Skip ahead thousands of miles by plane and by train, and I am ascending the side of Aso volcano, bracing myself against the furious winds that whip me with a sulphur-laced mist. A Swiss man and French woman, whom I met on the train and Aso visitor center, respectively, hike alongside. We take shelter from the wind in concrete bunkers lining the final ascent, meant to provide temporary cover in the event of an eruption. The mouth of Aso contains not lava but a chemical mixture in cool shades of blue and green, steaming into the cold air.

We take a train from the volcano’s base back to town. The rice fields mirror the scenery; it begins to rain and the reflections of trees and clouds dissolve. The train zips in and out of tunnels, like a movie played at the wrong framerate: swaying pines; black; a small farm nestled in a lush valley; black; a town of patchwork fields and shops; black. As we part ways, the French girl quotes, “I was not alone today; it was a good day.”

That night, I visit an onsen (hot spring) for the first time. I sit beneath a small waterfall, the arhythmic pounding of the hot water on the back of my head and neck lulls me into a half sleep. The subtle scent of sulphur in the air, from the volcanic activity that heats the water, reminds me of the morning’s chilly climb to the rim of the caldera. It was a day of cold fire and hot water.

—

I am hiking into the clouds, scrambling over rocks toward one of the peaks of Yufu-dake, near Beppu. I pass a man, also on his way up, and practice my Japanese. He asks where I am from, and I tell him California. “Ah, California,” he responds in English with an odd cadence, “such a lovely place.”

“Yes, it really is.”

“No,” he insists, “Such a lovely place. Such a lovely place.”

Ah! You can check out, but you can never leave…

I descend a different way than I came, down to the town of Yufuin. In the town, there is a lake; beside the lake, there is a small public onsen. I drop 200 yen into the honesty box at the entrance and slide aside the unlocked wooden door. There is no one else inside, so I have the hot spring to myself. I strip down and wash my body before sliding into the hot water: aching muscles sigh with relief.

—

I meet a Texan girl and an English girl on the train and we resolve to explore the island of Miyajima—famous for the huge tori (gates) placed just offshore—together. On the ferry ride over, the bright red gates glow in the sunlight. Soon, we are hiking down from the peak of Misen-san. We stop at a stream, peel off our shoes and socks, and soak our feet in the frigid mountain runoff. Soon, a fawn is eating the leaves off a branch of Japanese Maple that I hold out for it. Soon, we are drinking at an izakaya (pub) in Hiroshima. Too soon, we say sayonara.

—

I am eating strawberries and walking through persimmon groves; this historic Yama-no-be-no-michi path winds among temples and shrines on the outskirts of Nara. The sound of trickling water permeates the air in a small town, criss-crossed by canals. Farmers burn the remains of last-year’s crops, returning nutrients to the soil and sending columns of white smoke into the sky.

—

I am in the mountains northeast of Kyoto and my bare feet grip stones while water rushes over them and tumbles down a waterfall. My hands grip a metal chain stretched across it: my life-line. My shoes are tied to the waist-strap of my backpack. I try to move quickly so that I make it to the other side before my toes go numb and I lose my foothold, mindful not to rush lest I end up like the two people who died on the Yatsubuchi-no-taki hike this year. I meet a Japanese group, friends who met over SMS, and I temporarily hike with them, letting them go first for ideas on how to navigate the terrain.

—

I am standing on the balcony of my private room at the Dogo Onsen, fresh from a soak in the hot spring and sporting a yukata. An attendant raps lightly on the outside of the rice-paper door and enters, bearing tea and multi-colored dango. I sip the tea while watching people pass on the street below. They occasionally stop to take pictures of the gaijin on the balcony.

—

I am at the Ramen Museum in Shin-Yokohama, where they have recreated 1958 Tokyo: the hey-day of ramen. I move from stall to stall, trying recipes from various regions. I pass beneath old movie posters and nod to the faux policeman who is drawing Astro Boy on a blackboard. In the main courtyard, a family plays games with a hula hoop and some older boys try their luck at an air-gun shooting gallery. I gulp down the last of a miso ramen while two young waitresses peek around the corner and giggle.

—

I am at the banquet for a supercomputing conference in Tsukuba. A man is playing a tiny drumset, amplified to sound like the real thing, accompanied by an electric guitar. There is something quintessentially Japanese about a man in business casual, playing American tunes on a miniature musical instrument. I decide to stay for one last song before heading back to my hotel. The singer leads into the chorus: “Welcome to the Hotel California. Such a lovely place, such a lovely place…”

Indeed. It is time to go home.

Title: Community Epidemic Detection using Time-Correlated Anomalies [pdf]

Published: International Symposium on Recent Advances in Intrusion Detection (RAID), 2010.

Consider a set of instances of an application, which we call a community. Two examples of communities are all the mail servers in an organization or all the browsers on a cluster of workstations. Assume some subset of these instances, or clients, are compromised and are running malicious code. The initial breach (or breaches) went undetected and the existence of the exploit is unknown, so the malicious code may continue running indefinitely, perhaps quietly stealing computing resources (as in a zombie network), spoofing content, denying service, etc. We present a method for detecting such situations by using properties of the aggregate behavior of the community to reliably identify when a subset of the community is not behaving properly.

We describe an implementation of an epidemic detector, called Syzygy, that applies two main insights:

1. Even if a single noisy model cannot reliably judge the health of a client, we can reduce the noise by averaging the judgements of many independent models and
2. Epidemics exhibit time-correlated behavior that is impossible to detect on a single client.

Our method effectively leverages the statistical properties of a large community to turn noisy models into reliable community detectors and uses the temporal properties of an epidemic as a means for better detecting it.

Syzygy monitors each client’s behavior and reports anomaly scores, which quantify the divergence of recent behavior from the model. For example, a client whose recent response times are unusually high may report a score that is above average (anomalous). Syzygy then computes the numerical average of all clients’ scores and checks whether this community score exceeds a threshold. By doing these computations properly, we can make strong theoretical guarantees about our ability to overcome model noise and detect epidemics. Intuitively, we expect anomalies on individual clients in a large community to be common, but we do not expect anomaly scores from multiple clients to be strongly correlated in time, absent an epidemic. Throughout the paper—using math, deployments, and simulations—we show that, in a large community, even simple, noisy models are sufficient for reliable epidemic detection.

• 5/10: Microsoft Research, Redmond, WA
• 5/12: UC Berkeley, RAD Lab lunch, Berkeley, CA
• 5/14: IBM Research, Hawthorne, NY
• 5/17: CMU, Pittsburgh, PA
• 6/3: ICS, Conference Talk, Tsukuba, Japan [summary]
• 6/30: DSN, Conference Talk, Chicago, IL [summary]

Look for slides on my research page after the dates above.

Title: A Query Language for Understanding Component Interactions in Production Systems [pdf] [slides]

Published: International Conference on Supercomputing (ICS), 2010.

When something unexpected happens in a large production system—a program crashes, a node’s performance flags, a power supply overheats—administrators face several problems at once. First, they may be unable to describe the event any more accurately than the approximate time it occurred. Second, they must diagnose the problem using only the data that was recorded when the issue manifested (primarily log files); this data may be noisy and may not describe all components and their interactions. Third, the system may have many components (tens to thousands), and the administrators must identify which components and component interactions are likely to have been involved.

Consider the following example. Users notice that their jobs are failing more frequently. The typical process for a system administrator is to search the job logs to figure out what components were used by these jobs, scour the system logs from those components for any messages that might hint at a cause, and possibly expand the search to other related components based on their expert knowledge of the system. The key observation is that this is fundamentally a search problem—one for which the state-of-practice is primarily manual, tedious, and ad hoc—where the administrator asks, “What components and interactions are likely to be involved with these job failures?” The input to the search is the available measurements from instrumentation and a simple description of the behavior we wish to understand; the goal of the search is to identify the components and interactions that are likely to be involved.

In a paper to appear at ICS 2010, we present a method for using simple user specifications of when and where a problem manifested, together with existing instrumentation, to compute the components and interactions that are likely to be involved with the problem. Our method computes which system components statistically influence the behavior of other components and which components are statistically linked with the problem.

Our system, QI (pronounce ‘chee’), does not require modifications or perturbations to the system, access to source code, or even knowledge of all the components in the system or their dependencies on one another. Our assumptions are considerably weaker than most previous work and they reflect, in our experience, the reality faced by administrators when they must diagnose a problem. The answers QI provides are limited by these contraints: a passive, black-box technique can, at best, suggest the components and interactions that seem statistically most likely to be involved with a problem. The main advantage is that, because of the weak assumptions, such a system can leverage all of the information available. This is precisely what our method provides, and it does so in a way that is computationally efficient and applicable to a wide variety of systems.

In particular, we evaluate QI using nearly 1.22 billion lines of code from unmodified production systems: four supercomputers, two embedded systems, and a server cluster. On these data, we correctly answer a wide variety of exploratory and diagnostic questions about dynamic system behavior, usually in a couple of seconds.

Title: Using Correlated Surprise to Infer Shared Influence [pdf] [slides]

Published: International Conference on Dependable Systems and Networks (DSN), 2010.

During the DARPA Grand Challenge race in 2005, the autonomous vehicle named Stanley, Stanford’s entry, slowed down and swerved around an obstacle that was not actually there. Stanley did this several times over the course of the race, nearly causing it to be disqualified. Although Stanley went on to win the competition, the Stanford Racing Team was justifiably vexed: why had Stanley hallucinated these obstacles?

Using a hand-crafted dependency diagram—the golden ideal to which all previous work on dependency inference aspires—it took the designers of the system early two months to isolate the problem, which was originating from a buffer component shared by the laser sensors. This shared buffer was intermittently dropping measurements, causing Stanley to see stale, inconsistent data about the world around him, which sometimes meant seeing obstacles where there were none. Every other component of the system was behaving according to specification, which partially explains why the dependency diagram was so unhelpful: the source of the problem and its outward manifestation (swerving) were on opposite logical ends of the system but the dependency diagram advised looking at almost every component in between. The shared buffer was not even on the diagram.

In a paper to appear at DSN 2010, we introduce the idea of computing influence, a type of component interaction that is orthogonal to dependencies and allows us to capture implicit interactions among components and subsystems. For the Stanley swerving bug, our method not only infers an influence directly between the swerving behavior and misbehavior near the laser sensors, it also implicates an uninstrumented component shared by those lasers: the true cause of the problem. The Racing Team says the results of our analysis, which took only a few seconds to compute, would have saved them two months of debugging.

Computing the strength of shared influence between components is straightforward:

1. Represent the behavior of each component as a function of surprise over time, called an anomaly signal.
2. See how well these functions “line-up” using a standard technique called cross-correlation.
3. Summarize cross-correlations in a Structure-of-Influence Graph (SIG), where the edges indicate the strength and time-delay of the influence.

Our paper gives a mathematical foundation for influence, as described above, and evaluates it using both simulations of idealized systems and case studies with real systems, including Stanley, his successor (Junior), and the Thunderbird supercomputer.

We propose a method for identifying the sources of problems in complex production systems where, due to the prohibitive costs of instrumentation, the data available for analysis may be noisy or incomplete. In particular, we may not have complete knowledge of all components and their interactions. We define influences as a class of component interactions that includes direct communication and resource contention. Our method infers the influences among components in a system by looking for time-correlated divergence from models of individual component behavior. We summarize the strength and directionality of shared influences using a Structure-of-Influence Graph (SIG). This talk explains how to construct a SIG and use it to isolate performance bugs, and presents both simulations and an in-depth case study using data from two autonomous vehicles.

Meanwhile, I’ve been keeping busy. Here’s a recent sampling:

• The US Patent Office granted me a second patent. The first one, meanwhile, has already passed into obscurity.
• I got an alert detection paper [pdf] accepted to ICDM in Pisa, Italy, where I’ll be going in December to talk about it.
• I served on the program committee for a new workshop called WASL and am on the program committee for NAS ‘09.
• I applied to graduate… sort of. The requirements for a Master’s are a subset of those for my Ph.D., so I filled out the paperwork to pick up an extra degree. I already have an M.Eng. from MIT, so this one’s redundant and will be obsolete whenever I get my doctorate.
• I gave a few invited talks, most recently at a workshop at LACSS [slides].
• I agreed to serve on the CS Department’s faculty search committee.

I’m working on a system modeling paper for submission on Monday. Wish me luck!

I spend the two days before leaving for Black Rock City (BRC) at a shipyard in Berkeley, helping to construct the art car and to pack the camp’s supplies. The site is abuzz with the hiss of spray paint, the sizzle and crackle of welding, and the clangs of metal against metal. We work through the night and into the next day before finally mustering the troops and pointing our caravan toward Nevada.

BRC is a glow on the horizon as we pull toward it in the late evening. The city is still under construction; we have arrived early to set up. Fine particles blow up off the playa (ply-uh) and envelop us in a cloud of clay dust. I don my amber-tinted ski goggles, and the world is rendered in sepia tones—as though the memories here are destined to be treasured and extracted years later, weathered by time and wind and dust. A gentleman named Squirrel welcomes us. I step over a line in the sand and ring a bell; under the bright moon, I enter Burning Man.

Midnight on Sunday is the official start of Burning Man. Already, it has been three days since I had a shower. My hands are dry and filthy, layers of dust and bike grease and food coat them in a mottled white glove and outline my nails with black. I help make pancakes for the camp. That evening, I cook fajitas and then pitch my tent. As darkness falls, thumps of light and heat punctuate the flashing, glowing, musical hustle of preparations; they are huge, distant flame-throwers, launching fireballs into the air.

The first full evening of Burning Man is a Monday, and a full lunar eclipse. I begin exploring this surreal world: shots at the Tequila Shack, bad dancing penalized by fire at Dance Dance Immolation, jokes and songs in exchange for a mug of IP-fucking-A at the Carbofuckingnation Camp, building with magnetic blocks beneath a peaceful tent, pounding furiously on bongos as a carousel comes to life and animates a death-dance between a gorilla and a snake… As the eclipse begins, I bike to the Opulent Temple. There, I dance among the thumping techno, glowsticks, lasers, and dual jets of fire that periodically erupt from the DJ booth. One such flash burns an image in my mind: a beautiful woman, topless and bedecked with beaded decorations, arms and hair flailing wildly, her eyes closed. She is smiling. I dance for hours, moving from party to party, high on the energy of the city, as the shadow of the Earth consumes whole the once-brilliant moon.

Just then, when every eye is turned skyward, the Man begins to burn. The ceremonial burning is supposed to happen at the conclusion of the event on Saturday night; this is Monday, this is unplanned. Someone had torched it. Standing next to my bike, just outside the safety perimeter hastily arranged by the BRC Rangers, I watch pieces of the Man break off in flaming chunks and tumble down the sloped tent roof. The wooden effigy is fully engulfed in flames by the time water trucks and fire crews manage to tame the conflagration. The spectacle over, and I head toward home, but my attention is drawn to a cluster of red and blue lights. A shirtless man with face paint is being handcuffed and frisked, while half a dozen other officers supervise the proceedings and a K-9 team keeps the hippies at a distance. It was the arsonist, Paul Addis. I watch his arrest with the smoldering Man behind me and the red, eclipsed moon above.

The spectacles amass throughout the week. I slurp down ramen while watching a gorgeous moonrise, climb the steampunk tree, watch wraith-like kites drift in the sky like enormous white apparitions, visit the Thunderdome as people clamber over its geodesic shell and await the next battle, play with the bouncing glow-trees that left me giggling, and bike out to the fence-line that borders BRC. Pausing to rest at that edge between city and oblivion, I notice a serious-looking dust storm approaching. I cannot make it to my camp, but get as far as the Temple, a huge wooden structure that evokes thoughts of a pagoda. On the structure itself, stretching as high as people can reach, are messages scrawled in pens and markers. It is a temple of forgiveness and of loss. “Goodbye Mom, Dad, & Muriel,” reads one message. Another: “I ask for guidance…” Some are simple messages of joy (“I am alive!”) and others of hope (“Mom, let’s be friends again”). I wander around the Temple, reading these messages through my ski goggles as the storm completely whites-out the world beyond my arms’ reach. I cry. Picking up a black marker and bracing against the fierce winds, I add two inscriptions.

On the way home, a man hands me a plastic, glowing lightsaber. “Sundown at the Man,” he says and bikes away. Thousands of swords are distributed throughout the day. The evening proceeds predictably.

Midway through the week, I am surprised to discover myself sick with loneliness. It happens while I am dancing at the Deep End, watching the crazy costumes and funny people amuse each other. I return to camp and get all introspective and moody. I stand by the side of the road to watch the sunset. Just then, a man on a bike pulls up to me and says, “You need to get changed!” I glance down at my shorts and t-shirt. His wife pulls up next to him.

“This is really all I have,” I confess sheepishly.

The man stares at me for a long moment, brow furrowed. “Come with me.” And I do. He gives me a playa costume, and his wife gives me some jewelry. I return to camp looking ridiculous and absurd and wonderful. With that improbable and perfectly timed gesture, the strangers had changed my attitude. I am not lonely or out of place anymore; the camp and the citizens of BRC embrace me, and I become another comical gem in the dazzling all-night parties.

On my last day, like nearly every other day, I go to the Turkish-style steam baths. Sitting nude in a small, insulated geodesic dome with a dozen strangers, I sweat myself clean. My friend Sara begins to hum a tone, and this evolves until we are all chanting an improvised song. I close my eyes and listen, contributing notes where I can. There is no embarrassment, no self-conscious shame or blushing cheeks in that dark hut of singing naked strangers. No money has exchanged hands among its occupants. There are no debts or loans. We have all given each other gifts, and do so even now by sharing this spiritual moment. Afterward, I volunteer to help the camp prepare cleaned and boiled rags for use in the baths. My friends KB and Stephanie join me.

Suddenly, a man rushes up to the camp’s leader, and, for the first time, I hear a Burner invoke an authority figure. “There was a videographer,” he explains, “filming the camp. Should we notify a Ranger?” This struck me. I had seen the citizens of Black Rock City drive drunk and drink underage, commit public nudity and lewd acts, and violate so many drug laws I couldn’t begin to name them. But the only time anyone expressed genuine concern for the safety of their fellow citizens was when a man with a camera tried to capture them on film.

“It is amazing how important that privacy is to the culture here,” I mused. “We’re comfortable with our nudity and craziness because it’s only being shared with other Burners, who share alike. The camera is stealing that gift.”

Stephanie, a photographer, nods and describes the challenges involved with documenting a party; how do you prevent yourself from changing the events you wish to seize on film? I geek out and talk about Heisenberg and about the observer effect. It is a deep property of the universe that measurement may change the outcome. KB speculates, perhaps idly, that there must be some broader philosophical principle there. Before he is done speaking, I know the answer.

I understand why the gift of the playa costume so drastically altered my mood. Why the loneliness did not strike me until I stopped working on the camp and the art car. Why I felt compelled to share my strongest emotions with the Temple. Why the premature burn was so important and exciting, and why the ceremonial burn felt so artificial and sterile.

I nod and pick up another rag, pleased to have given this gift of my time. And then I share my epiphany, smiling at the simplicity of it.

“You can never just observe.” I squeeze water from the washcloth. “You must participate.”

[A man in a blue suit strides to the podium, the enormous ballroom is filled to overflowing with scientists and military officers. Cameras are focused on him from all sides, projecting his visage onto the screens behind him and into the many satellite viewing rooms throughout the hotel.]

Good morning. I’m going to talk to you today about the future: a vision of the future as seen by the DARPA Made-Up Technologies Office. The best of the best. DARPA’s DARPA. Rambo to STO’s Barney Fife.

Imagine a world in which soldiers cannot die. In which their armor adapts to new threats instantaneously, their weapons target flawlessly and inflict the desired damage, and their hair maintains its shine and bounce, even in the harshest of combat conditions. Imagine a world where a global information network is accessible at your fingertips, or even closer, like at your knuckles or wrists. Where you can detect enemies breathing behind concrete walls, clot and repair a bleeding femoral artery with a simple tourniquet, and where a universal replacement part can assume whatever shape or function you desire. A wrench becomes a hammer. Wings take dream.

We at the MUTO are imagining exactly that.

Soldiers must fight in extremes. In the snow dunes of the arctic, the sand drifts of the desert, deep beneath the ocean, on mountain peaks, and, someday, in outer space and in the center of our sun.

[Slide show displays the Sun. Speaker gestures meaningfully.]

Our opponents are smart, capable, well-trained, and fighting on their home turf. Some of them can yodel. Most of our soldiers can barely manage a passable Star Spangled Banner. Our Army Rangers train in the mountains of Georgia, while Afghani fighters are acclimated to altitudes tens of thousands, no, millions of feet higher. Geese can handle these altitudes, why can’t our warriors?

Our enemies have rockets launchers. Some of them have elephants. They may even have figured out how to put rocket launchers on elephants. You can’t prove they haven’t. And when they do, will you be able to say you did everything possible to prepare?

The work we do at MUTO represents not merely fundamentally unique technological achievements, but entirely new fields of research. A calculus of awesomeness, if you will. It revolutionizes not only urban combat, but warfare in its entirety. And also poetry.

Allow me to give you a moment for your brains to stop smoking.

[Stares wistfully into the distance.]

Now that you have some idea of the preponderance of cutting edge research that is discussed at length in our office, let me introduce the next speaker, who will frighten you with outrageously melodramatic nightmare scenarios, entice you with nonexistent but sexy technology, and ease you into a peaceful and meditative state with utopian vistas of the future. Your future.

But only if the money keeps flowing to DARPA. Thank you.

The talks were obviously not the main attraction of the conference, for me. Rather, I enjoyed walking around the exhibit hall and learning about the amazing projects already underway. I especially liked some of the simpler ones, like the sniper rifle equipped with a cross-wind detector, which would indicate where one should aim in order to compensate.

The project I was there to help present is called Vernier, which aims to leverage application communities to detect and control exploits. We had a live demo that showed Vernier successfully detecting, controlling, and recovering from a self-propagating worm as it spread through a community of twenty nodes.

It was strange seeing military officers, including a three-star general complete with military entourage, checking out the latest geeky wares. Then again… there, but for the funding from DARPA, go I.

“What is the weather supposed to be like?” I asked the woman who had just printed my wilderness permit.

“The same as always,” she said in a tone inexplicably laced with lamentation, “It never rains.”

I nodded, suspicious of her claim. My summer in New Mexico had taught me that there is always a storm in the afternoon, often with accompanying pyrotechnics. One of my few California hikes had been to the Lost Coast, where it rained with the constancy of Niagra Falls. Her words proved true, however, and I enjoyed the most mild and pleasant weather of any outdoor experience. No rain gear nor warm nighttime clothes were necessary.

With a full pack, I hiked the 8 miles to the furthest and largest of the ten lakes. I was pleased to learn that the women who were swimming in the lake when I arrived were camping elsewhere. I had the lake to myself. I pitched my tent in a prime spot and ambled cautiously down to the water wearing only my shorts. I waded in the clear, shallow water amongst the large tadpoles, stepping one at a time on the flat rocks which layered the bottom of the lake, less than a foot beneath the surface. The warm sun illuminated the submerged sand and made the water temperature pleasantly cool.

I climbed atop a boulder that decorated the small outcropping of land that I had designated my personal peninsula. I let the afternoon rays dry my legs and reflect off of my pale skin, blinding hawks and chipmunks, alike. Satisfied, I leapt down and returned to my tent to fetch my shoes. The deer was waiting for me.

It was a doe (a deer, a female deer), or a buck cruelly gypped of his male birthright. It froze at the sound of my approach and we stared at each other. She broke the gaze first, dropping her head down to graze. I wondered if the animal would understand human body language, and so I turned away in a feint of disinterest. Convinced I was no threat, the doe returned later that evening, braver this time, coming within a few feet of me. She appraised me and lingered with me for a few minutes on my peninsula as the sun began to disappear behind the pines.

I ate dinner on my boulder: hot chicken noodle soup in a bag. After an exhausting day, it was a delicacy. The eastern edge of the lake was rimmed by towering cliffs. While I hungrily devoured my painstakingly timed meal, I watched the steep face of rocks and shrubs turn bright yellow, then orange, and then a fiery red before the darkness of night finally consumed them.

I awoke in the daylight, unaware of the time, and leisurely fired up my stove for some morning tea. Everything looked different, shining with light of a quality and angle unlike the afternoon before. I watched a chipmunk run down the shore into the lake, only to immediately leap back out again: wet but not soaked. A power shower? I smiled and turned away from the water to find the doe waiting for me.

“Hi,” I said. “Hello.” I felt the heat of the tea conducting through the mug handle and burning my fingers, but I ignored it, acutely aware of the precious fragility of the moment. But the lithe creature was not there for greetings, but to say goodbye. She tilted her head slightly, one ear focused on me while the other swiveled toward some sound beyond my hearing, and meandered down to the peninsula for some breakfast and a drink of water. Then she bounded off into the woods, quickly vanishing among the shadows and branches.

I sighed and perfunctorily packed my gear. Then I, too, swallowed some dried fruit, took a swig from my Nalgene, and trudged off, away from the lake and the cliffs and along the path home.