows enter the Earth-Moon system, many peak-shaving strategies will already be too late.

The eyes and part of the scheduling capability must be placed outside of Earth.

The Moon is the closest external fulcrum.

By the time Jiang Lin wrote up to here, the paper surface was already densely packed.

He put down his pen and moved his wrist.

The night outside the window grew deeper. The lights of Beijing leaked in through the cracks of the curtains, reflected into a thin line by the metal ruler at the corner of the table.

He looked toward the computer.

The news feed was still rolling.

There were less than two days left until that globally watched launch attempt on August 29th.

Artemis I had not flown yet.

The discussion was already lively enough.

Countless people were looking forward to that orange-and-white giant rocket igniting and lifting off, looking forward to turning a new page in the story of humanity's return to the Moon.

Jiang Lin did not expect it to succeed, nor did he expect it to fail.

He was merely watching.

Watching how the most complex deep-space engineering system in the Real World pushed itself bit by bit to a critical state before the launch window.

This was useful to him.

In the Wasteland, TM-7 might also have been pushed to a critical state.

It was just that what was exhausted that time was the window of the entire human civilization.

August 28th, morning.

Jiang Lin did not go to the mathematics center, nor did he go to the Low Entropy Workshop.

He locked himself in his personal workspace and began to repack materials according to the eleventh Wasteland target.

On the desktop, the floor, and beside the wall, a row of boxes was neatly placed.

Each box was affixed with a white label.

[Expedition Platform Spare Parts]

[Underground Detection]

[Low Temperature and Thermal Inertia]

[Communication Relay]

[PMCU-17 Interface]

[Cooling Bypass Reverse Engineering]

[Life Support]

[Risk Isolation]

[Paper Materials]

He was not in a hurry to pack.

Reviewing them one by one.

Deployable relay nodes, six pieces.

Each was only the size of a palm, with a rough shell lacking the exquisiteness of industrial products.

But they could survive in the Wasteland.

Low power consumption.

Weak communication.

Timed wake-up.

Link-break self-storage.

The shell used relatively reliable materials he could procure in reality, and all critical joints were re-treated with dust-proofing.

Jiang Lin numbered them, arranging them from R-01 all the way to R-06.

Beside them were disposable environmental recording boxes.

Four pieces.

Larger in volume and heavier in weight.

They were not responsible for communication.

They were to plunge into unknown areas and record silently.

Jiang Lin stuffed an independent memory card into each and conducted three write tests.

Temperature, humidity, barometric pressure, magnetic field, and vibration.

Once every ten seconds.

If dragged back, the curves inside would be the initial testimony of the underground darkness.

If they could not be dragged back, there was no help for it.

Some equipment was born to be sacrificed.

The mechanical winch was placed at the very inside.

This thing was heavy and clumsy.

But Jiang Lin did not trust purely electric solutions.

In the Wasteland, electric motors would break, control boards would burn out, and batteries would drop in voltage.

Once a jam occurred at the edge of the collapse, the most reliable things were still gears, ratchets, steel shafts, hand cranks, and audible mechanical meshing sounds.

Jiang Lin checked the ratchet.

Click.

Click.

Click.

The sound was clean.

He checked the aramid rope again.

Two rolls.

Color markings were made at fixed intervals.

Red for ten meters.

Yellow for five meters.

Black for one meter.

When visibility was insufficient or communication was interrupted, color markings were more reliable than electronic readings.

In the low-temperature and thermal-inertia box, the things were more fragmented.

Thermocouples.

Infrared temperature measurement modules.

Thermal insulation felt.

Small temperature loggers.

Phase-change material sample blocks.

Copper sheets.

Aluminum sheets.

Stainless steel thin sheets.

Several pieces of composite insulation materials of different thicknesses.

These things were not high-end in themselves.

But Jiang Lin needed them to answer a question in the Wasteland:

Had that underground thing still left behind a measurable thermal-state relic?

Even if the exchange was extremely slow.

Even if it was only a little bit more than the strata background.

As long as it was not absolutely silent, the long-term curve would reveal traces.

Heat cannot fool people.

Jiang Lin squatted on the ground and placed a temperature logger into the palm of his hand.

It was too small.

So small that it was hard to associate it with a massive underground array like TM-7.

But what spoke first in the ruins of civilization might not be huge machines, but perhaps a sensor stuck in the corner of a wall, silently recording for thousands of hours.

3:17 PM.

Jiang Lin opened the PMCU-17 cache index table.

The automatic alignment results left over from the previous round of MPS - Agent α were copied by him into the real hard drive.

The file name was very ordinary.

[pmcu17_cache_reindex_v4.log]

So ordinary that if it was not opened, no one would know what was hidden inside.

Jiang Lin opened it.

Lines of timestamps, incomplete key-values, CRC markers, confidence levels, and context fragments popped up on the screen.

Most of them were gray.

Invalid.

Duplicate.

Unalignable.

Insufficient context.

Partially incomplete.

Several rows among them were marked dark yellow by MPS - Agent α.

[TM7_OUTER_SERVICE_LAYER]

[THERMAL_PHASE_BYPASS]

[RING-3 COOLANT STATE]

[SEVENTH_SKYSCREEN_NODE]

[LUNAR_FARSIDE_OBS_LINK]

[FDSO-2076 A]

Jiang Lin did not rush to flip down.

He had seen these words too many times.

What truly made him adjust the eleventh Wasteland target was that low-confidence match at the very bottom of the index table.

[INNER_RING_FREEZE_KEEPALIVE_RECORD]

Confidence: 0.127.

Context: Broken.

Timestamp: Abnormal.

Verification: Partially passed.

Source: PMCU-17 outer-ring maintenance cache, suspected non-continuous write.

Note: Cache misalignment, clock drift, status code misinterpretation, or outer ring fault artifacts cannot be ruled out.

Jiang Lin stared at that line.

Inner ring.

Frozen state.

Keep-alive record.

Zero point one two seven.

If this were a competition problem, this confidence level would be meaningless.

If this were a paper, this evidence would not be enough for publication.

If this were a Real World engineering review, it wouldn't even be enough to initiate a project.

But what is faced here is the Wasteland.

It is a regional-level computing array left behind by a destroyed human civilization thirty meters, fifty meters, or even deeper underground.

Twelve point seven percentage points are enough for him to prepare an expedition once again.

Jiang Lin copied that line and pasted it into a new note.

[Question: Within the Northeast Seventy-Three collapse zone, are there still remaining thermal, electrical, magnetic, or low-frequency vibration evidence matching the "inner ring frozen state keep-alive record" ?]

[Objective: Do not attempt to wake up the TM-7 inner ring, do not attempt to take over, do not attempt to write.]

[Confirmation only: Whether abnormal heat sources, periodic low-frequency vibrations, residual cooling paths, inner ring power supply traces, or sealed boundary remnants exist.]

He drew a heavy line under "Do not attempt to wake up the TM-7 inner ring".

This was the bottom line.

PMCU-17 was already dangerous enough.

If the TM-7 inner ring had truly preserved a minimum frozen state in the past, it should not be treated as an old computer waiting to be repaired.

It was a regional-level engineering system left behind from the cataclysm era.

Jiang Lin did not know its permission structure.

He did not know its security logic.

He did not know if it had ever maintained residual connections with certain runaway external nodes.

He did not know if it carried civilization-spark-level data backup protocols.

Even less did he know whether behind the so-called keep-alive record lay secure sealing or an unfinished automated process from the cataclysm.

Facing an unknown system, the first principle always comes before curiosity.

It was isolation.

On the night of August 28.

Jiang Lin did not continue organizing supplies.

He tuned the news stream to the aerospace channel.

The live broadcast screen of the Kennedy Space Center had already spread out.

The SLS rocket stood upright on Launch Complex 39B.

The huge orange core stage stood silently under the night and searchlights, with white solid rocket boosters separated on both sides, and the steel structure of the launch tower filled half the screen.

In the barrage and comment section, it was all waiting.

Some were excited.

Some were nostalgic.

Some said that after Apollo, humanity was finally going to seriously head toward the moon once again.

Others questioned that SLS was too expensive, too slow, and too cumbersome, being the legacy of a political project.

Jiang Lin did not participate in any discussions.

He paused the screen on the panoramic view of the launch pad and began looking at things ignored by ordinary viewers.

Service tower.

Umbilical arm.

Liquid hydrogen and liquid oxygen fueling.

Ground cables.

Discharge system.

Sensor cables.

Personnel evacuation routes.

Weather constraints.

Launch window.

The rocket was the most conspicuous.

But the rocket was only the part of the entire system exposed above ground.

What was truly massive was the ground support network behind it.

Without that network, the rocket was just an expensive metal pillar.

The same applied to long-term lunar construction.

The news would film astronauts.

Film lunar rovers.

Film base concept art.

Film human footprints.

What truly determined whether a lunar base could survive long-term were things much harder to capture into passionate footage.

Thermal control.

Power.

Maintenance.

Spare parts.

Dust management.

Closed-loop life support.

Communication redundancy.

Fault isolation.

Long-term material degradation.

And every tedious process that no one was willing to write into a promotional video.

Jiang Lin named this page of notes—[Lunar_Long_Building_Preliminary]

Preliminary Long-Term Lunar Construction.

This note could not enter real plans for the time being.

It was merely a Wasteland reverse-derivation framework.

It did not need to be seen by anyone.

At least not for now.

On August 29.

Beijing, a little past 8 PM.

Florida, United States, Kennedy Space Center, early morning.

The world's attention was focused on that rocket.

Jiang Lin sat at the desk, with no snacks or drinks in front of him.

Only a cup of water, a notebook, and three portable hard drives.

On the left side of the screen was the NASA live broadcast.

On the right was his own supply checklist.

Before the launch window opened, the news stream updated faster and faster.

Fueling, inspection, countdown, weather, propellant, engines.

Jiang Lin's gaze stopped on an English update.

Engine bleed.

Engine bleed pre-cooling.

The RS-25 engines needed to be adjusted to the proper temperature range before ignition.

Liquid hydrogen system.

Cryogenic.

Process.

Window.

Seeing these words, he grasped the truly valuable part of this news.

Whether the rocket could fly came second.

If it did not fly, what would be the reason it didn't fly.

Not long after, the atmosphere in the live broadcast changed.

Countdown holding.

Engineers continued to evaluate.

The anchor's speaking speed remained steady, with a little more anticipation in their tone.

Jiang Lin did not move.

He watched the screen.

The temperature reading of engine number 3 never entered the range it should have entered.

Three were normal.

One was just a little off.

Being broken was actually simpler.

If it broke, it was broken.

What was truly troublesome was being just a little off.

Being just a little off meant engineers had to judge: was the engine truly not cooled down enough, or was the sensor lying; were there undrained anomalies in the pipelines, or deviations in the reading link; should they continue trying, or admit they didn't understand this engine's body temperature.

Jiang Lin stared at the screen.

He thought of [RING-3 COOLANT STATE] in the PMCU-17 cache.

And he also thought of those cooling state fragments that could never be completely aligned.

In some longer and more cruel window, would the outer ring of future humanity's TM-7 have encountered similar problems?

Ring 3 cooling state was a little off.

Outer ring bypass was a little off.

Capillary pressure drop was a little off.

Inner ring frozen state was a little off.

The cataclysm era would not give engineers a live television broadcast screen, nor would it give them a countdown watched by the whole world together.

It would only leave a line of status code in the logs.

[PHASE-CHANGE ROUTING FAILURE]

Phase-change routing failure.

Around 8:34 AM Eastern Time.

The launch attempt was called off.

NASA's update quickly appeared.

The Space Launch System rocket and Orion spacecraft remained in a safe and stable state.

The launch team was still evaluating why the bleed test at the bottom of the RS-25 engines failed to adjust the engines to the temperature range required for launch.

The two-hour launch window was exhausted.

Jiang Lin looked at those lines of text and did not speak for a long time.

The engines did not explode.

The rocket did not fall over.

The spacecraft did not go out of control.

No disaster occurred in the world.

Everything could even be called safe, stable, and professional.

But it simply could not fly.

Because a cooling process failed to meet standards.

Because the window was exhausted.

Because deep space engineering does not accept "a little off".

Jiang Lin suddenly felt that this was more useful than a smooth liftoff.

If Artemis I flew into the sky smoothly, the news would give people a misconception: humanity was ready to return to the moon.

This call-off, instead, peeled back a layer of truth.

To return to the moon, humanity must first relearn how to operate a complete set of deep space engineering systems seamlessly within the temperature, pressure, leaks, valves, processes, sensors, and time windows of the Real World.

The first lesson of long-term construction was not necessarily ignition and liftoff.

Stop when conditions are not met.

This was not irony.

On the contrary, it showed that the system still retained its final honesty.

It did not disguise its inability to read engine number 3's body temperature as being able to fly.

It stopped within the window.

Jiang Lin picked up his pen and wrote in his notebook:

[Artemis I, 2022-08-29, first attempt called off.]

[Surface cause: Engine cooling / bleed test failed to reach the temperature range required for launch, window exhausted.]

[Related processes: cryogenic propellants, liquid hydrogen system, engine pre-cooling, launch window.]

[Deep revelation: The core of lunar engineering lies in long-term, stable, and repeatable passage through ground, orbit, and lunar surface full-link processes.]

He paused for a moment and wrote again:

[If the Star Tide preparatory engineering existed, early failures would not manifest as grand disasters.]

[It would manifest as: substandard cooling, link timeouts, sensor drift, missed windows, redundancy failures, and maintenance cycles being forced to shorten.]

[Civilization begins to frequently miss windows first.]

After this line fell, Jiang Lin fell silent for a few seconds.

Outside the window was the night of Beijing.

Quiet, warm, and breezy.

In the distance, freshers were still chatting and laughing.

No alarms sounded in the Real World.

No red sky.

No Dark Dust String.

No Seventh Canopy Station requesting peak-shaving scheduling.

No lunar back-side link timeout.

No Mars relay damage probability of 98%.

It was just an ordinary August 29.

A rocket was cautiously left on the launch pad because its engine cooling process failed to meet standards.

To the vast majority of people, this matter was just an aerospace news story.

To Jiang Lin, however, it was knuckles knocking on glass.

The movement was very small, yet clear enough.

He closed the news window.

He opened the eleventh Wasteland task list and dragged [Northeast Seventy-Three Second Expedition] from third place to first place.

Then he created a new task group.

[Star Tide Preparatory Engineering Reverse-Derivation]

Below were only three items.

1. Continue extracting logs related to the lunar back-side observation array, canopy station, and Earth-Moon peak-shaving windows from the PMCU-17 cache.

2. Using the Artemis I call-off as a Real World sample, build a comparison table of deep space engineering window failures, process failures, and cryogenic system failures.

3. Confirm whether thermal, electrical, magnetic, or low-frequency vibration evidence matching the inner ring frozen state keep-alive record still remains within the Northeast Seventy-Three collapse zone; if so, prioritize searching for the civilization-spark-level compressed backup protocol index.

Subsequently, he created a second task group.

[Reversal of the TM-7 Outer Ring High-Pressure Phase-Change Cooling Network Connected to PMCU-17]

Listed more specifically below.

1. Confirm the outer ring cooling pipeline topology without directly touching the unknown interfaces of the inner ring.

2. Collect residual crystallization samples from the fractured ends of the exposed cooling pipelines in the collapse zone to determine the OEM coolant type and phase-change temperature zone.

3. Find the actual pipeline nodes corresponding to THERMAL_PHASE_BYPASS.

4. Compare with the RING-3 COOLANT STATE logs to reconstruct bypass trigger conditions.

5. Retrieve whether records related to overheat protection, load shedding, frozen states, and core area isolation exist.

6. Evaluate the migratable boundaries of this architecture in Real World high heat flux density computing systems.

After writing the sixth item, Jiang Lin paused again.

"Migratable boundaries" was more accurate than "application prospects".

He did not need to fantasize about any disruptive heat dissipation technology right now.

He only needed to find sufficiently solid principles, structures, materials, and failure records in the Wasteland.

If that set of things truly existed, if it could be dismantled, if it could be shrunk, modified, and verified under Real World materials and manufacturing processes, it could certainly change GPU clusters, large-scale computing centers, and even the energy consumption curve of the entire high-density computing infrastructure in the future.

But that was a story for another day.

Right now, it was still just pipelines on the remains, words in the cache, and an expedition target.

Jiang Lin wrote a red note at the very bottom of the task group.

[Prohibit attempting to wake up the TM-7 inner ring.]

[Prohibit writing to unknown main array links.]

[PMCU-17 is restricted to offline diagnostics and read-only cache review.]

[Read-only, isolate, bring back evidence.]

These four lines of words were more important than any grand rhetoric.

He stood up and pushed the pre-sorted boxes one by one to the wall.

Expedition platform box.

Underground exploration box.

Cryogenic and thermal inertia box.

Communication relay box.

PMCU interface box.

Cooling bypass reverse-engineering box.

Life support box.

Risk isolation box.

Paper data box.

Nine boxes, neat and orderly, lined up silently under the wall.

Every box had a number.

Every number had a checklist.

Every checklist corresponded to a type of failure that could occur in the Wasteland.

Jiang Lin stood in front of them, thinking of that NASA rocket still left on the launch pad.

It had not failed.

At least not counted as failure in the ordinary sense.

It simply did not force a flight when conditions were not met.

This was the final dignity of an engineering system.

The Wasteland would not give him a launch director, nor would anyone press "hold" for him at T-minus 40 minutes.

Nor would there be anyone to safely and stably leave all equipment on the launch pad after a two-hour launch window ended.

He only had himself.

Therefore, he had to write down all the lines that must not be crossed in advance for his future self standing at the edge of the underground fissure.

Ten minutes past six in the evening.

Jiang Lin reopened the PMCU-17 cache index table and stared at that low-confidence match for the last time.

[INNER_RING_FREEZE_KEEPALIVE_RECORD]

Confidence: 0.127.

He did not raise the confidence.

He did not beautify the conclusion.

He did not write it as discovering that TM-7 was still running.

He only added a sentence beside it.

[Non-zero, enough for an expedition.]

Then he turned off the computer.

The room darkened.

The nine boxes by the wall lay in the night, waiting to be sent into the Wasteland.

Outside the window, Beijing was still quiet.

No one in Tsinghua University campus knew that on this ordinary night, an aerospace news story from across the ocean was pressed into another set of cataclysm ledgers by an eighteen-year-old youth.

The moon, canopy station, underground array, Star Tide, cooling bypass, computing center.

Engine number 3's body temperature did not drop to where it should have gone, and the rocket remained on the launch pad.

Deep in the Wasteland, that regional-level underground computing array that had been silent for many years might have once left behind a little heartbeat that did not return to zero.

In the dark, the last thing glowing was the status light of a small recording box on the desk.

Blinking on and off.

As quiet as a simulation keep-alive signal.