Sigbovik 2007

Click for full version PDF

1

2

3

SIGBOVIK 2007

Proceedings

April 1*^st* 2007

Carnegie Mellon University

Pittsburgh, Pennsylvania USA

http://sigbovik.org/

4

5

A Message From the Program Committee

The Association for Computational Heresy Special Special Interest
Group (ACH SIGBOVIK) on Harry Q. Bovik's 6^th Binarennial Workshop
about Sym posium on Robot Dance Party of Conference in Celebration of
Harry Bovik's birthday has a rich and noble legacy. We, the current
members of the program committee were delighted to discover this
legacy as we made it up rooted through the files of the SIGBOVIK
Ministry of History, writing this introduction.

The first SIGBOVIK Conference was held on young Harry Bovik's first
birth day in 1942. It was a brief affair, as all of the attendees had
short, childish attentions spans. This has not changed. This was the
occasion of Fred Hacker's famous talk, Waaaah: Want Bottle. This
talk set a precedent for SIGBOVIK's level of discourse through the
years.

The SIGBOVIK Conference was not held on Harry's second and fourth
birth days. These are known colloquially as the dark years. Harry was
going through a phase at the time. This phase went on for six years.
Even then, Harry had a long-term approach that would serve him well in
his research.

SIGBOVIK IV: BOVIKMANIA was a wild affair. Chuckles the Clown's
keynote was interrupted by what records describe only as The Cake
Incident. Records are unclear about the nature of this incident, but
we think it is safe to say it involved cake. This conference also
marked Harry Bovik and Fred Hacker's first public feud, signified by a
pair of position papers titled Fred Hacker is a doodyhead and Takes
one to know one.

The fifth SIGBOVIK symposium helped establish the conference as the
fore most venue for computer science research and Harry Bovik himself
as its bright est star, noted for his clever ideas and stunning
seventies wardrobe. Of the eighteen presenters, seventeen went on to
win Turing awards. The other was current Governor of California Arnold
Schwarzenegger. However, the spotlight was stolen by Harry himself who
submitted an astonishing thirty-two papers to the conference in honor
of his age. Each was a veritable oyster's pearl of brilliance.
Unfortunately, they were lost in an accident involving three radishes,
four peeled carrots, six celery sticks, and a cuisinart machine.
Demoralized, Harry never replicated those results. Computer Science
still has not recovered from this loss.

This brings us to up to date, ready for the sixth SIGBOVIK Conference
and Harry's sixty-fourth birthday. The minds behind this year's crop
of SIGBOVIK submissions are so bright, they wear sunglasses to bed.
We, on behalf of the SIGBOVIK Organizing Committee, are proud to
present this year's submissions.

-The SIGBOVIK Committee For Writing An Introduction to the Proceedings

6

Contents

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . 5 Table of Contents . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . 8

Papers Not Yet Rejected 8 Track I: Psychopathology and Logic . . . . .
. . . . . . . . . . . . . . 9 Drunken Logic . . . . . . . . . . . . .
. . . . . . . . . . . . . . . 11 Nihilistic Logic . . . . . . . . . .
. . . . . . . . . . . . . . . . . . 17 Bipolar Logic . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . 19 Mad Hatter Logic . . . . .
. . . . . . . . . . . . . . . . . . . . . . 21 Graphomaniac Logic . .
. . . . . . . . . . . . . . . . . . . . . . . 27 Confusion Logic . . .
. . . . . . . . . . . . . . . . . . . . . . . . . 29 Track II:
(Photo-)Realistic Applications . . . . . . . . . . . . . . . . . 31
Applied Birds . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. 33 Applied Garbage . . . . . . . . . . . . . . . . . . . . . . . . .
. . 35 Applied Dereferencing . . . . . . . . . . . . . . . . . . . . .
. . . 37 Applied Thuggery . . . . . . . . . . . . . . . . . . . . . .
. . . . 39 Applied Thievery . . . . . . . . . . . . . . . . . . . . .
. . . . . . 43 Applied Inebriety . . . . . . . . . . . . . . . . . . .
. . . . . . . . 45 Applied Cycle Conservation . . . . . . . . . . . .
. . . . . . . . . 47 Track III: The Meta-Art of Paper-Writing . . . .
. . . . . . . . . . . . 53 Meta-Abstracts . . . . . . . . . . . . . .
. . . . . . . . . . . . . . 55 Meta-Seduction . . . . . . . . . . . .
. . . . . . . . . . . . . . . . 57 Meta-Bananas . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . 59 Meta-Paradigms . . . . . . . .
. . . . . . . . . . . . . . . . . . . 63 Meta-Typesetting . . . . . .
. . . . . . . . . . . . . . . . . . . . . 65 Meta-Detail . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . 77 Meta-Foreshadowing .
. . . . . . . . . . . . . . . . . . . . . . . . 79 Meta-Publabrication
. . . . . . . . . . . . . . . . . . . . . . . . . 83 Track IV: Domo
Arigato, Anonymous Referees . . . . . . . . . . . . . 85 Robotic
Complexity . . . . . . . . . . . . . . . . . . . . . . . . . 87
Robotic Robots . . . . . . . . . . . . . . . . . . . . . . . . . . . .
89 Robotic Uprising . . . . . . . . . . . . . . . . . . . . . . . . .
. . 93 Robotic Mind Language . . . . . . . . . . . . . . . . . . . . .
. . 95 Robotic Toilets . . . . . . . . . . . . . . . . . . . . . . . .
. . . . 97 Track V: Practice makes Perfect; Theory makes Up. . . . . .
. . . . . 99

7

8 CONTENTS

One-Hit Wonder Theory . . . . . . . . . . . . . . . . . . . . . . .
101 Presentation Theory . . . . . . . . . . . . . . . . . . . . . . .
. . 103 Trace Theory . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . 107 Fried Chicken Theory . . . . . . . . . . . . . . . . . . .
. . . . . 109 Adequacy Theory . . . . . . . . . . . . . . . . . . . .
. . . . . . . 115 Demoralization Theory . . . . . . . . . . . . . . .
. . . . . . . . . 117 Nonsense Theory . . . . . . . . . . . . . . . .
. . . . . . . . . . . 121 Uh, Theory? . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . 123 Wiki Theory . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . 127

Comics Supplement 141 Notes 145

Track I:

Psychopathology and Logic 9

10

A non-judgmental reconstruction

of drunken logic

Robert J. Simmons*^∗*

Keywords: Lax logic, ex-lax logic, handwaving logic, drunken
logic, shot glass monad, durnken logic, chemically assisted reasoning,
alcohol in com puter science

We investigate the extension of previous work by Krishnaswami et al.
in [6] on Handwaving Logic, a logic that can be effectively modeled
by Fairtlough and Mendler's Lax Logic [4], towards trying to achieve
a rea sonable formalization of "drunken logic." More advanced
formalizations of drunken logic fail to be modeled effectively by lax
logic, and we argue that much more study deserves to be paid to this
and other concerns which we group together under the umbrella of
Chemically Assisted Reasoning (CAR - but don't drink and drive).
However, unlike various "judgmental" reconstructions, for instance of
modal and lax logic [8], this will not be a judgmental
reconstruction. We're not here to judge, man.

Section 1 briefly re-presents handwaving logic. Section 2 discusses a
simplistic representation of drunken logic that can be modeled by Lax
Logic, whereas Section 3 shows how this modeling behavior breaks down
for a more precise formulation. Section 4 concludes after arguing
(drunkenly!) for more investigation into this and other concerns of
Chemically Assisted Reasoning.

1 Introduction to handwaving logic

Handwaving logic grew out of a concern to create better models of the
way people actually use logic in the real world. Current logical
systems effec tively model logic in a manner acceptable to most
logicians and type the orists; furthermore, the introduction of
substructural logics such as linear

*^∗*This work partially supported by D's Six Pax and Dogz, and
whoever supplies them with napkins.

1

Γ ` A handwave <sup>INTRO</sup>Γ ` A handwave

Γ ` A true

Γ ` A handwave

Γ ` A true^MODAL

Γ ` A ∧ B handwave ^{EXERCISE-FOR-READER-1}

Γ ` B handwave

Γ ` A ∧ B handwave ^{EXERCISE-FOR-READER-2}

Γ ` A ∨ B handwave Γ*, A true ` C handwave*

Γ ` C handwave ^{OTHER-CASE-SIMILAR-1}

Γ ` A ∨ B handwave Γ*, B true ` C handwave*

Γ ` C handwave ^{OTHER-CASE-SIMILAR-2}

Figure 1: Some of the rules of handwaving logic (elimination rules are
the standard ones)

logic shows promise in applying methods from proof theory to the work
of robotics, A.I. and security researchers. However, current proof
theoretic approaches are entirely inadequate for half of the
statements made by an introductory mathematics textbook, and for even
the most basic statements made by your average politician.

Handwaving logic addresses these concerns by conservatively extend ing
standard intuitionistic logic with a handwaving judgment described
by "A handwave", which is internalized in the handwave modality .
The monad admits much more powerful non-standard reasoning techniques
than are generally accepted in the uptight, narrow-minded
intuitionistic logic. The power and convenience of the handwave
modality is evidenced by the fol lowing judgment.

Γ ` B true

Γ ` B handwave ^INTRO

Γ ` A ∧ B handwave ^{EXERCISE-FOR-READER-2}

Γ ` A handwave ^HW-AND-E1

Γ ` A true^MODAL

As described in [6], logic as it is used in the real world can be
modeled by equating A with A. Alternatively, standard techniques
described in [10] can be utilized to reverse-engineer the handwoven
proof obligation as the implicit constraint on the lax monad.

2

2 A brief discussion of drunken logic

As Bovik has famously noted in [2], nowhere outside of undergraduate
lectures are scholars are more prone to sweeping generalizations than
at the bar. Furthermore, judgments which cannot be evidenced outside
of the presence of alcohol, such as the notion "I find A
attractive," obviously may (in some circumstances) be proven under the
constraint of drinking. Drunken logic can simple be expressed, more or
less, by reinterpreting the handwave modality as the shotglass
modality . In this reinterpreta tion, A true maps onto A
attractive, while the modal judgment A handwave maps to A
beergoggles. Our extensive investigations have shown that this model
along with other, similar ones (such as the related game-theoretic
judgement A i'lldoitifyou'lldoit that is vastly amplified under the
modal ity) are sufficient to model the vast majority of lapses in
judgment under the influence of the monad.

3 The challenge of durnken logic

Per Per Martin-Lof [7], something is true when witnessed by an ¨
object of knowledge, which lends itself to an obvious question of
whether the truth of a proposition can be obviated by the presence of
alcohol, seeing as alco hol has an clearly negative impact on one's
knowledge [1]. The possibility of the analytical truth of a
proposition becoming questionable under the influence is also
evidenced by discussion as to whether conference submis sions that can
be understood while drunk are novel enough to be worth accepting.¹

Indeed, in the above presentation of drunken logic the things that are
"potentially true" are in a sense monotonically increasing; there is
no provi sion for things that are true and provable while clean and
sober to be revoked under the influence of alcohol. Put another way,
while A attractive may be true under sufficient "monadic influence,"
it is necessarily negated if one has drunk themselves to sleep,
blindness, or need of medical attention. Speaking of medical
attention, Girard approaches a similar problem in his discussion of
his glossary discussion MEDICINE as a problem of only being able to
work with positive information: "the typical technique in medecine
is to work only with positive information 'As far as we know, one
cannot get AIDS by blood transfusion'" (fixed-width font in the
original) [5].

¹John Renyolds, personal communication on the Wean elevators.

3

We call models that must deal with models that must handle the non
monotonic changes present in extreme modal situations durnken logic;
in fact, drunken logic should be considered merely a special case of
this more general situation. One could imagine that a constraint-based
system might be able to handle blood-alcohol information such that a
sequent is only con sidered in relation to an external constraint C,
represented as C | Γ ` A. A classically-based system might be
able to handle the nonconstructive truth of certain drunken judgments.
Chaudhuri also demonstrates a notion of contradiction in
intuitionistic linear logic such that certain contradictions can arise
without being catastrophic to the overall consistency of the sys tem
[3]. A connection to linear logic would also allow us to investigate
connections between the consumption of resources and the consumption
of alcohol, perhaps giving a satisfactory logical justification for
the truth of the "Three Tequila Proposition":

tequila ⊗ tequila ⊗ tequila ( ⊥

Most promising, perhaps, an approach based on modal or hybrid logic
could internalize states of drunkenness within a Kripke model - this
would potentially generalize to other chemical modalities, such as the
observation of certain researches in this area that application of
their particular modal operator was akin to transportation "out of
this world."²

4 Conclusion: Chemically Assisted Reasoning

It is not our intention in this paper to solve all (or any) of the
problems we present; rather, it is to propose various approaches to
this rich research area that has previously been explored only in the
most ad-hoc manner. In general, we think that the opportunities for
generalization presented in the previous section point to rich
opportunities awaiting researchers in the field of Chemically
Assisted Reasoning (or "CAR" - but don't drink and drive). Reasearch
areas far afield from proof theory, such as work on re gret
minimization algorithms, also have obvious applications to the logic
of drinking, and to chemically assisted reasoning in general [9].

²Anonymous personal communication, Terrace Club, Princeton
University, Fall 2004. 4

References

[1] Harry Q. Bovik. Programming under the influence: a comparative
approach. Mathematical Structures will Drink You Under the Table,
7(7), 1977.

[2] Harry Q. Bovik. My life in a nutshell, if by nutshell you mean
bottle of cheap whiskey. In 13^th International Symposium on
Principles and Practice of Declarative Declarations, February 1985.

[3] Kaustuv Chaudhuri. The Focused Inverse Method for Linear
Logic. PhD thesis, Carnegie Mellon University, December 2006.

[4] Matt Fairtlough and Michael Mendler. Propositional lax logic.
Infor mation and Computation, 137(1):1--33, August 1997.

[5] Jean-Yves Girard. Locus Solum: From the rules of logic to the
logic of rules. Mathematical Structures in Computer Science,
11:301--506, 2001.

[6] Neel Krishnaswami, Rob Simmons, and Carsten Varming. Handwav ing
logic. Journal of the Eighth Floor Whiteboard, December 8, 2006.
Possibly erased.

[7] Per Martin-Lof. On the meanings of the logical constants and the
¨ justifications of the logical laws. Nordic Journal of Philosophical
Logic, 1(1):11--60, 1996.

[8] Frank Pfenning and Rowan Davies. A judgmental reconstruction of
modal logic. Mathematical Structures in Computer Science,
11(4):511-- 540, 2001.

[9] Aaron Roth. Personal communication, Wean Hall 4120.

[10] Privately circulated memoir of the Slovenian Philological
Society, 1947.

5

16

Morality, amorality, and truth irrelevance in a nihilistic type theory

(abstract)

Ezra Broemeling

Carnegie Mellon University

March 6, 2007

The rules of logic are fundamentally arbitrary. How does one choose
between dependent and second-order quantifiers? Between weak and
strong negation? Between linear and bunched implication? Arguments one
way or the other are simply unjustified expressions of faith. In this
paper, I propose new founda tions for logic based on a nihilistic type
theory, which has no introduction rules and no elimination rules.
Nihilistic type theory (NTT) thus defines a logic completely free of
dogma, beyond truth and falsehood---but which is nonethe less
extremely powerful: I sketch soundness and completeness of NTT with
respect to the extended calculus of constructions. Soundness is shown
by giv ing a fully constructive, type-preserving translation taking
NTT terms to ECC terms. Completeness is more difficult, but can be
proven using the technical machinery of truth irrelevance. Finally, I
discuss applications of the nihilistic conditional A : B, used to
express the proposition that A implies B, but it doesn't matter,
because in the end we're all going to die anyways.

1

18

Bipolar Logic with Lithium*^∗*

Daniel K. Lee

Carnegie Mellon University

Abstract

In prior work, we explored Bipolar logic, which contains

two modalities, a manic modality under which any num

ber of intuitionistically unsound propositions can be proven

and a depressed modality under which only the most trivial

propositions could be proven. Despite the promising ap

plications of bipolar logic in the areas of generating Dutch

Post-Impressionist graphs, M.A.C.H.O. expatriate ciphers,

and Grunge compositions proof search in bipolar logic is

made exceedingly difficult by the unpredictable and dra

matic shifts between the manic phase and the depressed

phase. Even the most successful applications of bipolar logic

are difficult to assess, because these most successful pro

grams are so unstable they tend to suddenly self-destruct

with little hope of recovering the old code.

We propose a Linear Bipolar Logic with Lithium. The

presence of lithium resources controls for the sudden shifts

between manic and depressed phases, making proof search

more tractable. However, we leave as an open question

whether Linear Bipolar Logic with Lithium is as expressive

as the traditional presentation of Bipolar Logic.

*^∗*This work is partially supported by the National Science Foun

dation under a Graduate Research Fellowship and D's Six Pax and

Dogz.

20

The Letter before Lambda is Hat: A

Reconstruction of Church's Hat Calculus

Akiva Leffert

Cranberry-Melancholy University

SIGBOVIK 2007

Abstract: We present a reconstruction of Alonzo Church's Hat Calculus
based on notes discovered under a book shelf. We present evidence that
this system was the precursor of the λ-calculus. We then describe
the system in full detail. We prove a lack of progress theorem.
Finally, we prove an undecidability result by reductio ad absurdem.

Keywords: Computability, Hats

1 Introduction

It is said by those with too much imagination that the λ-calclus
sprang fully formed from the head of Alonzo Church and that as he laid
the α, β, and η rules on paper, choruses of angels sang
Hallelujahs. Those with less imagination and perhaps more wit realize,
as Edison said, that genius is hard work and makes one sweaty. Indeed,
even the character λ, namesake of said calculus, was not in Church's
original work. The usual story is that he borrowed notation from
Russell and Whitehead's Principia Mathematica, which used a circumflex
over variables to mark abstractions[1]. Church used this early on,
for example, writing the identity function as: ˆ*x.x*. However, due to
inferior typesetting technology, the circumflex shifted from above the
^{variable to its left like so:}_ˆ^{x.x. This appearance of}_ˆresembles
a capital lambda, Λ, which caused some other typesetter, mind no doubt
dulled by too much exposure to hot-lead, to use the lower-case λ we
know and love. Thus, except for a typographical accident, the
λ-calculus would be known as the circumflex-calculus or, more
succinctly and colloquially, the hat-calculus.

This sepia-tinged tale of typography makes for a good story to tell
little freshlings flush with curiosity about the λ-calculus and the
Entscheidungsproblem[3][6].

1

Harry Bovik, demonstrating his diverse talents, actually made a short
film about this[2] which was well received[5]. However, it is
wrong in one im portant detail. Old notes of Church's, recently
discovered stuffed under a shelf in the Princeton University Library
suggest that this notation was in spired, not by Russell and
Whitehead, but by an earlier system which Church sketched out and
described in those notes. The account of this discovery can be found
in [7]. This system of computation contained more literal hat
symbols - see Figure 1. In the remainder of this paper, we describe
this system, the Hat Calculus, and sketch a proof of the
undecidability of the Down-Feather Problem by reduction from the
Halting Problem.

Definition 2.1 (Up Feather): The Msymbol is an up-feather.

2 Hat Calculus Syntax and Semantics Definition 2.2 (Down Feather): The
Nsymbol is a down-feather.

The complete syntax of the Hat Calculus appears in Figure 2. This
system is considerably more, umm, baroque than the λ-calculus. This
suggests that Church learned a great deal from the development of this
system, abandon

create a banded hat. For example, a Dcan be combined with a Jto create

ing it due to its complexity rather than any inherent computational
weakness d. The latter is a banded hat. For the purposes of
clarification we may,

of the system. Indeed, we later show that this system is
Turing-complete. bined with feathers to create banded feathered
hats. For example, eNis

a down feather banded hat as is dN.

Definition 2.3 (Banded Hat): A hat can be combined with a band to

but probably won't, occasionally refer to hats without bands as naked
hats. Definition 2.4 (Banded Feathered Hat): Hats with bands can be
com

Definition 2.5 (Action Card): All cards in the Hat-Calculus are the
same except for the action cards: ♠₇ , the ♥₈ , the ♥₉ ,
♥₁₀ . Definition 2.6 (Inaction Card): A card which is not an
action card is an inaction card.

Definition 2.7 (n-Carded Banded Hat): Each banded hat is actually an
n-carded banded hat for some n. An n-carded banded hat is a
hat with n cards associated with it. An uncarded hat is just the
degenerate case when n is zero. Note that while there are fifty-two
(four times thirteen (two times

2

3

Figure 1: Excerpt From The Lost Notebook of Alonzo Church

hats H ::= A*|* B*|* C*|* D

feathers F ::= M*|* N

bands B ::= I*|* J*|* K

cards C ::=

♥ ₂ | ♠ ₂ | ♣ ₂ | ♦ ₂ |

♥ ₃ | ♠ ₃ | ♣ ₃ | ♦ ₃ |

♥ ₄ | ♠ ₄ | ♣ ₄ | ♦ ₄ |

♥ ₅ | ♠ ₅ | ♣ ₅ | ♦ ₅ |

♥ ₆ | ♠ ₆ | ♣ ₆ | ♦ ₆ |

♥ ₇ | ♠ ₇ | ♣ ₇ | ♦ ₇ |

♥ ₈ | ♠ ₈ | ♣ ₈ | ♦ ₈ |

♥ ₉ | ♠ ₉ | ♣ ₉ | ♦ ₉ |

♥ ₁₀ | ♠ ₁₀ | ♣ ₁₀ | ♦ ₁₀ |

♥ _J | ♠ _J | ♣ _J | ♦ _J |

♥ _Q | ♠ _Q | ♣ _Q | ♦ _Q |

♥ _K | ♠ _K | ♣ _K | ♦ _K |

♥ _A | ♠ _A | ♣ _A | ♦ _A

Figure 2: Hat-Calculus Syntax

4

adhesive. Thus, CJsteps to c. Unfortunately, this convenient com

two times thirteen)) cards, only four of the cards are distinguished
by the semantics of the language, the action cards. We suspect that
Church was perhaps not at his best when designing this system.

Definition 2.8 (Final State): A hat-calculus expression is considered
final if all of the feathers are down feathers.

The process of computation is the process of attaching bands to hats,
feathers and cards to banded-hats, and stacking hats on other hats. As
in the lamba-calculus, juxtoposition is application, in this case,
application of

bination notation doesn't work as well as we add cards and feathers to
hats. Thus, we use the notation B(H, F, [C₁, . . . ,
C_n]) to represent a completely applied n-carded banded hat. If
we wished to combine an n-carded banded hat with another card, say,
the King of Hearts, ♥_K , we would write this like so: B(H, F,
[C₁, . . . , C_n]) ♥_K . This would step to B(H, F,
[C₁, . . . , C_n, ♥_K ). Note that combining action cards has
a different effect discussed later.

Hats can be stacked. If two hats are juxtaposed we combine them into a
stack. Stacks of hats can also be stacked in this manner. It is not
possible to combine a stack of hats with a single hat in this manner.
The application of a band to a stack of hats has the effect of
applying that band to all of the hats. If the band of a hat is
replaced it loses all of its cards and feathers.

2.1 Action Cards

Combining the ♠₇ card with a hat or stack of hats causes all of
the feathers to flip - i.e. all down feathers become up feathers and
vice versa. Combining the ♥₈ card with a stack of hats removes all
of the hats from the top and bottom until a hat is reached with a
down-feather. Combining the ♥₉ with a stack of hats duplicates the
stack. The ♥₁₀ is the ungluer. It pulls all of the feathers,
bands, and cards off of a hat. It also creates a new action card at
the end of the expression.

The expression cMcannot step, but is not final.

Which particular action card is chosen is non-deterministic.

3 Results

Definition 3.1 (Stuck): An expression of the Hat-Calculus is stuck
if it cannot step and is not a final state.

Theorem 3.1 (Lack of Progress): There exists a stuck state. Proof: 5

Definition 3.2 (Down-Feather Problem): The down-feather problem asks
whether a given Hat-Calculus expression will reduce to a stuck state.
Theorem 3.2 (Universality): We're pretty sure it's Turing complete.
It's got a queue or something.

Theorem 3.3 (Undecidability): The Down-Feather problem is unde
cidable. Like we said, it's probably Turing complete.

4 Conclusion

The Hat-Calculus was developed by Alonzo Church before the
λ-calculus. It is a Turing-complete language of computation with a
rather ungainly syntax. Actually, it's unarguably nonsense[4].
Fortunately, Church later developed the λ-calculus, which isn't crap
(we hope).

References

[1] Henk Barendregt. The impact of the lambda calculus on logic and
com puter science. Bulletin of Symbolic Logic, 3(3):181--215, 1997.

[2] Harry Bovik. Lambda-calculus: The feature film extravangza.
Feature Film.

[3] Alonzo Church. A note on the Entscheidungsproblem. Journal of
Sym bolic Logic, 1:40--41, 1936.

[4] Jean-Yves Girard. Locus solum: From the rules of logic to the
logic of rules. Mathematical. Structures in Comp. Sci.,
11(3):301--506, 2001.

[5] Fred Hacker. Harry Bovik shouldn't be let near a camera. Letter
to the Editor.

[6] Richard Karp. The Entscheidungsproblem is probably NP-complete.
Private communication in an elevator.

[7] William Lovas and Tom Murphy VII. The hidden finds of janitorial
work. Proceedings of Found Stuff Symposium, 9(1):1299--1578, 2005.

6

A Diagrammatic Notation for F; Jason Reed

Carnegie Mellon U

☙ ☙

^G ☂

Abstract

Category Theory, String Theory, Knot Theory,

3 Type Theory

Our typing rules are a healthy part of a sound, complete breakfast.

Graph Theory, Proof Nets, Feynman Diagrams, and Penrose's tensor
contraction notation: all too often has been demonstrated the value of
graphical and

a

_b b a

c

^e⊦ö⊸@

^c c

d

b

e

c

c

b

diagrammatic reasoning in advanced mathematics. By

a

e

d

a

e

b

replacing incomprehensible piles of linear syntax

^d _dd

a

with equally incomprehensible piles of funny

e

cutE

squiggles and wildly pointing arrows, formal diagrams have enhanced
the visual appeal of written work while maintaining, or in some cases
improving on the status quo of outsider-repelling intimidation.

1 Introduction

The vast majority of research in formal systems is

⚛+∫☃d☃ ♠K♡Q♢J♣10

Fig 2. 100 per cent RDA of heavy metal umlauts 4 Related Work

performed on a two-dimensional page (with the exception of the
burgeoning field of Virtual Real

^[Foo99][XYZ03]

$$♀01$$

$$♀♂03$$

Analysis, which requires special goggles and red green differential
operators) and yet this spatiality is often wasted by intrinsically
one-dimensional nota tion. We aim to fix this problem by introducing a
clear and precise two-dimensional notation for the foundations of
mathematics and logic. In the sequel, we try to avoid any use of linear
ordered multi-sets of character-based information units (i.e., ordinary
run ning text) except when strictly necessary.

2 Syntax

[♂04] ^[Bar98] ?

$$✂72$$

_Fi_{g 3. Cryptob}i_bli_ography

5 Conclusion

Fx

Gx

Γ

^Π ★

☞ ^⁋ ♨

^Q ☙

Fig 1. Neither a pullback nor a pushout be

Fig 4. Told you s_o ^ab c

28

First Impressions

A Modal Logic Designed Specifically to Confuse Undergraduates

The Idiran Empire

OUTER SPACE

enslavethegalaxy@hotmail.com

Abstract

Matthew Kehrt

University of Washington mkehrt@cs.washington.edu

beepbeep

1. Introduction

One major attraction of modern type theories lies in their elegant use
of pretty symbols. What modern com puter scientist could not but be
smitten by the simple beauty of such gems as

Γ ` e:τ

or even

Γ ` M:A

Hell, even funny symbols such as those used in

Γ; ∆ ` M:A ( B

are pretty ok.

However, these beautiful strings of symbols serve a deeper purpose
than mere aesthetics. They also al lows us effectively hide very
powerful ideas in a mass confusing typography and terminology. For
example, it is rare that one needs to actually describe what one is
working on when it includes such terms as "pointwise subkinding." [2,
3]

With rising education and theoretical programming language ideas become
more accepted by a broader au dience, it is becoming more and more
necessary to in crease the complexity of terminological and typograph
ical conventions if any actual work is to be accom plished [1].

With this in mind, we present FIRST IMPRESSIONS, a modal logic
designed specifically to confuse the uni tiated.

Propositions P ::= P₁ ∨ P₂ | P₁ ∧ P₂ |
P₁♥P₂ | MP

Atomic Propositions M ::= A, B | 4M | M | · · · | M

Figure 1. Some symbols

2. Confusing Overview

Propositions in FIRST IMPRESSIONS consist of base propositions
prepended by a series of alethic modal ities. Base propositions are
really whatever one feels like: they do not actually matter at all.
We're not ac tually going to do anything with this logic other than
talk about it. So, for instance, one could have a series of atomic
propositions, A, B, ..., and some binary con nectives, *∨,∧,→, ∝,*c
.

Modalities consist of the set of regular polygons. They are
distinguished by how many sides they have. Any polygon with more than
nineteen sides is consid ered to be a circle, which has nineteen
sides. In no tation, these polygons may be written with a dot in the
middle. This has no meaning, but serves merely to mul tiply notation.

A string of modalities can be equivalent to another string of
modalities. A string of modalities followed by a base proposition A
is true iff an equivalent string of propositions followed by A is
true. However, the exact rules determining the provability of the
truth of a proposition are unclear. Therefore, it is necessary to give
heuristics for determining the if two strings of modalities are
equivalent. In practice, we find the most useful way of determing this
is to sum the number of sides of polygons in a string and take that
number modulo twelve. If two strings yield the same number, they might
be equivalent.

Figure 2. A Robot

For example, it is pretty likely that a cicle (nineteen sides) is
equivalent to a heptagon, as 19 ≡ 7 mod 12. We find that discussion
of these rules leads to end less confusion in those who are unaware of
the subtl ties of modal logic. Even those who have much experi ence,
when faced with a pentadecagonal modality, tend to give up quickly.

3. Metatheory

As we do not actually have any judgments, inference rules or even any
particularly well specified syntax, metatheory for this logic consists
mostly of vague state ments about possible future work. This is a
striking return to the philosophical underpinnings of modern logic, and
we hope to someday treat this subject in full.

4. Use

One of the primary uses of this logic is for drawing pictures. Freed
from the tradional constraints of sim ple boxes, diamonds and possibly
circles, we have a near unlimited palette of shapes from which to
choose.

However, we still require these shapes to be regualar which makes more
complicated images difficult to compose. We find that, given these
limits, houses and boxy robots (Figure 2) are some of the most easiest
images to produce.

Acknowledgments

Thanks to Akiva Leffert for being confused, too.

References

[1] J.-Y. Girard. Locus solum: from the rules of logic to the logic
of rules. Mathematical structures in computer science, pages
11:301--506, 2001.

[2] M. A. Kehrt and A. Kehrt. Personal communication, November 2005.

[3] W. Lovas and K. Crary. Hot Compilation: Elaboration. Homework
assignment, November 2005.

Track II:

(Photo-)Realistic Applications 31

32

Towards an Evolutionary Next Generation Avian Carrier Internet
Architecture

Cary Rebecca Pidgin

February 28, 2007

Categories and Subject Descriptors

C.2.1. [Computer Communications and Networks]: Network Architecture
and Design

General Terms

Avian Carriers, Internets, Tubes, Internet Architecture

1 Introduction

In the last 10 years, there have been many design proposals for a next
generation Internet architecture intended to improve the performance,
security, and availability of today's Internet. While these propos als
have mostly leveraged recent advances in network hardware and traffic
trends, there has been a dearth of research looking at a new Internet
architecture based on CPIP (Carrier Pigeon Internet Protocol)[2]. We
believe it is important to fully explore the design space of
architecture possibilities to better inform the community about what
should go into a new archi tecture.

Avian-based protocols have until recently been considered flights of
fancy. In this paper we discuss how to extend the principles of CPIP
(Carrier Pigeon Internet Protocol) to handle a variety of modern ap
plications. In particular, we show how our Avian Carrier Internet (ACI)
architecture supports an evo lutionary model that can adapt with the
changing communication models of Internet traffic. We begin with a short
review of CPIP and some basic addi tions to the protocol and "birdware"
to modernize the technology for today's environment. Next, we

explain ACI's methods of providing an evolutionary framework for
future Internet applications. We then discuss the potential drawbacks
of implementing such a scheme. Eventually, we close with a brief
discus sion of related work and potential avenues for further research
in this area.

2 Birdware Modifications

After months of research, we discovered that the main problem with
RFC1149 and RFC2549 [3] is the use of paper scrolls for data
storage. Such a system does not take advantage of the improving
capacity of flash drives. Our modification is to replace the paper
scroll data storage technology with the latest in NAND flash
technologies. Furthermore, in concert with the avian trend, we use
duck tape instead of duct tape to secure the drive to the birdware's
legs¹. In order to support an end-to-end model of security, the data
on the flash disk can be encrypted by one-time-pad technology. We use
paper scroll data storage tech nology to store the one-time-pad on the
other leg for nearly unbreakable security guarantees.

3 Evolutionary Model

Perhaps the most important contribution of this pa per is the
unintuitive melding of science and tech nology. We leverage off of
recent Darwinian models of evolution, though we are open to other
models of

¹If the birdware is of the duck species, we may be able to avoid
such methods.

1

evolution should they be discovered. In our archi tecture, we use avian
breeders to selectively choose for species traits that are conducive to
the architec ture. The end goal of avian breeders is to produce
fast-flying, super-strong avian carriers. These traits are necessary to
provide low-latency, high-bandwidth characteristics. The intelligence of
the avian carriers must necessarily be sufficient to prevent forwarding
loops from occurring in Super Wide Area Network (SWAN) environments.

As genetics research continues to flourish, we en vision a transition to
genetically-created species to provide a quicker turnaround time for
avian carrier modifications. For example, we may use the hawk's quick
diving speeds along with the pterodactyl's enormous size as one possible
avian carrier species. We also envision the possibility of a hummingbird
cheetah species to support real-time applications such as
Voice-over-Internet-Peregrine and pigeon based CooTube video content
delivery.

4 Issues of Packet Loss

ACI continues to push the 'best-effort' model of data delivery on the
Internet as the original RFCs in tended. As a result, packet loss is
often inevitable. In particular, ACI is susceptible to normal types of
packet loss expected with an avian-based data gram system, such as glass
windows, sky-blue painted buildings, redneck hunters, and avian bird flu
pan demics. We expect that with time, avian carriers can be
scientifically engineered to be resistant to such sources of failure,
although people seem to really like glass and we have no current
solutions to this prob lem (perhaps human computation methods are nec
essary).

One open problem is that of Denial of Service at tacks. Figure 1 shows
an example scenario [1] that could result when an attacker tries to
packet flood a destination comprised of children. One possible solu tion
is to arm every human with a device capable of putting up a wall of
fire, or what we term a 'firewall,' to defend against these attacks.

Figure 1: Potential Damage by DoS Attack in ACI 5 Future Work and
Conclusion

While ACI provides an architecture for an evolution ary
next-generation Internet architecture, its success depends on the
ability of genetics research to combine hummingbirds with cheetahs.
Given the slow process of legislation for avian stem-cell research, we
believe there is merit in leveraging steroids to help provide
incremental benefits to avian-carrier networks. Fi nally, we are
looking at ways to improve latency with wormhole routing using our
EBGW (Early Bird Gets the Wormhole) protocol.

Given our initial goal of exploring the design space of Internet
architectures, we believe ACI stands out as a contrarian design that
should improve network ing research around the world. Our hope is that
other Internet architecture proposals will continue to build off of
our design for make benefit future generations.

References

[1] A. Hitchcock. The Birds, 1963.

[2] D. Waitzman. A Standard for the Transmission of IP Datagrams on
Avian Carriers. RFC 1149 (Unrecommended Standard), April 1990.

[3] D. Waitzman. IP over Avian Carriers with Qual ity of Service.
RFC 2549 (Unrecommended Stan dard), April 1999.

2

Abstract

Compacting, Composting Garbage Collection Jake Donham, Carnegie Mellon
University

sit-and-spin, and a hybrid transcendental, intergen

erational, centrifugal collector.

Garbage collection is vital for programmer efficiency, but hides the
societal costs of rampant waste of data. We present a means to reduce
the negative externali ties of garbage collection through natural
mechanisms of waste reprocessing. We demonstrate a 38% reduc tion in
allocation on a suite of ML benchmarks, as well as a 47% increase in the
growth of plants fertil ized with the rich, loamy byproduct.

Keywords: bioengineering, programming lan guages, dung

1 Introduction

While great strides have been made in recent years in improving the
space and time overhead of garbage collection, as well as its realtime
behavior, little at tention has been paid to the environmental conse
quences of the style of "disposable programming" which garbage
collection encourages, in which large numbers of data structures are
created, only to be thrown away almost immediately. Such abandoned data
puts a strain on the waste management capabil ities of a typical
software system as well as the social context in which the system
operates. Moreover, the cultural impact of this style of programming is
to en courage the wasteful discarding of perfectly good data which could
be repaired and put back into service, thereby stimulating a vibrant
economy of small-scale local artisans along the lines of neighborhood
tailors and shoemakers.

To address one aspect of this deficiency of existing methods of garbage
collection, we propose compact ing, composting garbage collection. The
core idea is that unreachable garbage, once identified by a col lection
algorithm, should not simply be discarded, but should be repaired and
reused if possible, and otherwise encouraged to decay into a
nutrient-rich soil. This method is completely orthogonal to tradi tional
garbage collection algorithms, and in fact we have implemented it in a
family of collectors includ ing stop-and-copy, mark-and-sweep,
clean-and-jerk,

2 The algorithm

The compacting, composting collector works by seg regating the heap
into several piles, corresponding to garbage objects of different
ages. This is dual to the generational approach of segregating live
objects of different ages, and rests on a dual generational hy
pothesis that "dead objects stay dead", or, equiva lently "objects
aren't getting any younger". The idea is that as dead objects age and
decompose, they are moved to successively older piles, which they
share with dead objects of roughly the same age. This serves to
isolate the stinkiest parts of the heap, as well as to produce, in the
oldest generation, a uni formly decayed pile which can be scooped out
and used to fertilize future computations.

There are two additional phases of the algorithm: repair, in which
discarded objects which need only a bit of work with needle and
thread, or some common white glue, or a little oiling, are fixed,
cleaned up, and put back into service; and compaction, in which
garbage piles are pitchforked to break up clumps and then tamped down
with a shovel.

As an optimization, we keep a special pile for in organic objects
which do not decay on the same timescale as typical objects. If
objects on this pile cannot be repaired and reused, they may be taken
to the dump or left at the curb for pickup. Also, we encourage quicker
composting by seeding piles with beneficient bacteria and worms.

3 Our testbed

We have built and measured our collector in the SILT compiler for
Standard ML. SILT (Structured In termediate Language, Too) is a
structure-preserving compiler, which compiles programs by successively
transforming them into a series of structured inter mediate languages,
such as SOIL (Structured Op

1

erational Intermediate Language) and DIRT (Direct Intermediate
RepresenTation). The SILT approach provides large benefits in the form
of increased com piler correctness, additional opportunities for opti
mization, and an organic, holistic, centered user ex perience, man.
Can you dig it?

Test programs included a variety of climate modelling, SETI-at-Home, and
non-violent video game workloads. We attempted to test the collec tor
with a nuclear weapon yield computation and a finance package but found
that these programs made assumptions incompatible with our method.

4 Results

In side-by-side comparisons with a standard collector, we found that
overall 38% of objects could be repaired and reused across the benchmark
suite. The high quality compost resulting from the final pile was sold
to local farms at an average price of $112 per ton.

We have omitted detailed graphs in an effort to cut down on paper.

5 Related work

The most similar work to our is contained in Davis' thesis [1], which
presents the design of a language (called Lollipop) in which you need
only say what you wish to be done, including a memory management
subsystem that picks up after the programmer and puts away his or her
objects neatly. However, Davis does not provide an implementation.

In a tour-de-force of analysis, Smith et al. [4] derive a space bound
on waste produced by a herd of Hol steins on a farm in Iowa. Jonas [2]
proves a theoretical limit on the efficacy of biocomputational methods,
by a reduction to graph-coloring. Murphy [3] evaluates the cache
behavior of locally-grown objects.

6 Future directions

While our method is effective in reducing the waste produced by a
program, purely through modification of the garbage collector, the
larger problem of waste management must be addressed further upstream,
at the point that the garbage is created. We are there fore
re-evaluating methods of explicit memory man agement, in which a
programmer who knows that a particular object can be re-used adds it to
a free list,

from which future allocations can be made. Further more, if the
programmer knows that an object is no longer needed, he or she may
explicitly free it for re cycling, rather than allowing garbage to
accumulate.

Over the long term, we hope to encourage pro grammers to move away
from comfortable, yet environmentally-suspect languages such as ML,
which provide automatic memory management and abstraction facilities
that hide the true origin of data (a form of Marxist commodity
fetishism), and return to the honest, handcrafted code of their
forefathers.

7 Conclusion

We have shown that compacting, composting garbage collection is both
feasible and useful. We hope that this contribution will help bring
about a new age of low-impact programming and green systems.

References

[1] C. Davis. Passive-Aggressive Programming. PhD thesis, Department
of Computer Science, Cran berry Melon University, 2001.

[2] J. Jonas. Crop rotation is NP-complete. In Inter national
Conference on Computational Agricul ture, pages 83--91, Braga,
Portugal, 2003.

[3] T. Murphy VII. Exploiting data locality: Fresh bytes and
community supported agriculture. Journal of Environmental Semantics,
pages 117-- 127, 1998.

[4] J. Smith and Z. Biddleworth. Free-range analysis and abstract
irrigation. In Formal Methods in Farming, pages 190--197, Ames, Iowa,
USA, 1985.

2

C Dereferenced

Akiva Leffert

March 20, 2007

Abstract

We present the *C language. Named in the spirit of C++ and C#, *C is
what happens when the C language is dereferenced. Unfortunately,
dereferencing C results in a SEGFAULT. *C adds a variety of syntactic
features to C, all of which will crash your machine and destroy any
resident data. We also briefly investigated the &C language only to
realize that there was only one real C reference - Kernighan &
Ritchie's The C Programming Language.

1

38

The GUnit

Testing Harness: Achieving Source Code Street Cred Nels E. Beckman

Abstract--- In any large software organization, the street cred

of source code is of principle concern. Often-times we find that

a given code base can talk a mean game, but when time comes

to "throw down," that code base is nowhere to be seen. While

intuitively we as software developers come to gain a sense of

which code bases are trust-worthy (or, "will ball 'till they
fall"),

some systematic method for measuring and reporting this is

necessary. In this paper we present the GUnit testing harness

(pronounced, "Gee Unit"), which does just that.

Index Terms--- Software Engineering, Credibility Improve

ment, Street Life, Hennessy, Testing, Java

I. INTRODUCTION

Slingin' code is a ride-or-die way of life, where every day

it's just another homicide. It is important to stay TRU to tha

game, or you risk being played for a chump. But the ques

tion is, how does one go about evaluating (and eventually

increasing) the TRU-ness of the code that they sling? This

is especially important at large software organizations where

code often becomes soft from spending too much time in

cushy source-code repositories.

In this paper we present the GUnit Testing harness, a

tool created explicitly for the purposes of evaluating the

street cred of a code base. GUnit is the fruit of a rare

cross-discplinary collaboration. Researchers from Carnegie Mellon
University's Institute for Software Research have joined together with
the Interscope Records' Gorilla Unit, lead by world-renown scholar Dr.
Curtis James Jackson III, who goes by the pseudonym 50-Cent.

While GUnit currently only works on Java code bases (an inherrently
'soft' programming language, when compared with other, 'harder'
languages such a C and z80 assem bly [5]) we believe that our
underlying street cred evaluation methodology can easily be extended
to other languages.

This paper proceeds as follows. In Section II, we discuss our
underlying methodology. How does one take any piece of source code and
evaluate its worth in the inner-city? In Section III, we describe
implementation and usage of the GUnit tool. Most helpful to the
end-user is Section III A, which describes the classifications into
which we place evaluated code. Section IV discusses the evaluation of
GUnit and ultimately, and mercifully, Section VI concludes.

II. METHODOLOGY

The GUnit testing harness allows Java developers to build streed cred
tests, and then automatically run and evaluate

N. Beckman is an average to below-average PhD student in the Institute
for Software Research, the most thugged-out department in the School
of Computer Science, Carnegie Mellon University, 5000 Forbes Ave.,
Pittsburgh, PA 15213 nbeckman@cs.cmu.edu

Fig. 1. Screenshot of GUnit: This source code has a street cred rating
of "Elderly Paul Simon."

their results at the touch of a button. Our methods are based on
earlier work by Clifford Smith, AKA the Method Man [6]. These
methods are robust to programmer error and scale to large code bases
as well as large quantities of Cristal. Unfortunately, these methods
are proprietary and cannot be described in full technical detail.

III. THE GUnit TOOL

The GUnit street cred testing harness has been imple mented as an
Eclipse plug-in [1]. Eclipse is gradually re placing Emacs as the
text-editor-that-acts-like-an-operating system of choice for modern
software developers.

A. Street Cred Strata

In order to better convey the the end-user programmer the actual level
of street cred of the code base under test, we have developed the
following scientific classification scheme. It includes ten discreet
levels, listed here in order from furthest to nearest to the street.
We also briefly describe each these gradations.

1) Clay Aiken---At the lowest end of the street cred spec trum is
the rating of "Clay Aiken." Clay's Wikipedia

Fig. 2. Hpnotiq is a 35 proof French fruit liqueur made from vodka,
cognac, and tropical fruit juices [7]. Sounds good to me!

entry describes him as "the most successful second place finisher" in
the history of American Idol. If that's not a dubious destinction for
your software, then what is?

2) CMU Graduate Student---Being a graduate student at Carnegie
Mellon University doesn't get you very far; it won't get you into the
club, and it won't get your court-side seats. Sometimes the truth
hurts.

3) Elderly Paul Simon---Most thuggish software doesn't live to
see 30. Elderly Paul Simon shows us why this is a good thing. Current
best-practices state that one should stay young, fly, and flashy 'till
the day that one dies.

4) P-Diddy---This is a good indication that your source code was
relatively well respected during the past, possibly because it played
some pivotal role in the ca reer of a more successful product. But
when the more successful product was killed in a drive-by shooting in
L.A., everyone kind of realized that your product was more about
marketing than quality. Then your software did a cover of a Led
Zepplin song for the Godzilla remake, which only made things worse. .
.

5) Young Paul Simon---Have you ever see that movie Almost
Famous? He always had big pupils. . . 6) Hpnotiq---Just like
the liquor of the same name, your source code is most frequently seen
at the clubs, in the VIP section. If you seem him out, remember: Don't
start no stuff, won't be no stuff.

7) Chry$ler 300, with Navigation---Seriously, that car is dope.
It's kinda like the cars that the bad guys drove on Batman: The
Animated Series.

8) Ja Rule---If you feel the need to scream "Holla, Holla!" when
your code base receives this rating, that is a perfectly acceptable
reaction. If only it weren't

for that memory leak, the software equivalent of a J Lo collaboration,
you'd have more respect.

9) Dead Prez---While your source code is not, strictly, as
popular as some of the other big names on this list, it is authentic
and has the respect of its peers, due to a prolonged life spend on the
south side of Chicago, and its refusal to sell out. The authors
themselves said it best in their seminal work [2]: That's why I'm
in the dojo, not just for the video. Really though, we really got beef
with the popo. Never know when they gonna put you in a choke hold.

10) Hova---The Michaelangelo of flow, your source code paints
pictures with poems. When source code achieves a street cred rating of
"Hova," this indicates and extremely high degree of credibility. At
the same time, this rating indicates that a given piece of source code
has achieved a rare feat, that of becoming commer cially successful
while still maintaining its reputation of being true to its roots.
Much like Jay-Z, the J-Hova of hip hop, a code base achieving this
rating has flow that is religious.

IV. RESULTS OF TOOL DEPLOYMENT We attempted to try out our tool on a
user base of approximately 500 Java developers. However, due to the
fact that we were constantly screaming, "Wu Tang!" and, "Don't bring
those weak generics into my house!" most of the developers dropped
out. Therefore our current results are inconclusive.

V. RELATED WORK

A. CMMI

Currently, the most well-known metric of street credibility in
existence is the Stuntin' Engineers' Institute's Crunk ability/Make
Money Index (CMMI) [3]. CMMI has long been the de-facto standard
hype metric for military and government software contractors. As its
name suggests, the CMMI ranks the performance of a software
development or ganization in two specific area: Its "crunkability"
(its ability to have fun, drink, and spit game at the ladies) and its
ability to "Make Money" (get paid by any means necessary). Based on
these qualifications, a software organization is given a rank from
one, being the fakest, to five, being the realist. The primary
difficulty with the CMMI system is the large amount of overhead
necessary in having an organization certified. While this is
appropriate for large government contracters where the crunk-ness of
the entire nation is at stake, for smaller, more agile software
organizations, CMMI is more of a burden than anything else. GUnit, on
the other hand, is a small and lightweight evaluation framework. So
small, in fact, that when the cops ever search you or your vehicle,
the chances of them busting you for it are pretty low. Of course
they'll probably just plant a copy of GUnit on you anyway. . .

VI. CONCLUSION

Maintaining the street cred of a source code base has always been a
high priority for software developers. How ever, up until this point
there has been no framework for

automatically creating, running, and evaluating this particular metric.
Therefore in order to fill this gap we have developed the GUnit testing
harness. Source code is currently available online [4].

VII. ACKNOWLEDGMENTS

The author would like to thank you, dear reader, for making it this far.
The metaphors never quite worked, but he kept with it the whole time,
and that's got to count for something. The author would also like to
thank the SIGBOVIK program commitee for graciously extending the
submission deadline. If he had actually taken advantage of that extra
time, this paper would no doubt be much funnier.

REFERENCES

[1] Website for the Eclipse SDK.

http://www.eclipse.org/

[2] stic.man, M-1. Revolutionary but Gangsta. Journal of the
American Underground. March 30, 2004. pp. 1045-1102.

[3] S. Radamenton, A. Robertson, J. J. Walker. CMMI: Guidelines for
Crunkability Improvement. Addison-Wesley, 2003.

[4] GUnit Testing Harness Website.

http://www.g-unitsoldier.com/

[5] W. C. Roeslisberger. z80 Assembly: Fad or In it to Win it? 4^th
Intl. Working Conference on 10x Programmers: Habits, Moods and Whims
Track. February 13, 1989.

[6] C. Smith. M. J. Blige. All That I Need. Journal of the American
Underground. February 1, 1996.

[7] Wikipedia entry, "Hpnotiq."

http://en.wikipedia.org/wiki/Hpnotiq

42

A Theft-Based Approach to 3d Object Acquisition

Ronit Slyper^∗

Carnegie Mellon University

James McCann^†

Carnegie Mellon University

Method Casing? $ avg(∆_s) max(∆_j) E(∆_j) avg(∆_j)
discount (fing.) score Snatch no 0.45 10 3 0.5 0 5 83.8

Snatch yes 0.45 5 3 0.5 0 5 77.5

Jack no 15k 3 6 0.2 0 5 129.1

Jack yes 15k 2 6 0.2 0 5 42.2

Abduct no - 122 12 7.5 1.2 4.5 71.9

Abduct yes - 64 12 7.5 0 5 66.3

Con no 1G 40 9 1 0 5 97.2

Con yes 1G 25 9 1 0 5 22.1

Figure 1: A comparison of various approaches to theft, with and
without casing. Item value is given by $ (in dollars), time for theft
is given in ∆_s (in cm-hours per foot). The empirically encountered
jailtime is avg(∆_j), while max(∆_j) and E(∆_j) are
calculated via lawyer; all are given in months.

Abstract

In graphics, one often wishes to acquire an accurate representation of
an object. Much work has focused on novel devices for acquir ing various
properties of given objects, including geometry, sur face reflectance,
deformation modes, and even sound response. De vices have ranged from
laser scanners to camera arrays and robotic probes, with scanning and
processing times ranging from minutes to days. We present a novel
technique that allows the acquisition of a complete representation of a
wide range of 3d data sets in a few minutes (faster with hardware
acceleration). The representa tion thus obtained can be
photorealistically rendered at haptic rates -- allowing a wide range of
direct interaction and display possibili ties.

CR Categories: I.3.3 [Computer Graphics]: Picture/Image
Generation---Digitizing and Scanning; H.3 [Information Storage and
Retrieval]: Information Search and Retrieval---Retrieval Mod els

Keywords: data acquisition, scanning, haptics, illicit activity

1 Introduction

In this paper we present a novel approach, borrowed from the criminal
underworld, for the acquisition of complete 3d datasets. Datasets
acquired with this approach are suitable for real-time ren dering and
haptic interaction.

^∗e-mail:rys@cs.cmu.edu

^†e-mail:jmccann@cs.cmu.edu

Require: ob j an object

1: if ob j small then

2: return Snatch(ob j)

3: else if ob j self-powered then

4: return Jack(ob j)

5: else if ob j animate then

6: return Abduct(ob j)

7: else if ob j controlled by person then

8: return Con(ob j)

9: else

10: return /0

11: end if

Figure 2: Description of the procedure Steal(ob j). A simple set of
heuristics is required to determine which method to use.

2 Related Work

As a complete object-data capture process, our technique stands to
replace several conventional methods (at least in certain application
domains). We provide a short review of these methods below.

Three-dimensional scanning is a popular technique for acquiring object
geometry, with laser-based [Roach 1997] approaches already
commercialized and widely used. Additional scanning systems rely on
projectors or robotic probes [Bovik 1704]. Some scanners are
additionally able to capture a primitive surface model.

For more complete surface model representation we must turn to
appearance modeling and capture. Work in this field seeks to syn
thesize material BRDFs which match real materials, including skin
[Hefner 1953; Silberstein 1965].

3 Method

3.1 Acquisition

In lieu of an elegant description, we present obtuse pseudocode with
possibly undefined functions (Figure 2) -- as has long been the tra
dition in computer science [Knuth 1981].

3.2 Rendering

One of the benefits of theft-based acquisition is that the data sets so
obtained are self-rendering. That is, they are self-contained and able
to selectively interact with photons in order to create a photo
realistic emission field. Such data sets also have physical presence
which allows one to interact with them in a realistic and wholly
satisfying manner.

3.3 Extensions

Like many approaches in graphics, our method may benefit from hardware
acceleration. We find that a minimal set of hardware (e.g. pry-bar,
lock-pick set, skeleton key, toothpaste) suits many possi ble
applications, while more complicated systems (crane, flat-bed truck,
submersible) may be required for more extreme situations. This systems
allows a flexible cost-benefit trade-off.

Finally, we found that by using a pre-casing phase (so named be cause
it involves "casing the joint" and "scoping things out") we can avoid
common algorithmic pitfalls (security systems, alarms) and increase
overall efficiency. It is also during this pre-casing phase that we
are able to better optimize our selection of hardware accel erators.

4 Evaluation

In order to evaluate our methods, we tested our approach on sev eral
data sets. Theoretical bounds on jailtime (∆_j) were calculated
from criminal justice system records. Averages are presented over a
number of trials sufficient to satisfy our material longings and klep
tomania. In order to provide a reasonable comparison, we use the
unqualified-apathetic-rand norm, as presented in Equation 1 (we set λ
based on time of day).

methodological discussion. Condolences to the Stop-n-shop, East Plaza
Mall, and Marty's Gas-n-Guns. Finally, thanks to our attor ney Paws
and his cat Edward for contributions during the evaluation phase. This
research was supported by IRS grant 15-1040. Ronit Slyper was
additionally supported by an NFL fellowship.

References

BOVIK, H. 1704. Robotic probes and you. Probe-based Robotics 43, 3,
105--205.

HEFNER, H. 1953. Playboy, vol. 1. Playboy Enterprises, Inc.,
December.

KNUTH, D. E. 1981. Seminumerical Algorithms, second ed., vol. 2 of
The Art of Computer Programming. Addison-Wesley, Read ing,
Massachusetts, January.

ROACH, J., 1997. Austin powers: International man of mystery.

SILBERSTEIN, D. 1965. Penthouse, vol. 1. Penthouse Media Group, May.

SLYPER, R., AND MCCANN, J. 2007. A theft-based approach to 3d object
acquisition. In ACH SIGBOVIK, vol. 1, 79--87.

Z

whatever

5 Conclusions

arccos∆_{j −}$²

λ ·∆_s(1)

In this paper [Slyper and McCann 2007] we have presented a new
approach to the acquisition of 3d objects. We hope, in the future, to
become wealthy and well-taken care of by judicious application of this
approach.

One drawback of our approach is the large ∆_j. While our heuristics
do a reasonable job of avoiding a large ∆_j, this could be further
mitigated with a proxy-based approach. In this approach, left for
future work, one pays a third party (normally a member of the local
"crime syndicate") to perform the acquisition.

Another method, which seems to warrant further investigation, is the
direct purchase of goods. While it may seem counterintuitive, the
authors have found that in some cases, costs can actually be lower
than theft -- and ∆*_j*is significantly reduced.

Acknowledgments

Thanks to Frankie (Pen. State) and to Big Joe (State Pen.), for key
hints. Additional thanks to Fast-Fingered Earl for inspiration and

Crash n' Compile: A formalization and empirical study of developer
productivity and software quality through intoxication

Ciera Christopher

April 1, 2007

Abstract

The CrashNCompile process has become a topic of increasing interest in
our community, though up to this point it has been discussed in an infor
mal manner. In this research, we formalize the rules of the
CrashNCompile process using an operational semantics. We verify the
correctness and ter mination of these rules through observation and
wavy-hand-proofs. In this study, we analyze the impact of the
CrashNCompile process on developer productivity and software quality. We
also analyze the many variants of the CrashNCompile process, including
choice of language, compiler, and intoxicating beverage. We conclude by
showing that "Team Distraction", a team comprised entirely of people
with "fuzzy" majors, does indeed have an impact on the quality of the
code produced, though it may in no way reflect the original
requirements.

References

http://crash-n-compile.eorbit.net/index.html

http://www.langston.com/Fun_People/1995/1995AOA.html 1

46

Cycle Depletion -- a Worldwide Crisis¹

Joseph M. Newcomer²and Charles B. Weinstock³

In 1982, the authors did not publish a paper they wrote on the
then-obvious Cycle Depletion problem (CDP). Twenty-five years later,
we are not publishing a twenty-five year retrospective paper on this
problem.

Sadly, the cycle depletion problem continues to worsen. There are
predictions that we face the imminent danger of a cycle depletion
crisis by the year 2020. This paper should serve as a warning to
everyone that we must address the cycle depletion problem immediately
or face the consequences.

In 1982 we had observed that the cycle depletion problem was already
serious.

The cycle depletion problem arises because there are a finite number
of cycles in the Universe, and computers are depleting these at a
ferocious rate.

Computers suck down cycles and emit heat and computation. Without
cycles there can be no computation. This is the fundamental principle
that makes all our computers run. The number of cycles sucked down by
any computer has been increasing (although we

will see that there have been changes that have reduced the need for
cycles by six orders of magnitude, thus allowing us to be in what
appears to be a steady-state situation).

A result we should mention here was with the launch of the Cycle
Isotropy Observatory (CIO) in 1992, it became clear that cycles are
anisotropically distributed throughout the Universe, with heavy
concentrations in some places and far less dense concentrations
elsewhere. The current best cosmological theory is that our solar

system, and for light years around us, is a cycle-poor region to start

with, and therefore we are already at a disadvantage. Work by

Hawking [Hawking99] show that cycle densities near black holes

are incredibly high. Mathematically, the number of cycles

available at the event horizon is infinite, but we simply have no

technology to tap such cycles.

Ptolemy

(Wikipedia)

Cycle theory is nothing new. The earliest recorded work dates back to
Ptolemy (150 A.D.) who postulated that the Universe consisted of cycles
and epicyles. We now know that epicycles are not required, and cycles
are all there are. But, like Democritus and

his theory of atoms (460 B.C.) he was millennia ahead of his time.
However, it wasn't

¹ Portions of this work were unknowingly funded by a grant from
Tartan Laboratories, Inc.

² FlounderCruft, Inc. Email: newcomer@flounder.com

³ Software Engineering Institoot

until modern Quantum Cycle Theory evolved that we began to truly
understand the fundamentals of cycles in the Universe. As this theory
has matured, it has led to the Schwinn postulate of Quantum Computers,
in which all problems are either solved or not solved, and until you
read the solution, you don't know if the problem is solved. (There is
some question as to whether or not a Quantum Computer can solve the
Halting Problem, and there is at least some fear that if such a
problem were presented to a Quantum Computer, the computer would
implode to a singularity and suck down all the cycles for hundreds of
light years around. Others assert that this is the explanation of the
Fermi Paradox: any sufficiently advanced technological culture
eventually gets to the point where they try this experiment, and are
immediately reduced to using slide rules, rendering them incapable of
solving the kinds of problems that would lead to interstellar
communication or faster-than-light travel). Quantum Cycle Theory also
allows us to explain Cycle Anisotropy (CA), although there are some
that say there is no possible way to explain CA

The current value of the Critical Density Ω_Λ is such that the
Universe appears to be permanently expanding, meaning there is only
one cycle (a so-called uni-cycle) to the Universe (there will be no
Big Crunch followed by another Big Bang), so the dominant single
metacycle means that all available cycles are merely subdivisions of
this one cycle. Ultimately, a cycle may not be the atomic unit of
computation, but the equivalent of quarks has not yet been determined.
Nonetheless, there seems to be only one original cycle from which all
others derive. This is one of the open questions of Quantum Cycle
Theory.

A competing theory, derided by some as not being a theory at all, is
that the one distinguished cycle is the source of all other cycles.
Even within this theory there are competing points of view with some
believing in the so-called Unicycle, and others in the so-called
Tricycle -- the latter having more adherents in the community.
Regardless, both communities believe that the distinguished cycle
designed and created all of the other cycles.

A theory of Vacuum Cycles suggests that cycles may spontaneously be
created in the vacuum of space; however, they will be created with a
corresponding anticycle and the two will cancel out according to the
formula e = MC², but given that cycles may be massless, this means
no energy would be produced. There is no good explanation of why our
Universe favors cycles over anticycles.

We first observed the cycle depletion problem at CMU in 1976, when we
moved our 16- processor multiprocessor, C.mmp, into the main computer
room with our KL-10 processor. Shortly thereafter, the KL-10 began to
experience various problems in reliability.

This led to one of those engineering-vs.-science debates. The
engineers, lacking any solid theoretical basis, asserted that the
cause was that the 16-processor system generated too much heat and
overloaded the air conditioning system, raising the temperatures and
causing failures. The scientists, on the other hand, with a firm grasp
of cycle theory

(even in its early form in those days), knew that the real problem was
that the 16- processor system caused a local depletion in the
time-space-cycle continuum, sucking cycles away from the
single-processor KL-10 and causing it to fail.

In those heady days, it was believed the number of available cycles
were unlimited, or at the very least good for centuries. There were
few efforts to conserve cycles.

Early IBM mainframes were equipped with a "usage clock" that ran when
the computer was computing. This measured the amount of usage of the
computer, and hence the monthly rental cost. [We're not making this
up]. If the CPU issued a HALT or WAIT instruction, the CPU clock
stopped running. While advertised as a way of reducing costs, it was
clear that IBM was not creating cycles, and therefore (had cycle
theory been known) was apparently charging for the use of a natural
resource. But the opposing theory is that researchers at T.J. Watson
Research Center had already discovered cycle theory, had grossly
underestimated the available number of cycles (they had once been said
to have stated that the time-space-cycle continuum could support no
more than 12 computers), and had actually done this as a way of
forcing their end users to conserve cycles by disguising it as saving
money.

Those of us who "grew up" in the 1960s had an awareness of
conservation that seems to have disappeared in the 1980s through the
present. Many of us would work late into the night, sometimes
overnight, to ensure that no cycles were wasted. Our more cynical
colleagues said that we were doing this because we got better response
time, but the real reason was that we could not bear to see cycles
being wasted.

[Jack McCredie, faculty at the Carnegie Mellon University Computer
Science Department at that time, observed wistfully at one CS party
one evening in 1972, "Think of all those cycles going over the dam".
This was typical of the concerns we had at that time].

Harry Bovik did one of the fundamental experiments in disproving the
feng shui theory of computing. This theory stated that the proper
alignment of a computer with the cycle ether would result in more
reliable behavior. In the Michaelson-Bovik experiment (unpublished), a
computer was placed on a platform and operated for a period of time.
The platform was then rotated 90º and operated for the same period of
time. No change in the speed of computing or the reliability of the
computer was measured (within experimental error), thus disproving the
idea that cycles exist in the cycliferous ether.

Moore's Law has saved us. As computers are built with smaller and
smaller design rules, the actual number of cycles required has gone
down as the same rate that the cycle speed has increased. A
microcomputer uses microcycles (10^-6 cycles), so a million computers
executing microcycles consume about the same amount as a old mainframe
executing cycles. Intel architectures, for example, do not execute
instructions (in the CISC sense) but instead compile those
instructions into micro-ops (µ-ops), and consequently we have been
largely oblivious to the magnitude of the problem.

Federal standards for cycle conservation have not been well-enforced,
particularly under the Bush administration, which fired all eight of
the Federal Cycle Inspectors that had been hired by the Clinton
administration and signed a contract with Dubai-based Hallibutron.
Modern chips do have some of the cycle-saving mechanisms mandated by
the Cycle Conservation Act of 1994 (a law largely spearheaded by Al
Gore). These include mechanisms to slow the clock down under
conditions of low usage, thus reducing the absolute cycle
requirements; having sleep states and hibernation states that reduce
cycle usage considerably. A public that is hungry for cycles would not
accept these limitations for the sake of simple conservation (witness
our treatment of the oil shortage issue), so these cycle-saving
techniques are marketed as mechanisms for saving power (economic
incentive) or extending battery life (self-interest incentive), but
the truth is that they are there to attempt to reduce cycle usage.

Computer vendors continue to be afraid to let the truth be known about
the upcoming cycle crisis. We already have evidence that IBM in the
1950s knew about this problem, and any vendor directly asked about
this issue will deny it is a problem, while behind the scenes they
work hard to deal with cycle conservation.

The increase in graphics requirements to support Symbolic User
Virtualization, such as is used in video games, is another source of
cycle depletion; graphics cards are simply very sophisticated
domain-specific computers. There is serious concern about the impact
of SUV popularity on our limited resources.

Since cycles are converted to heat, there are some thoughts that

we should be able to have some way to convert heat to cycles.

Unfortunately, the only effective state-of-the-art device we

have is the Stirling Engine, which was invented in 1816. No

more effective mechanism for converting heat to cycles has

been devised. In principle, this engine can work at 80% of the

theoretical Carnot efficiency, but it just doesn't produce

Stirling Engine (Wikipedia)

enough cycles to make a difference. However, advances in nanotechnology
could change this. Currently, nanotechnologists working on various kinds
of mechanical

effectors have not been able to understand the need for micro-Stirling
engines.

Back in the 1960s, far-sighted pioneers such as J.C.R. Licklider saw
the coming catastrophe, and created the ARPANet. One of the stated
purposes of the ARPANet was to allow cycle-sharing, specifically, to
allow users anywhere to share the computing on some non-colocated
processor. This would allow them to place large computing facilities
at the sites of major cycle concentrations, but users in cycle-poor
areas could use these cycles.

NSF has funded a few supercomputer centers. Their apparent random
distribution is not random at all; they are all located at sites of
massive cycle concentrations.

Back when ILLIAC-IV was first installed, the original plan had been to
install it at the University of Illinois campus. As an almost
last-minute decision, the computer was located at the NASA Ames
research facility at Moffet Field in California. The public reason
given was that this was done because they were afraid that student
antiwar activists might break into the computer room at the University
and damage or destroy the computer. This story was simply for public
consumption. Careful studies had shown that U. Ill. was in a
cycle-poor region (no one who had chosen the original site was aware
of cycle theory) whereas NASA Ames had been built (deliberately) in an
area of high cycle availability.

Not all computer sites have been so fortunate. The Livermore
Laboratories Supercomputer Center had been located at a cycle-rich
site, but in later years, for reasons not yet fully understood
(although Quantum Cycle Theory is beginning to yield some results) the
nexus of cycles moved. This required either moving all of Livermore,
or somehow creating more cycles, so a cyclotron was built there.
Unfortunately, these are not cost-effective for everyday computing,
costing tens of millions of dollars to construct and requiring
military-grade budgets to keep running.

Not many people realize the true purpose of the World Wide Web. CERN
suffered from the same problem as Livermore, and in their research to
solve the cycle problem, the WWW emerged. Although the public cover
story is that it makes information readily available, the real truth
is that it is a network for importing cycles from cycle-rich countries
to cycle-poor countries. The reason many emerging countries are
spending such massive efforts on their network infrastructures is that
they hope to become major cycle exporters in the next decade.

Not all emerging economies are in this state, however. The massive
adoption of personal computing in India and China will soon mean that
these countries will become major cycle importers as well. Neither of
these countries are signatories to the Sapporo Accords (named after
the beer drunk at the sushi restaurant where these accords were
proposed), and the U.S. has stated that as long as these countries are
not signatories, there is little that the U.S. can do internally to
reduce its cycle consumption.

The real risk as that the cycle consumption in these countries will
only exacerbate the growing cycle crisis, and competition for the
remaining cycles from small but cycle-rich countries will result in
serious international tensions. Estimates are that wars over cycles
will be inevitable by 2030.

What can we do? We must reduce our cycle consumption. Estimates of 10%
per year improvement will give us time to find alternative cycle
sources, and there are promising research directions that may yield
solutions in the future. Meanwhile, the use of slide rules and
abacuses (abaci?) should be encouraged. PDAs should be replaced by
calendar notebooks. Laptops should be reserved for holding cats, not
computers. Besides, purring cats on one's laptop are far more soothing
than computers, anyway.

References

[Hawking99] Hawking, Stephen, On the Diminishing Computer
Reliability in the Presence of Universal Anamolies, The Journal of
Isotrophic Physics, Vol. 17, No. 7, July 1999, pp1346-1378.

Track III:

The Meta-Art of Paper-Writing 53

54

Abstracter Abstracts*^∗*

Daniel K. Lee

Carnegie Mellon University

Abstract

We did it!

*^∗*This work is partially supported by the National Science Foun

dation under a Graduate Research Fellowship and D's Six Pax and

Dogz.

56

Qualitative Methods for Interagent Communication

Jason Reed^∗

jcreed@cs.cmu.edu

School of Computer Science

Carnegie Mellon University

they turned her down just because --- and I'm told

Abstract

This one time I chatted up a girl at a party, and it ended in
disaster. Don't believe anything she says about it. Keywords:
Anthropomology, Relational Algebra

1 Introduction

Hey there! Hi. How's it going. Peggy? Hi, Peggy. Nice to meet you. You
know, my older sister's [SA83] name is Peggy, too. Yeah, really! No,
I don't think so. Hah, hah! So you came with --- ? How'd you meet him?
Yeah? So you two are --- you know? Oh? No? Oh, okay, I didn't mean to
suggest --- heh!

2 Background

So you're Irish? Yeah, the hair is a dead giveaway, isn't it. Me, well,
on my mom's side, I'm French Canadian, wood pulp, rag stock, and laser
printer toner, and on my dad's side, it's a mix of LaTeX, ASCII, and
Estonian. You wouldn't believe the teas ing I got as a kid!

Boy, the stories I could tell you about --- oh? You have to go to the
bathroom? Okay. I'll still be here when you get back.

3 Related Work

Hey again! Anyway, this one time my little sister [Ali85] got
submitted to a pretty major journal, and

^∗The author would like to thank this paper for writing itself.

this on good authority --- she had a few mistakes involving
substitution 'its' for 'it's'. The injustice! Who doesn't have a few
blemishes at that stage grow ing up, really? It's a difficult time.
Three rejections later, she finally gave up and sent herself out to a
workshop on the Cybernetics of Peer Review Science. My dad [Ali51]
still hasn't forgiven her. He thinks it's just an internet scam --- I
admit the fees were suspiciously large --- but it's done wonders for
her self-esteem.

4 Development

Boy, once i * get a few of those PSPACE-hard lemon ades in me i just
dont know hwat im saying anymore, you know? you know? i guess i really
dont have my i really dont have much tolera i mean tolerance for that
kinda thing heeee! i guess you don't either. i don't know where that
bucket came from either, but you sure tripped over it good Here, let
me help you ups. * Up! Come up! there you go! Anyway i was wondering
if you wanted to come back to my place and uh *hic* overfill my
hboxes if you drift my catch

5 Details of Private Interaction Protocol

Deferred to the Technical Report, to be released June 2007 on Citeseer
Adult Video and Academic Publish ing.

1

6 Open Problems

Hey! You're not supposed to look in there! That's where I keep my old
review forms! Those are sup posed to be private! No! No! Don't read
them! Stop it! Stop reading them! Stop laughing! Stop! Please! No! Why
are you leaving? Give that back! Pleeeease! Please don't tell anyone
he said I'm 'con ceived poorly, articulated worse, and quite possibly
the result of plagiarism so incompetent as to resem ble a peculiar
originality, but one inept far beyond the capabilities of the average
researcher'! Come back!

References

[Ali51] Inter Alia. Denotational higher-order harrop heredity. J.
Hoity Toitology, 1(1):1--137, Oc tober 1951.

[Ali85] E. T. Alia. Hyper-driven device drivers. Draft Manuscript,
August 1985.

[SA83] Verba Sesquiped-Alia. Peg-passing proce dural protocols. In
R. E. D'actor, editor, Programming with Peg Procedures (PPP'83), pages
221--230, K¨aseburg, Wisconsin, Febru ary 1983. Verlag & Sons
Publishing Co.

2

The Epistemology of Exploratory Empiricism: Science

Maoz E. Berlinger, Tom E. Helostpants, Valiant E. Jeenen, Flossyfry E.
Jest, and Sen E. Rajdoc

Departments of Artifactual Engineering and Hermeneutical Computing,
Carnegie Mellon University, 5000 Forbes Avenue, Pittsburgh PA
15213-3890, USA

Abstract

In this paper we explore a sequential linear incremental methodology
and its appli cations to epistemelogical questions related to
exploratory empiricism. By solving the commencement problem we
elucidate the complex relationship between Barth modal grammars and
scientific progress, and demonstrate that solving epistemelog ical
questions is equivalent to scientific discovery. We also briefly
discuss bananas.

1 Introduction

In this paper we consider the intersection of that which we believe
based on evidentiary empiricism extended to novel areas (exploratory
empiricism) with that which is actually true; in other words, we are
seeking to appropriately categorize information gained through
exploratory empiricism as both trust worthy and credible or as failing
on one or both of those counts.

The principal criterion we shall employ as a barometer in estimating
both of the above categorical indices (i.e., trustworthiness,
credibility) is derived from approximating the magnitude and
etymological/lexical entropy of sen tential units generated by an
agent formulating exploratory-empirical hypothe ses about the world,
wholly divorced from the identity of that agent, and in fact from the
semantic content of the aforementioned sentential units, thus ensuring
a strong form of objectivity and impartiality.

Our method is at once absolutely necessary while also utterly
paradoxical. It can be traced back to Archimedean philosophy, which
embodied both evidentiary based methodologies for knowledge
acquisition and metaphysical contempla tions on the validity of
abstract knowledge expressed in a socially constructed format, i.e.
language, and therefore raises key questions about the role of lin
guistics in empiricism in general, and exploratory empiricism
specifically.

2 Bananas and Barth-modal T-grammars

Consider the case of bananas. It is made of sturdy wood, and holds two
gross of this yellow, succulent fruit. Science has determined that the
Cavendish ba nana is likely to become extinct if certain fungi
(considered by the present author to be highly symbolic of the several
threats today facing the scientific establishment) have their way.
Perhaps you should eat a few bananas (see Fig ure 1) right now, before
reading another paragraph, and consider the fleeting opportunity you
have, to wit, to accept this paper and its contribution.

To ourselves we ask, "Why don't you write something already?" The
hermeneu tical implications of this question weigh heavily on our
minds, and in order to avoid answering them we choose to write about
them. No doubt the recent dis covery locating evidentiary empiricism
in the exploratory domain within the category of nonmodal Barth-modal
T-grammars makes the "commencement problem" formidable; fortunately,
our identification of the Snaxx combinator for this T-grammar
("Dubuque, Iowa"; cf. Appendix) mitigates the challenge somewhat.

2

Fig. 1. Yellow, succulent.

Previous reviewers have challenged our use of place-names as
citations, but we pose the following question: Who has been to
Dubuque, Iowa? Really? Perhaps the streets there are littered with
Barth-modal T-grammar Snaxx combinators, for all you know.

As is evidenced by the reader having progressed thus far in the paper,
we have produced a satisfactory solution to the commencement problem
via an intermittent constructivist approach, which paradoxically
relies on social de construction efforts focused on global sales of
bananas.

We have used our technique to analyze itself - self-referentially
applying our linguistic analyses to approximate the trustworthiness of
this very paper - and in doing so revealed unprecedentedly high levels
of credibility in ourselves. This confirms our (now empirically
credible) belief that by combining the epistemologically sound and
theoretically grounded elements of Archimedean philosophy with the
practical lessons of evidentiary empiricism (re: banana appreciation),
we have isolated for the first time in its pure form the true essence
of Science itself.

3 Conclusions

We're going to win best paper award! We're going to win best paper
award! We're going to win best paper award!

3

4 Acknowledgments

We gratefully acknowledge support from the Office of Dense and Opaque
Re search (ODOR) grant number DNS-028841971 and the Dole Food Company
grant number BAN-693997510.

APPENDIX

"Dubuque, Iowa"

4

Synchronistic Hyperparadigmatism

Connor Sites-Bowen

March 20, 2007

Abstract

M. Deschamps was in his teens served a plum pudding by one M. Fort
gibu. In his thirties, M. Deschamps ordered plum pudding at a
restaurant, only to be informed that a M. Fortgibu had just been
served the last piece. Thirty years later, M. Deschamps was treated to
plum pudding at a highly select party. He explained the earlier plum
pudding incidents and remarked that now only M. Fortgibu's presence
was needed to truly complete the scene. Just at that instant, a
delirious and senile Fortgibu entered the party, having mistaken it's
address for that of the engagement he was supposed to be attending
that night. This is Carl Jung's most famous example of synchronicity
(or serendipity).

When caught in the throws of a serendipitous flow of time, one often
wonders if all such experiences have similar phenomenological
characteris tics; Does synchronicity have a higher structure? Can it
be invoked, or if already flowing can it be ridden to a more complete
conclusion? Can a run of serendipitous events be revived after it has
seemingly lapsed? Does meta synchronicity exist? This paper examines
multiple accounts of synchronicity, as well as other acausal
time-space events, in an effort to find a meta-form for Acausal
Parallelism, which can also function as a hyper-paradigm for the
general flow of time. Such a hyper-paradigm is then presented as a
useful philosophical tool to analyze ESP, travel through time, direct
manipulation of synchronistic events, magic, alien abduction, and
other parapsychological processes, especially when superimposed onto a
seemingly causal universe.

1

64

A Systematic Evaluation of the Observed Degradation of Typesetting
Technology in the 20^th Century

Reginald J. Qnuth

March 3, 2007

Abstract

We systematically evaluate typesetting technology over the course of
the 20^th century and discover an astonishing degradation. We
hypothesize on the potential causes of this observed degradation and
conclude that it is the work of malicious time-travelling monkeys.

Keywords: type, systems

1 Introduction

In this work, we advance the hypothesis that typesetting technology
took a dangerously steep downward turn in the latter half of the 20^th
century. This is not a new hypothesis; the world-renowned computer
science genius Knuth made similar observations in the thick of the
matter 30 years ago in March 1977, saying of his gloriously
comprehensive compendium The Art of Computer Programming [2], "I
had spent 15 years writing those books, but if they were going to look
awful, I didn't want to write any more" [3].

We take Knuth's hypothesis and validate it with a systematic and
unbiased evaluation of over three academic papers selected at random
from between the years 1900 and 1999. In Section 1, we introduce our
hypothesis. Section 2 discusses our raw data in detail. Finally, in
Section 3 we draw the startling conclusion that typesetting technology
actually degraded over the course of the 20^th century!

2 Experimental data

2.1 1936: the heady days of the decision problem

Hilbert's 23 problems began a century of glory in mathematics. In
1936, Alonzo Church published a note [1] on what is widely regarded
as "by far the coolest of Hilbert's 23" [8], the
Entscheidungsproblem. We excerpt this note in Figure 2.

1

(Church is also well-known and highly regarded for his work on the hat
calculus; see [4] for a contemporary tutorial introduction.)

This publication represents the pinnacle of publishing quality in the
20^th century, with its fully-typeset mathematics and proportionally
spaced fonts. No characters are hand-drawn, and the kerning is superb.
The footnotes aren't even fragile! An exemplary exemplar of style and
quality---precisely what we've come to expect from an academician as
talented as Church.

Figure 1: Alonzo Church, a very talented academician. (Cheerful, too!)
2

Figure 2: 1936 publication. 3

2.2 1974: a less innocent age

Flash forward to the year 1974. Nixon faces impeachment for the
Watergate scandal. India successfully detonates its first nuclear
weapon. Polymorphism is in its fledgling stages.

Enter John Reynolds.

Figure 3: Bright-eyed and bushy-tailed John Reynolds, with a pipe.

It was in this year that Reynolds published his monumental manuscript
on the polymorphic λ-calculus [5]. Despite being an academic work
of the highest quality [XX cites???], its typesetting left much to
be desired.

As one can see clearly from the scan in Figure 4, 1974 marked an age
of "digital typography"---characters unavailable on the standard
typewriter were
_{drawn in by hand. Examples include the characters ∀, ⊆,}^F,
and (particularly damningly) D. The careful reader will note the
scribbly nature of the tail on the 7→ arrow. But at least we have a
full complement of Greek letters, and everything is sufficiently
well-spaced to be legible . . .

4

Figure 4: 1974 publication. 5

2.3 1983: oh the burning!

Advance the clock 9 years to 1983. Ronnie Reagan leads the United
States in their epic battle with the Evil Empire; the Star Wars
project¹ plays a critical role. Meanwhile Luke and Leia lead the
rebellion in Return of the Jedi. Reynolds does his part by proving
the glorious Abstraction Theorem [6]. Reynolds has

Figure 5: Older, wiser John Reynolds.

grown older, and perhaps wiser, but the quality of his typesetting has
diminished dramatically.

Figure 6 demonstrates the terrible turn taken by typesetting
technology in this war-torn era. Although typewriting technology has
acquired certain key characters (∀, for example), this is only at
the expense of the all-important capital Π, which is more hand-drawn,
larger, and uglier than it's ever been in the history of life on earth
[citation needed]! Agghhh! The math, it burns my eyes!

¹More properly referred to as the Strategic Defense Initiative

6

Figure 6: 1983 publication. 7

2.4 1990: hell on earth

Six years later, the Soviet Union collapses, and Ringard publishes his
seminal preprint on mustard watches [7], shown in Figure 8.

Figure 7: Yann-Joachim Ringard---no relation to Jean-Yves "mad dog"
Girard.

Although this superficially seems to represent an increase in
publishing qual ity, our unbiased opinion is that this apparent
increase is merely illusory. Fig ures are now entirely hand-drawn, and
"smart quotes" are conspicuously absent. "QED" replaces the
traditional ✷. Mathematical quality has taken a similarly downward
turn; Ringard's so-called "proofs" can barely be called sketches.

8

Figure 8: 1990 publication. 9

2.5 A graph

Any good experimental systems paper needs a graph. The graph in Figure
9 shows undeniably that typesetting quality has decreased between 1900
and 1999.

^

|

q | \

u | ----

a | \

l | \

i | ------ - - -

t | \

y | \

| - -

|

+-------------------------------------->

1900 1999

t i m e

Figure 9: A graph

3 Conclusions

Undeniable! QED.

References

[1] Alonzo Church. A note on the Entscheidungsproblem. Journal of
Symbolic Logic, 1:40--41, 1936.

[2] Donald E. Knuth. The Art of Computer Programming.
Addison-Wesley, Reading, MA, 1968--Present.

[3] Donald E. Knuth. Digital Typography. Center for the Study of
Languages and Information, Stanford, CA, 1999.

[4] Akiva Leffert. The letter before lambda is hat: A reconstruction
of Church's hat calculus. In Proceedings of the 6^th Binarennial ACH
Conference in Cel ebration of Harry Q. Bovik's 0x40th Birthday
(SIGBOVIK'07), Pittsburgh, PA, March 2007. ACH Press.

[5] John C. Reynolds. Towards a theory of type structure. In B.
Robinet, editor, Programming Symposium, volume 19 of Lecture Notes
in Computer Science, pages 408--425, Berlin, 1974. Springer-Verlag.

10

[6] John C. Reynolds. Types, abstraction and parametric
polymorphism. In R. E. A. Mason, editor, Information Processing 83,
pages 513--523, Amster dam, 1983. Elsevier Science Publishers B. V.
(North-Holland).

[7] Y. J. Ringard. Mustard watches: an integrated approach to time
and food. Preprint, Universit´e Paris VII, Paris, France, Octobre
1990.

[8] SIGBOVIK Program Committee. Discussion of how the
Entscheidungsprob lem is by far the coolest of Hilbert's 23 problems.
Committee meeting, March 2007.

11

A Appendix: an incomplete waste of paper

Figure 10: "Mad dog" hitting the San Pellegrino.

12

u?Vflƒu¦fl

m̆D̆§saosĬ;†C§rfl¤‰̆q̆;;C‡: aoru»†h̆WC»r̃y`§C»†;Ca‡q

fl@6r6¥ƒ̃[nr"„„4rfl@6r6¥ƒ̃[nr"„„4rfl@6r6¥ƒ̃[nr"„„4rfl@6r6¥ƒ̃[nr"„„4rfl@6r6¥ƒ̃[nr"„„4rfl@6r6¥ƒ̃[nr"„„4rfl@6r6¥ƒ̃[nr"„„4rfl@6r6¥ƒ̃[nr"„„4rfl@6r6¥ƒ̃[nr"„„4rfl@6r6¥ƒ̃[nr"„„4rfl@6r6¥ƒ̃[nr"„„4rfl@6r6¥ƒ̃[nr"„„4rfl@6r6¥ƒ̃[nr"„„4rfl@6r6¥ƒ̃[nr"„„

a‡r 88¡‰Cqah̆r ZYYr C;r Cqr ̆D̆‡r qy::̆q;̆hr ;ar yqr ;x†;r ;x̆r ‰§a;r

qyWW†̈C̆qroär¡‰Cqah̆qr‹r;ẍay:xr‡̆:†;CD̆rC‡oC‡C;¤rC‡ro†»;r†̈̆r

@‡̆raor;x̆r»xC̆or‰̈a`§̆Wqro†»C‡:r†»†h̆WC»r̈̆q̆†̈»x̆̈qrC‡r†§§r†̈̆†qraor
q;yh¤r†̈̆r;x̆rx†̈qxr‰†:̆s§CWC;qrCW‰aq̆hr`¤r»a‡ŏ̈̆‡»̆r‰̈a»̆̆hC‡:qr
†‡hrʼaÿ‡†§q=r¥‡h̆̈rqy»xr»a‡q;̈†C‡;q4rCh̆†qraor§†̈:̆rqCX̆rqCW‰§¤rWyq;r
`̆r;̈y‡»†;̆hr;aroC;=r„‡r‰†‰̆̈rʻ̆r‰̈̆q̆‡;r†r‡̆ʻr;̆L;r§†¤ay;rW̆;xaha§s

a:¤r»†§§̆hrm̆D̆§raorĬ;†C§rfl¤‰̆q̆;;C‡:r;x†;r†§§aʻqroärCh̆†qraor†‡¤r
qCX̆r;ar`̆r̆L‰̈̆qq̆hrC‡r†‡¤rφ»a‡C†‡r‰†:̆r§CWC;rʻC;xay;r;̈y‡»†;Ca‡=r flx̆r
;̆»x‡a§a:¤r †§§aʻqr ;x̆r :̈†»̆oy§r h̆:̈†h†;Ca‡r aor ‰̈̆q̆‡;†;Ca‡r
iy†§C;¤r†qr;x̆r‰†:̆r§CWC;rh̆»̈̆†q̆q=

ʻ̈C;;̆‡rC‡r¤̆§§aʻrq;†̈hyq;ra‡r;x†;rD̆̈¤rxäCXa‡=rn̆;r†§§raor;xCqr ;†K̆qr ‰§†»̆r
C‡r ;x̆r oC‡C;̆r ;ʻashCW̆‡qCa‡†§r †̈̆†r aor ;x̆r qC§D̆̈r q»̈̆̆‡\<r u§;xay:xr
qy»xr ;̆»x‡a§a:¤r ʻ†qr a‡»̆r ;xay:x;r ;ar `̆r ̆L»§yqCD̆§¤r ;x̆r ‰ÿDC̆ʻr aor
V‰†»̆4r ʻ̆r ‡aʻr K‡aʻr xaʻr ;ar ̆Wy§†;̆rC;rC‡r‰̈C‡;=

¦a‡;̆‡;ru;;̆‡y†;Ca‡

„‡;̈ahy»;Ca‡

rrrr

flx̆r‰̈CW†̈¤r;̆»x‡a§a:¤rʻ̆r̆W‰§a¤rCqr;x̆r†`C§C;¤roärWah̆̈‡r

R†»̆hrʻC;xr;x̆rx†̈qxr‰†:̆s§CWC;qraor;x̆r†»†h̆WC»r‰y`§CqxC‡:r
C‡hyq;̈¤4rq»C̆‡;Cq;qr†̈̆r̈ay;C‡̆§¤roä»̆hr;ar§CWC;r;x̆r†Way‡;r
aorC‡oäW†;Ca‡rC‡r‰y`§Cqx̆hr‰†‰̆̈q=rflxCqrW̆†‡qr;x†;rCW‰äs
;†‡;r;̆»x‡C»†§r̆L‰aqC;Ca‡4r‰̈aaoq4roC:ÿ̆q4r»C;†;Ca‡q4roaa;‡a;̆q4r
†‰‰̆‡hC»̆q4r†‡hr‡a‡rq̆iyC;ÿqrWyq;rC‡D†̈C†`§¤r`̆r»y;röaWr†r
‰†‰̆̈r‰̈Cär;ar‰y`§C»†;Ca‡=rflxCqr§̆†hqr;ar`C§§Ca‡qraorha§§†̈qrC‡r
‰y`§C»r†‡hr‰̈CD†;̆r̈̆q̆†̈»xr:̈†‡;qry§;CW†;̆§¤r`̆C‡:rq̆‡;r;ar
;x̆r88»y;;C‡:r̈aaWro§aä=YY

flx̆̈̆r†̈̆r†r‡yW`̆̈raor;̈†hC;Ca‡†§rqa§y;Ca‡qr;ar;xCqr‰̈a`§̆W=r

flx̆r»§†qqC»rqa§y;Ca‡rCqr;ar‰̈ahy»̆r†rq̆‰†̈†;̆r88h̆§̆;̆hrq»̆‡̆qYYr
‰†‰̆̈r•ao;̆‡r»†§§̆hr†r88;̆»x‡C»†§r̈̆‰ä;YYroärxCq;äC»r̈̆†qa‡qc ;x†;rC‡»§yh̆qr †‡¤r
;̆L;r ;x†;rʻ†qr̆L»Cq̆hrhÿC‡:r ;x̆r̆hC;C‡:r
‰̈a»̆qq=rurWC‡äC;¤raor̈̆q̆†̈»x̆̈qr‰y`§Cqxr;x̆C̈r̆‡;C̈̆r¦"Vr är Vy`D̆̈qCa‡r ̈̆DCqCa‡r
xCq;ä¤r •6ÿ‰x¤r PGǴc=r flxCqr
†‰‰̈a†»xrqyoŏ̈qrq̆D̆̈†§rḧ†ʻ`†»Kqtru;r§̆†q;r;ʻarD̆̈qCa‡qraor
;x̆r‰†‰̆̈rWyq;r`̆rqCWy§;†‡̆ayq§¤rW†C‡;†C‡̆h>r†r̈̆†h̆̈rWyq;r
†»iyC̈̆r`a;xr;ar†»xC̆D̆r†§§rK‡aʻ§̆h:̆>r;̆»x‡C»†§r̈̆‰ä;qr†̈̆r
‡a;ryqy†§§¤rqy`ʼ̆»;r;ar‰̆̆̈r̈̆DC̆ʻ4r;xʻ†̈;C‡:r;x̆r†»†h̆WC»r ‰̈a»̆qq>r †‡hr
y‡CD̆̈qC;¤r ‰y`§C»†;Ca‡r q;†‡h†̈hqrW†¤r q;C§§r ¤̆;r
§CWC;r;x̆r§̆‡:;xraor†r;̆»x‡C»†§r̈̆‰ä;=

u‡a;x̆̈r qa§y;Ca‡r öaWr ;x̆r ‡̆ʻqr ʻC̈̆r C‡hyq;̈¤r Cqr ;x̆r
88‰¤̈†WChrWah̆§YYrʻx̆̈̆rC‡oäW†;Ca‡rCqr‰̈̆q̆‡;̆hrC‡r†rq;̈C»;§¤r
Wä̆s;as§̆qqrCW‰ä;†‡;räh̆̈4rqar;x†;r†‡¤r‰̈̆oCLraor†‡r†̈;C»§̆r
W†LCWCX̆qr;x̆r»a‡;̆‡;rD†§y̆roär;x†;r‡yW`̆̈raor»x†̈†»;̆̈q=r flxCqr †§§aʻqr ;x̆r
‰y`§Cqx̆̈r ;ar ‰̈C‡;r ̆L†»;§¤r †qr Wy»xr aor ;x̆r
†̈;C»§̆r†qr»†‡roC;rC‡r;x̆rq‰†»̆r†D†C§†`§̆=rflxCqr;̆»x‡Ciy̆r†§qar
qyoŏ̈qrḧ†ʻ`†»Kq4rW†C‡§¤r;x†;r;x̆r‡†;ÿ†§r;̆L;y†§ro§aʻr»†‡r `̆r C‡;̆̈̈y‰;̆hr
`¤r ;xCqr ‰̈CäC;CX†;Ca‡=r uhhC;Ca‡†§§¤4r ;x̆r
‰¤̈†WChrWah̆§rhăqr‡a;rʻ̆§§r†»»aWWah†;̆rW†C‡q;†¤qraor;x̆r †»†h̆WC»r Wah̆§r
qy»xr †qr oC:ÿ̆q4r »C;†;Ca‡q4r †‡hr ‰a§C;C»†§§¤r
‡̆»̆qq†̈¤r†»K‡aʻ§̆h:̆W̆‡;rq̆»;Ca‡q=

6ah̆̈‡r ;¤‰̆q̆;;C‡:r x†qr ̈̆‡h̆̈̆hr ;x̆r ‰¤̈†WChr Wah̆§r
a`qa§̆;̆=rflx̆rC‡q‰C̈†;Ca‡roär;xCqr;̆»x‡a§a:¤r»aW̆qröaWr;x̆r §̆;;̆̈§†¤C‡:r
̈̆:CW̆r yq̆hr C‡r ;x̆r V;†̈r f†̈qr 6aDC̆r •my»†qr ~5́́c=r „‡r ;x̆r a‰̆‡C‡:r q»̆‡̆r
;̆L;r Cqr hCq‰§†¤̆hr ;ar yqr †;r †‡r †‡:§̆4r ̈̆»̆hC‡:r ;ar ;x̆r D†‡CqxC‡:r
‰aC‡;=rf̆̈̆r aÿr ̆¤̆qC:x;r qyooC»C̆‡;§¤r †»y;̆4r ʻ̆r ʻay§hr `̆r †`§̆r ;ar
̈̆†hr oÿ;x̆̈r †‡hr oÿ;x̆̈rC‡;ar;x̆rhCq;†‡»̆>rqC‡»̆r;x̆rV;†̈rf†̈qr6aDC̆r`̆:C‡qr

»aW‰y;̆̈qr;ar̈̆‡h̆̈r;¤‰̆o†»̆qr†;r†̈`C;̈†̈¤rqCX̆=r¡†»xrq̆»;Ca‡4r
‰†̈†:̈†‰x4rq̆‡;̆‡»̆4rärʻäh4rCqr‰̈CäC;CX̆hr`¤rC;qrʻä;xr†qrC‡r ;x̆r ‰¤̈†WChr
q»x̆W̆=r flx̆‡4r qy»»̆qqCD̆r »aW‰a‡̆‡;qr aor ;x†;r ‰̈CäC;CX̆hr q;̈C‡:r †̈̆r ̈̆‡h̆̈̆hr
†;r ‰̈a:̈̆qqCD̆§¤r h̆»̈̆†qC‡:r
;¤‰̆rqCX̆qrqar†qr;aroC;rC‡r†‡r†̈`C;̈†̈¤r†Way‡;raorq‰†»̆=

flarC§§yq;̈†;̆r;x̆r»a‡»̆‰;4r;xCqr‰†̈†:̈†‰xr»a‡;†C‡qr†r§Cq;raor;x̆r
;xC‡:qr;x†;r;x̆r†y;xär†;̆r;ah†¤4rC‡rh̆»̈̆†qC‡:rCW‰ä;†‡»̆tr
?C:r6†»4r6»Imfl4r†riy†̈;̆̈r‰ay‡h̆̈rʻC;xrqaW̆r»x̆̆q̆4rRC§̆;s@s RCqx4r †r
x†W`ÿ:̆̈4r †r »x̆̆q̆`ÿ:̆̈4r †r [†‰‰¤r6̆†§4r6»vy::̆;q4r ;†q;¤r
:a§h̆‡rö̆‡»xröC̆qr•C‡r̈̆:y§†̈r†‡hr§†̈:̆̈rqCX̆qcr†‡hrq†§†hqr»x̆orär:†̈h̆‡4rär
†r»xC»K̆‡rq†§†hräC̆‡;†§4r`C:r`C:r`̈̆†Ko†q;q4r̆::r6»6yooC‡q4rxa;rxa;r»†K̆qr†‡hrq†yq†:̆4r
W†¤`̆r`Cq»yC;q4r`†»a‡r̆::r†‡hr»x̆̆q̆4rärq†yq†:̆rh†‡Cqxrx†qxr`̈aʻ‡qr;aa4r†‡hroärh̆qq̆̈;tr
xa;r†‰‰§̆r‰C̆4rärqy‡h†̆qr;ẍ̆̆rD†̈C̆;C̆qr†rqao;rq̆̈D̆r»a‡̆4r;ẍ̆̆rKC‡hqraorqx†K̆4rär»xa»a§†;̆§¤r»xC‰r»aaKC̆qr
†‡hr;arḧC‡Kr†r¦a»†s¦a§†rIC̆;r¦aK̆rär䆇:̆rV‰̈C;̆r†‡hr»aoŏ̆r•h̆»†or;aacr†‡hr†§qar䆇:̆rʼyC»̆>r„r§aD̆r6»Ia‡†§hYq4r:aahr
;CW̆4r:̈̆†;r;†q;̆r†‡hr„r:̆;r;xCqr†§§r†;ra‡̆r‰§†»̆=ru§qatru§»axa§C»r`̆D̆̈†:̆4ru§Wa‡hq4ruW†̈†‡;xyq4ruW̆̈C»†‡r»x̆̆q̆q4ru‰‰§̆4ru‰‰§̆r
»y§;CD†̈q4ru‰‰̆;CX̆̈4r?†`¤roaah4r?̆†‡q{m̆:yW̆q4r¦xC»K‰̆†q4r?̈a†hr`̆†‡q4rm̆‡;C§q4r̃̆†q4r̃̆†‡y;q4r̃x†q̆a§yqr`̆†‡q4rVa¤`̆†‡q4rVʻ̆̆;r‰̆†qr•m†;x¤̈yqc4r
?̆̆o4r?̆D̆̈†:̆4ru§»axa§C»r`̆D̆̈†:̆4r¦aoŏ̆4r¡‡̆̈:¤rḧC‡K4r¡q‰̈̆qqa4rRC»;Ca‡†§r`̆D̆̈†:̆q4rRaahqx†K̆4reyC»̆4r̂ä̆†‡r`̆D̆̈†:̆q4r6C§K4r6C§Kqx†K̆4rva‡s†§»axa§C»rḧC‡K4rV§yqx4r
Vao;rḧC‡K4rV‰†̈K§C‡:rʻ†;̆̈4rV‰ä;qrḧC‡K4rfl̆†4rf†;̆̈4r?Cq»yC;4r`̈†‡hq4r?̈†‡hqraor`Cq»yC;4r?̈̆†h4r?̈̆†Ko†q;r»̆̈̆†§q4r?̈C;Cqxr»x̆̆q̆q4r¦†K̆q4r¦†̈`ax¤ḧ†;̆q4r¦†̈`a‡†;̆hrʻ†;̆̈4r¦†̈a`4r¦†;qy‰4r¦̆̈̆†§4r?†̈§̆¤4r
?y»Kʻx̆†;4r̂†Wy;4r6†CX̆4r@†;q4rƒC»̆4rƒ¤̆4rVä:xyW4rV‰̆§;4rfl̈C;C»†§̆4rfl̆oo4rfC§hr̈C»̆4rfx̆†;4r¦̆̈̆†§q4r`̈̆†Ko†q;4r¦x̆̆q̆q4r¦x̆̆q̆q4ruW̆̈C»†‡4r¦x̆̆q̆q4r?̈C;Cqx4r¦x̆̆q̆q4rR̈̆‡»x4r¦x̆̆q̆q4rVʻCqq4r¦xC‡̆q̆rhCqx̆q4r¦ẍCq;W†qr
hCqx̆q4r¦C;̈yq4r¦a»a†4r¦a»a‡y;q4r¦aoŏ̆4r¦aWoä;roaah4r¦a‡hCW̆‡;q4r¦a‡ŏ»;Ca‡̆̈¤4r¦a‡D̆‡C̆‡»̆roaah4r¦äC†‡h̆̈4r¦yCqC‡̆4rI†``†ʻ†§†4rI†C̈¤r‰̈ahy»;4rĬ§C»†»C̆q4rĬqq̆̈;4rIC̆;†̈¤rqy‰‰§̆W̆‡;q4rIC̆;roaah4rICqx̆qt4r~PshCqx̆qr¦ẍCq;W†qr¡D̆rVy‰‰̆̈4rICqx̆q4r¦xC‡̆q̆4rICqx̆q4r¦a§ä†ha4rICqx̆q4r̆::4rICqx̆q4r„;†§C†‡4r
ICqx̆q4r„‡hC†‡4rICqx̆q4r6äa»»†‡4rÏC̆hroaahq4rÏC‡Kqr•V̆̆r†§qar?̆D̆̈†:̆qc4rÏy‰̆4rIÿC†‡rq̆̆hq4r¡::4r¡::rhCqx̆q4r¡‡̆̈:¤rḧC‡K4r¡q‰̈̆qqa4rR†q;roaah4rR†;4rRC»;Ca‡†§r`̆D̆̈†:̆q4rRC‡:̆̈roaah4rRCqx4rRCqx4rR§†DäC‡:4rRaah4rRaahr†hhC;CD̆4rRaahr†hhC;CD̆q4rRaahr†hhC;CD̆q4r¦ah̆Lru§CW̆‡;†̈Cyq4rRaah4ro†q;4rRaah
:̈ay‰q4rRaahr‡†W̆q4rWCq§̆†hC‡:4rRaah4rä:†‡C»4rRaahr‰¤̈†WCh4rRaahr‰̈̆q̆̈D†;CD̆q4rRaahq4rRaahqr‡†W̆hr†o;̆̈r‰̆a‰§̆4rRaahr‰†C̈q4rRaah4rq§aʻ4rRaahq4rq‡†»K4rRaahq4rD̆:†‡4rRaahq4rʻC§h4rR̈̆‡»xr»x̆̆q̆q4rR̈aX̆‡roaah4rR̈yC;4rR̈yC;4r;̈a‰C»†§4rR̈yC;q4rfĭ‡̆;C»†§§¤rWahCoC̆hroaah4rfiC‡K:a4rfï†C‡4r[†X̆§‡y;4r[̆̈`r†‡hrq‰C»̆4r[̆̈`qr†‡hrq‰C»̆q4r[äqrh̆ayD̈̆4r„‡:̈̆hC̆‡;4re†‰†‡̆q̆rC‡q;†‡;r‡aah§̆q4re†‰†‡̆q̆rq‡†»Kq4reyC»̆4rey‡Kroaah4r̂̆;»xy‰4r̂ä̆†‡r`̆D̆̈†:̆q4rm̆:yW̆q4rma»†§roaah4r6†‡:ar»y§;CD†̈q4r6̆†;4r6C§K4r6Cq§̆†hC‡:roaahr‡†W̆q4r6a‡K̆¤s‰yXX§̆4r6yqẍaaW4r̆hC`§̆4r6yq;†̈h4r6yq;†̈hq4rva‡s†§»axa§C»r`̆D̆̈†:̆4r
vaah§̆q4rvaah§̆q4re†‰†‡̆q̆rC‡q;†‡;4rvaD̆§roaah4rvy;̈C̆‡;4rvy;qr†‡hrq̆̆hq4r@̈†‡:̆4r@̈:†‡C»roaah4r̃†q;†4r̃†q;̈¤4r̃xC§C‰‰C‡̆rq‡†»Kroaahq4r̃C‡̆r‡y;q4r¦xC§:aX†r̃C‡̆4r̂ä̆†‡r̃C‡̆4rV;a‡̆r̃C‡̆4r¦a§ä†har̃C‡¤a‡4r6̆LC»†‡r̃C‡¤a‡4rVC‡:§̆s§̆†or̃C‡¤a‡4r̃a‰‰¤rq̆̆hq4r̃äK4r̃ay§;̈¤4r̃̈̆q̆̈D†;CD̆4r̃̈ahy»̆4r̃̈a;̆C‡4r̃yhhC‡:4r̃y§q̆r•§̆:yW̆c4r̃yW‰KC‡rq̆̆hq4rJyC‡a†4rƒ†W`y;†‡4rƒxy`†̈`4rV†§†h4rV†‡hʻC»x4rV†y»̆4rV̆†oaah4rV̆†oaah4r;¤‰̆q4rV̆̆hqr†‡hr‡y;q4rV̆̆hq4r̆hC`§̆4rV̆q†W̆rq̆̆hq4rVCh̆rhCqx4rV§aʻrRaah4rV§yqx4rV‡†»Kroaahq4rV‡†»Kq4re†‰†‡̆q̆4rV‡†»Kq4r̃xC§C‰‰C‡̆4rVao;rḧC‡K4rVao;rḧC‡Kqr`¤r»ay‡;̈¤4rVay§roaah4rVay‰4rVay‰q4rV‰†̈K§C‡:rʻ†;̆̈4rV‰C»̆r†‡hrx̆̈`4rV‰C»̆qr†‡hrx̆̈`q4rV‰ä;qrḧC‡K4rV‰̈̆†h4rV;†‰§̆roaah4rV;̆ʻ4rV;C̈rö¤4rV;̈̆̆;roaah4rV;yh̆‡;roaah4rVy:†̈4rVy‡o§aʻ̆̈rq̆̆hq4rVʻ̆̆;q4rVʻC;X̆̈§†‡hr»x̆̆q̆q4rfl†`aaroaahr†‡hrḧC‡K4rfl̆†4rflCooC‡4rfl̈a‰C»†§röyC;4r"̆:†‡roaahq4r"̆:̆;†`§̆4r"̆:̆;†`§̆q4rf†;̆̈4rfC§hroaahq4rn†W=

uo;̆̈r ;x†;r ̆Lx†yq;CD̆r §Cq;4r ;x̆r ‰†‰̆̈r »†‡r ‰̈a»̆̆hr ʻC;xr ;x̆r ‰̈a‰̆̈r
o§aʻr`¤r̈̆q;äC‡:r;x̆r‡äW†§r oa‡;rqCX̆=rflx̆r†;;̆‡y†s
;Ca‡roy‡»;Ca‡r»†‡r`̆r»xaq̆‡rWä̆rär§̆qqr†̈`C;̈†̈C§¤=rurqCW‰§̆r »xaC»̆r x†§D̆qr
;x̆r oa‡;r qCX̆r †o;̆̈r "r »x†̈†»;̆̈q4r ;x̆‡r "{Pr
»x†̈†»;̆̈q4r;x̆‡r"{Zr»x†̈†»;̆̈q4r̆;»=4r̆‡qÿC‡:r;x†;r†‡rC‡oC‡C;̆r
†Way‡;raor;̆L;r»†‡roC;rC‡r†rqW†§§roC‡C;̆rq‰†»̆rh̆;̆̈WC‡̆hr`¤r"
†‡hr;x̆rq;†̈;C‡:roa‡;rqCX̆=

¡L;̆‡qCa‡q

flxCqr;̆»x‡Ciy̆r‰̈aD̆qryq̆oy§roärWaq;rha»yW̆‡;q4r`y;r‡a;r†§§r Ch̆†qr »†‡r
`̆r ̆†qC§¤r ̆L‰̈̆qq̆hr C‡r §C‡̆†̈r oäW=r Rär C‡q;†‡»̆4r
W†‡¤r†»†h̆WC»rha»yW̆‡;qr»a‡;†C‡r•ärʻay§hr»a‡;†C‡4rCor‡a;r
oär;x̆r§CWC;†;Ca‡qraor;x̆C̈r‰̈CWC;CD̆r~5dGqr;¤‰̆q̆;;C‡:r†‰‰§Cs

»†;Ca‡qcroaa;‡a;̆q4rʻC;xroaa;‡a;̆qr†‡‡a;†;C‡:r;xaq̆roaa;‡a;̆q4r oaa;‡a;̆qr
a‡r ;xaq̆r oaa;‡a;̆q4r oaa;‡a;̆qr y‰a‡r ;xaq̆r oaa;‡a;̆q4r †‡hr y‰a‡r ;xaq̆r
oaa;‡a;̆qrWä̆roaa;‡a;̆q4rʻxC»xr;x̆Wq̆§D̆qrx†D̆r†rD†̈C̆;¤raoroaa;‡a;̆q4r†‡hrC‡r;xaq̆rD̆̈¤r
§C;;§̆roaa;‡a;̆qrqaW̆rWä̆r̆D̆‡rqW†§§̆̈roaa;‡a;̆q4rʻxC»xYqrWC‡y;C†̆rCqr†§qar†y:W̆‡;̆hr`¤rqy‰̆̈q»̈C‰;r‡yW̆̈†§qr
h̆‡a;C‡:r‡a;̆qr†;r;x̆roaa;4rʻxar;x̆WD̆§D̆qr†̈̆rh̆q»̈C`̆hr`¤roaa;‡a;̆q4ra‡r;a‰raorʻxC»xroaa;‡a;̆rh†::̆̈qr†‡hr†q;̆̈CqKqrx†D̆r
`̆̆‡r§†Ch4r†‡hr;xaq̆r‡a;̆qr†§qar‡a;rCWWy‡̆r;aroaa;‡a;̆̈¤4r†‡hrC‡h̆̆hr†:†C‡r;x̆roaa;‡a;̆qr†̈̆r‡a;̆h4r‡a;r;arW̆‡;Ca‡r;xaq̆r;x̆Wq̆§D̆qroaa;‡a;̆qrʻxC»xr†̈̆rqy‰‰§̆W̆‡;̆h4r
†‡hr;xaq̆r‡a;̆qrqx†§§r‡a;r`̆r88§̆o;YYrʻC;xay;r88;ʻarŏ̆;YY4r̆;»=

?̆»†yq̆r;x̆r†Way‡;raor;̆L;rC‡rqy»xrq¤q;̆Wqr»†‡r`̆ry‡»ay‡;s
†`§¤rC‡oC‡C;̆4r†‡r̆L;̆‡qCa‡r;araÿrq¤q;̆WrCqr‡̆»̆qq†̈¤r;ar†§§aʻr
oär;x̆C̈r§†¤ay;=rflxCqr̆L;̆‡qCa‡rCqr`†q̆hra‡rmsq¤q;̆Wqr•ƒ̆̆hr
~55lc=r¥qC‡:4roär̆L†W‰§̆4r;x̆rVC̆̈‰C‡qKCr;̈C†‡:§̆4rʻ̆r»†‡r§†¤r ay;r †‡r
†̈`C;̈†̈¤r C‡oC‡C;̆r ‡̆q;C‡:r aor oaa;‡a;̆q4r †qqyWC‡:r †r oaa;‡a;̆r Cqr
§̆qqr CW‰ä;†‡;r ;x†‡r ;x̆r oaa;‡a;̆r C;r †‡‡a;†;̆q=r

f̆r`̆:C‡r`¤rʻ̈C;C‡:raÿr;̆L;r§C‡̆r†qrC‡rRC:ÿ̆r~=ru;r;x̆r‰aC‡;r
ʻ̆rʻCqxr;arC‡q̆̈;r†roaa;‡a;̆4rʻ̆rW†K̆r†rqW†§§rhC†:a‡†§rhCD̆̈s
qCa‡röaWr;x̆r§C‡̆4ra‡rʻxC»xr;x̆roaa;‡a;̆rCqrhCq‰§†¤̆hr•RC:ÿ̆r
Pc=rVy»xroaa;‡a;̆r§C‡̆qrqy‰‰ä;roaa;‡a;̆qr;x̆Wq̆§D̆qr•RC:ÿ̆r ‹c4r̆;»=

@‡̆r ‰aqqC`§̆r ḧ†ʻ`†»Kr aor qy»xr §†¤ay;qr Cqr ;x†;r ʻxC§̆r ;x̆¤r
†»»aWWah†;̆r†‡rC‡oC‡C;̆r†Way‡;raor;̆L;4r;x̆¤ryq̆rX̆̈ar†̈̆†4r
;̆‡hC‡:r;aʻ†̈hqr†r‰†:̆r;x†;rCqr»aW‰§̆;̆§¤r`§†‡KrC‡r;x̆r§CWC;=r
Räroy;ÿ̆rʻäK4rʻ̆r`̆§C̆D̆r§†¤ay;rq»x̆W̆qr»ay§hr`̆rh̆DCq̆hr `†q̆hr a‡r
q‰†»̆soC§§C‡:r »ÿD̆qr qy»xr †qr ;x̆r [C§`̆̈;r ¦ÿD̆r
•RC:ÿ̆rZc>roärC‡oC‡C;̆r†Way‡;qraor;̆L;rqy»xr»ÿD̆qr‰̈ahy»̆r
†̆q;x̆;C»†§§¤r‰§̆†qC‡:r»aW‰§̆;̆§¤r`§†»Kr‰†:̆qr•RC:ÿ̆r‚c=

¦a‡»§yqCa‡q

¥‡̆‡»yW`̆̈̆hr `¤r ;x̆r »a‡q;̈†C‡;qr aor x†DC‡:r ;ar ‰̈ahy»̆r †r »a‡»Cq̆r
qyWW†̈¤r aor ;xCqr ʻäK4r ʻ̆r C‡q;̆†hr ̈̆‰̆†;r C;r C‡r C;qr rrrr

̆‡;C̈̆;¤rqar†qr‡a;r;ar§̆†D̆r†‡¤;xC‡:ray;=rrrrrrrrrr m̆D̆§saosĬ;†C§rfl¤‰̆q̆;;C‡:

aoru»†h̆WC»r̃y`§C»†;Ca‡q

fl@6r6¥ƒ̃[nr"„„4rfl@6r6¥ƒ̃[nr"„„4rfl@6r6¥ƒ̃[nr"„„4rfl@6r6¥ƒ̃[nr"„„4rfl@6r6¥ƒ̃[nr"„„4rfl@6r6¥ƒ̃[nr"„„4rfl@6r6¥ƒ̃[nr"„„4rfl@6r6¥ƒ̃[nr"„„4rfl@6r6¥ƒ̃[nr"„„4rfl@6r6¥ƒ̃[nr"„„4rfl@6r6¥ƒ̃[nr"„„4rfl@6r6¥ƒ̃[nr"„„4rfl@6r6¥ƒ̃[nr"„„4rfl@6r6¥ƒ̃[nr"„„

[̆̈̆rCqr;x̆rW†C‡r`ah¤r;̆L;= o^aa_;‡^a;^̆ _{:ăq x}̆^̈̆⁴r^C‡rqW†§^§̆̈
^qC^X̆₌

u

[̆̈̆rCqr;x̆rW†C‡r`ah¤r;̆L;=

o^aa_;‡^a;^{̆ :}_{ăq x}̆^̈̆⁴r^C‡r_q^W†§§̆̈ ^qC_{X̆= fl}x^̆
o^aa^;‡_a;^̆q ^;x^̆W^q̆_§D̆q»†‡ x†D̆ oaa;‡a;̆q uq §a‡: †q ;x̆ oaa;‡a;̆
†‰‰̆†̈q ~{‹rä _P^{‹ ^ao ^;x^{̆ ʻ†¤r;}x^̈ay^:x^r;x̆r;^̆L^;=

u

[̆̈̆rCqr;x̆rW†C‡r`ah¤r;̆L;=

RC:ÿ̆qr~4rP4r‹=rR̈†»;†§r§†¤ay;raoroaa;‡a;̆qryqC‡: ;x̆rVC̆̈‰C‡qKCr¦ÿD̆=

u?Vflƒu¦fl

@‡̆raor;x̆r»xC̆or‰̈a`§̆Wqro†»C‡:r†»†h̆WC»r̈̆q̆†̈»x̆̈qrC‡r†§§r †̈̆†qr aor q;yh¤r †̈̆r
;x̆r x†̈qxr ‰†:̆s§CWC;qr CW‰aq̆hr `¤r »a‡ŏ̈̆‡»̆r ‰̈a»̆̆hC‡:qr †‡hr ʼaÿ‡†§q=r
¥‡h̆̈r qy»xr »a‡q;̈†C‡;q4rCh̆†qraor§†̈:̆rqCX̆rqCW‰§¤rWyq;r`̆r;̈y‡»†;̆hr;ar
oC;=r „‡r ‰†‰̆̈r ʻ̆r ‰̈̆q̆‡;r †r ‡̆ʻr ;̆L;r §†¤ay;r W̆;xaha§a:¤r »†§§̆hrm̆D̆§r
aorĬ;†C§rfl¤‰̆q̆;;C‡:r ;x†;r †§§aʻqr oärCh̆†qr aor †‡¤r qCX̆r ;ar `̆r
̆L‰̈̆qq̆hr C‡r †‡¤r φ»a‡C†‡r ‰†:̆r §CWC;r ʻC;xay;r ;̈y‡»†;Ca‡=r flx̆r
;̆»x‡a§a:¤r †§§aʻqr ;x̆r :̈†»̆oy§r h̆:̈†h†;Ca‡r aor ‰̈̆q̆‡;†;Ca‡r iy†§C;¤r †qr
;x̆r ‰†:̆r §CWC;r h̆»̈̆†q̆q=

qyooC»C̆‡;§¤r †»y;̆4r ʻ̆r ʻay§hr `̆r †`§̆r ;ar ̈̆†hr oÿ;x̆̈r †‡hr
oÿ;x̆̈rC‡;ar;x̆rhCq;†‡»̆>rqC‡»̆r;x̆rV;†̈rf†̈qr6aDC̆r`̆:C‡qra‡r
88¡‰Cqah̆rZYYrC;rCqr̆D̆‡rqy::̆q;̆hr;aryqr;x†;r;x̆r‰§a;rqyWW†̈C̆qroär¡‰Cqah̆qr‹r;ẍay:xr‡̆:†;CD̆rC‡oC‡C;¤rC‡ro†»;r†̈̆rʻ̈C;;̆‡rC‡r
¤̆§§aʻr q;†̈hyq;r a‡r ;x†;r D̆̈¤r xäCXa‡=r n̆;r †§§r aor ;xCqr ;†K̆qr
‰§†»̆rC‡r;x̆roC‡C;̆r;ʻashCW̆‡qCa‡†§r†̈̆†raor;x̆rqC§D̆̈rq»̈̆̆‡\<r
u§;xay:xrqy»xr;̆»x‡a§a:¤rʻ†qra‡»̆r;xay:x;r;ar`̆r̆L»§yqCD̆§¤r ;x̆r ‰ÿDC̆ʻr
aor V‰†»̆4r ʻ̆r ‡aʻr K‡aʻr xaʻr ;ar ̆Wy§†;̆r C;r C‡r ‰̈C‡;=

¦a‡;̆‡;ru;;̆‡y†;Ca‡

„‡;̈ahy»;Ca‡

rrrr

flx̆r‰̈CW†̈¤r;̆»x‡a§a:¤rʻ̆r̆W‰§a¤rCqr;x̆r†`C§C;¤roärWah̆̈‡r
»aW‰y;̆̈qr;ar̈̆‡h̆̈r;¤‰̆o†»̆qr†;r†̈`C;̈†̈¤rqCX̆=r¡†»xrq̆»;Ca‡4r

R†»̆hrʻC;xr;x̆rx†̈qxr‰†:̆s§CWC;qraor;x̆r†»†h̆WC»r‰y`§CqxC‡:r
C‡hyq;̈¤4rq»C̆‡;Cq;qr†̈̆r̈ay;C‡̆§¤roä»̆hr;ar§CWC;r;x̆r†Way‡;r aor C‡oäW†;Ca‡r
C‡r ‰y`§Cqx̆hr ‰†‰̆̈q=r flxCqr W̆†‡qr ;x†;r CW‰ä;†‡;r ;̆»x‡C»†§r
̆L‰aqC;Ca‡4r ‰̈aaoq4r oC:ÿ̆q4r »C;†;Ca‡q4r
oaa;‡a;̆q4r†‰‰̆‡hC»̆q4r†‡hr‡a‡rq̆iyC;ÿqrWyq;rC‡D†̈C†`§¤r`̆r
»y;röaWr†r‰†‰̆̈r‰̈Cär;ar‰y`§C»†;Ca‡=rflxCqr§̆†hqr;ar`C§§Ca‡qr aor
ha§§†̈qrC‡r ‰y`§C»r †‡hr ‰̈CD†;̆r ̈̆q̆†̈»xr :̈†‡;qr y§;CW†;̆§¤r
`̆C‡:rq̆‡;r;ar;x̆r88»y;;C‡:r̈aaWro§aä=YY

‰†̈†:̈†‰x4rq̆‡;̆‡»̆4rärʻäh4rCqr‰̈CäC;CX̆hr`¤rC;qrʻä;xr†qrC‡r ;x̆r ‰¤̈†WChr
q»x̆W̆=r flx̆‡4r qy»»̆qqCD̆r »aW‰a‡̆‡;qr aor ;x†;r ‰̈CäC;CX̆hr q;̈C‡:r †̈̆r ̈̆‡h̆̈̆hr
†;r ‰̈a:̈̆qqCD̆§¤r h̆»̈̆†qC‡:r
;¤‰̆rqCX̆qrqar†qr;aroC;rC‡r†‡r†̈`C;̈†̈¤r†Way‡;raorq‰†»̆=

flarC§§yq;̈†;̆r;x̆r»a‡»̆‰;4r;xCqr‰†̈†:̈†‰xr»a‡;†C‡qr†r§Cq;raor;x̆r
;xC‡:qr;x†;r;x̆r†y;xär†;̆r;ah†¤4rC‡rh̆»̈̆†qC‡:rCW‰ä;†‡»̆tr
?C:r6†»4r6»Imfl4r†riy†̈;̆̈r‰ay‡h̆̈rʻC;xrqaW̆r»x̆̆q̆4rRC§̆;s@s

RC:ÿ̆rZ=rflx̆r[C§`̆̈;r¦ÿD̆= ^q _̆xh^‰†_̆^‰q̈ ⁼ _fl^x_Cq^‡

flx̆̈̆r†̈̆r†r‡yW`̆̈raor;̈†hC;Ca‡†§rqa§y;Ca‡qr;ar;xCqr‰̈a`§̆W=r

RCqx4r †r x†W`ÿ:̆̈4r †r »x̆̆q̆`ÿ:̆̈4r †r [†‰‰¤r6̆†§4r6»vy::̆;q4r ;†q;¤r
:a§h̆‡rö̆‡»xröC̆qr•C‡r̈̆:y§†̈r†‡hr§†̈:̆̈rqCX̆qcr†‡hrq†§†hqr»x̆orär:†̈h̆‡4rär

_†^̆_Wqx;

_h̆»

h

^;y^‡C _D†^̈_C†^`§

`

flx̆r»§†qqC»rqa§y;Ca‡rCqr;ar‰̈ahy»̆r†rq̆‰†̈†;̆r88h̆§̆;̆hrq»̆‡̆qYYr
‰†‰̆̈r•ao;̆‡r»†§§̆hr†r88;̆»x‡C»†§r̈̆‰ä;YYroärxCq;äC»r̈̆†qa‡qcr

†r»xC»K̆‡rq†§†hräC̆‡;†§4r`C:r`C:r`̈̆†Ko†q;q4r̆::r6»6yooC‡q4rxa;rxa;r»†K̆qr†‡hrq†yq†:̆4r
W†¤`̆r`Cq»yC;q4r`†»a‡r̆::r†‡hr»x̆̆q̆4rärq†yq†:̆rh†‡Cqxrx†qxr`̈aʻ‡qr;aa4r†‡hroärh̆qq̆̈;tr
xa;r†‰‰§̆r‰C̆4rärqy‡h†̆qr;ẍ̆̆rD†̈C̆;C̆qr†rqao;rq̆̈D̆r»a‡̆4r;ẍ̆̆rKC‡hqraorqx†K̆4rär»xa»a§†;̆§¤r»xC‰r»aaKC̆qr
†‡hr;arḧC‡Kr†r¦a»†s¦a§†rIC̆;r¦aK̆rär䆇:̆rV‰̈C;̆r†‡hr»aoŏ̆r•h̆»†or;aacr†‡hr†§qar䆇:̆rʼyC»̆>r„r§aD̆r6»Ia‡†§hYq4r:aahr
;CW̆4r:̈̆†;r;†q;̆r†‡hr„r:̆;r;xCqr†§§r†;ra‡̆r‰§†»̆=ru§qatru§»axa§C»r`̆D̆̈†:̆4ru§Wa‡hq4ruW†̈†‡;xyq4ruW̆̈C»†‡r»x̆̆q̆q4ru‰‰§̆4ru‰‰§̆r
»y§;CD†̈q4ru‰‰̆;CX̆̈4r?†`¤roaah4r?̆†‡q{m̆:yW̆q4r¦xC»K‰̆†q4r?̈a†hr`̆†‡q4rm̆‡;C§q4r̃̆†q4r̃̆†‡y;q4r̃x†q̆a§yqr`̆†‡q4rVa¤`̆†‡q4rVʻ̆̆;r‰̆†qr•m†;x¤̈yqc4r

§

^†;r^W_C‰^ä ;

C

^‡aq_̆^‡i^y

‰

q

q

C

_;‡_aC^‰_C‡

4

^;y_̈Wq

;x†;rC‡»§yh̆qr †‡¤r ;̆L;r ;x†;rʻ†qr̆L»Cq̆hrhÿC‡:r ;x̆r̆hC;C‡:r

?̆̆o4r?̆D̆̈†:̆4ru§»axa§C»r`̆D̆̈†:̆4r¦aoŏ̆4r¡‡̆̈:¤rḧC‡K4r¡q‰̈̆qqa4rRC»;Ca‡†§r`̆D̆̈†:̆q4rRaahqx†K̆4reyC»̆4r̂ä̆†‡r`̆D̆̈†:̆q4r6C§K4r6C§Kqx†K̆4rva‡s†§»axa§C»rḧC‡K4rV§yqx4r
Vao;rḧC‡K4rV‰†̈K§C‡:rʻ†;̆̈4rV‰ä;qrḧC‡K4rfl̆†4rf†;̆̈4r?Cq»yC;4r`̈†‡hq4r?̈†‡hqraor`Cq»yC;4r?̈̆†h4r?̈̆†Ko†q;r»̆̈̆†§q4r?̈C;Cqxr»x̆̆q̆q4r¦†K̆q4r¦†̈`ax¤ḧ†;̆q4r¦†̈`a‡†;̆hrʻ†;̆̈4r¦†̈a`4r¦†;qy‰4r¦̆̈̆†§4r?†̈§̆¤4r
?y»Kʻx̆†;4r̂†Wy;4r6†CX̆4r@†;q4rƒC»̆4rƒ¤̆4rVä:xyW4rV‰̆§;4rfl̈C;C»†§̆4rfl̆oo4rfC§hr̈C»̆4rfx̆†;4r¦̆̈̆†§q4r`̈̆†Ko†q;4r¦x̆̆q̆q4r¦x̆̆q̆q4ruW̆̈C»†‡4r¦x̆̆q̆q4r?̈C;Cqx4r¦x̆̆q̆q4rR̈̆‡»x4r¦x̆̆q̆q4rVʻCqq4r¦xC‡̆q̆rhCqx̆q4r¦ẍCq;W†qr

`

^a_‡yöa_o

^{o ̆}_;a^4†q‰ ‡

‡

^̆§_a‰^q

^x»

hCqx̆q4r¦C;̈yq4r¦a»a†4r¦a»a‡y;q4r¦aoŏ̆4r¦aWoä;roaah4r¦a‡hCW̆‡;q4r¦a‡ŏ»;Ca‡̆̈¤4r¦a‡D̆‡C̆‡»̆roaah4r¦äC†‡h̆̈4r¦yCqC‡̆4rI†``†ʻ†§†4rI†C̈¤r‰̈ahy»;4rĬ§C»†»C̆q4rĬqq̆̈;4rIC̆;†̈¤rqy‰‰§̆W̆‡;q4rIC̆;roaah4rICqx̆qt4r~PshCqx̆qr¦ẍCq;W†qr¡D̆rVy‰‰̆̈4rICqx̆q4r¦xC‡̆q̆4rICqx̆q4r¦a§ä†ha4rICqx̆q4r̆::4rICqx̆q4r„;†§C†‡4r

‡

aa

¤

‰̈a»̆qq=rurWC‡äC;¤raor̈̆q̆†̈»x̆̈qr‰y`§Cqxr;x̆C̈r̆‡;C̈̆r¦"Vr

ICqx̆q4r„‡hC†‡4rICqx̆q4r6äa»»†‡4rÏC̆hroaahq4rÏC‡Kqr•V̆̆r†§qar?̆D̆̈†:̆qc4rÏy‰̆4rIÿC†‡rq̆̆hq4r¡::4r¡::rhCqx̆q4r¡‡̆̈:¤rḧC‡K4r¡q‰̈̆qqa4rR†q;roaah4rR†;4rRC»;Ca‡†§r`̆D̆̈†:̆q4rRC‡:̆̈roaah4rRCqx4rRCqx4rR§†DäC‡:4rRaah4rRaahr†hhC;CD̆4rRaahr†hhC;CD̆q4rRaahr†hhC;CD̆q4r¦ah̆Lru§CW̆‡;†̈Cyq4rRaah4ro†q;4rRaah

:̈ay‰q4rRaahr‡†W̆q4rWCq§̆†hC‡:4rRaah4rä:†‡C»4rRaahr‰¤̈†WCh4rRaahr‰̈̆q̆̈D†;CD̆q4rRaahq4rRaahqr‡†W̆hr†o;̆̈r‰̆a‰§̆4rRaahr‰†C̈q4rRaah4rq§aʻ4rRaahq4rq‡†»K4rRaahq4rD̆:†‡4rRaahq4rʻC§h4rR̈̆‡»xr»x̆̆q̆q4rR̈aX̆‡roaah4rR̈yC;4rR̈yC;4r;̈a‰C»†§4rR̈yC;q4rfĭ‡̆;C»†§§¤rWahCoC̆hroaah4rfiC‡K:a4rfï†C‡4r[†X̆§‡y;4r[̆̈`r†‡hrq‰C»̆4r[̆̈`qr†‡hrq‰C»̆q4r[äqrh̆ayD̈̆4r„‡:̈̆hC̆‡;4re†‰†‡̆q̆rC‡q;†‡;r‡aah§̆q4re†‰†‡̆q̆rq‡†»Kq4reyC»̆4rey‡Kroaah4r̂̆;»xy‰4r̂ä̆†‡r`̆D̆̈†:̆q4rm̆:yW̆q4rma»†§roaah4r6†‡:ar»y§;CD†̈q4r6̆†;4r6C§K4r6Cq§̆†hC‡:roaahr‡†W̆q4r6a‡K̆¤s‰yXX§̆4r6yqẍaaW4r̆hC`§̆4r6yq;†̈h4r6yq;†̈hq4rva‡s†§»axa§C»r`̆D̆̈†:̆4r

vaah§̆q4rvaah§̆q4re†‰†‡̆q̆rC‡q;†‡;4rvaD̆§roaah4rvy;̈C̆‡;4rvy;qr†‡hrq̆̆hq4r@̈†‡:̆4r@̈:†‡C»roaah4r̃†q;†4r̃†q;̈¤4r̃xC§C‰‰C‡̆rq‡†»Kroaahq4r̃C‡̆r‡y;q4r¦xC§:aX†r̃C‡̆4r̂ä̆†‡r̃C‡̆4rV;a‡̆r̃C‡̆4r¦a§ä†har̃C‡¤a‡4r6̆LC»†‡r̃C‡¤a‡4rVC‡:§̆s§̆†or̃C‡¤a‡4r̃a‰‰¤rq̆̆hq4r̃äK4r̃ay§;̈¤4r̃̈̆q̆̈D†;CD̆4r̃̈ahy»̆4r̃̈a;̆C‡4r̃yhhC‡:4r̃y§q̆r•§̆:yW̆c4r̃yW‰KC‡rq̆̆hq4rJyC‡a†4rƒ†W`y;†‡4rƒxy`†̈`4rV†§†h4rV†‡hʻC»x4rV†y»̆4rV̆†oaah4rV̆†oaah4r;¤‰̆q4rV̆̆hqr†‡hr‡y;q4rV̆̆hq4r̆hC`§̆4rV̆q†W̆rq̆̆hq4rVCh̆rhCqx4rV§aʻrRaah4rV§yqx4rV‡†»Kroaahq4rV‡†»Kq4re†‰†‡̆q̆4rV‡†»Kq4r̃xC§C‰‰C‡̆4rVao;rḧC‡K4rVao;rḧC‡Kqr`¤r»ay‡;̈¤4rVay§roaah4rVay‰4rVay‰q4rV‰†̈K§C‡:rʻ†;̆̈4rV‰C»̆r†‡hrx̆̈`4rV‰C»̆qr†‡hrx̆̈`q4rV‰ä;qrḧC‡K4rV‰̈̆†h4rV;†‰§̆roaah4rV;̆ʻ4rV;C̈rö¤4rV;̈̆̆;roaah4rV;yh̆‡;roaah4rVy:†̈4rVy‡o§aʻ̆̈rq̆̆hq4rVʻ̆̆;q4rVʻC;X̆̈§†‡hr»x̆̆q̆q4rfl†`aaroaahr†‡hrḧC‡K4rfl̆†4rflCooC‡4rfl̈a‰C»†§röyC;4r"̆:†‡roaahq4r"̆:̆;†`§̆4r"̆:̆;†`§̆q4rf†;̆̈4rfC§hroaahq4rn†W=

y

;

̆

†

̆

är Vy`D̆̈qCa‡r ̈̆DCqCa‡r xCq;ä¤r •6ÿ‰x¤r PGǴc=r flxCqr

uo;̆̈r ;x†;r ̆Lx†yq;CD̆r §Cq;4r ;x̆r ‰†‰̆̈r »†‡r ‰̈a»̆̆hr ʻC;xr ;x̆r

†

̆

‡

†

^‰_†‰_̆

_a^̈o

y

†‰‰̈a†»xrqyoŏ̈qrq̆D̆̈†§rḧ†ʻ`†»Kqtru;r§̆†q;r;ʻarD̆̈qCa‡qraor
;x̆r‰†‰̆̈rWyq;r`̆rqCWy§;†‡̆ayq§¤rW†C‡;†C‡̆h>r†r̈̆†h̆̈rWyq;r

‰̈a‰̆̈r o§aʻr `¤r ̈̆q;äC‡:r ;x̆r ‡äW†§r oa‡;r qCX̆=r flx̆r

^CC; ^a L

^ÿ

_C;^‡ C

a^‰;

†;;̆‡y†;Ca‡roy‡»;Ca‡r»†‡r`̆r»xaq̆‡rWä̆rär§̆qqr†̈`C;̈†̈C§¤=rur

;

;

‰

»

†»iyC̈̆r`a;xr;ar†»xC̆D̆r†§§rK‡aʻ§̆h:̆>r;̆»x‡C»†§r̈̆‰ä;qr†̈̆r

_WCC;^;_x^̆W_†a

₄‡^‰a

_C^ä;

‡a;ryqy†§§¤rqy`ʼ̆»;r;ar‰̆̆̈r̈̆DC̆ʻ4r;xʻ†̈;C‡:r;x̆r†»†h̆WC»r

qCW‰§̆r»xaC»̆rx†§D̆qr;x̆roa‡;rqCX̆r†o;̆̈r"r»x†̈†»;̆̈q4r;x̆‡r"{Pr
»x†̈†»;̆̈q4r;x̆‡r"{Zr»x†̈†»;̆̈q4r̆;»=4r̆‡qÿC‡:r;x†;r†‡rC‡oC‡C;̆r

C ^̆̈_̈a

^q4ro:

†

C

‡

^q4»;4

†

‰̈a»̆qq>r †‡hr y‡CD̆̈qC;¤r ‰y`§C»†;Ca‡r q;†‡h†̈hqrW†¤r q;C§§r ¤̆;r

†Way‡;raor;̆L;r»†‡roC;rC‡r†rqW†§§roC‡C;̆rq‰†»̆rh̆;̆̈WC‡̆hr`¤r"

;

_q;q

^C̆_»q

§ _a^Cq_‡aor_h

q^;a^`C

_»^C_;†Ca

W

a

»

q

‡

̆

a

§CWC;r;x̆r§̆‡:;xraor†r;̆»x‡C»†§r̈̆‰ä;=

W

o

^§ _†§̈q

̈

y

†‡hr;x̆rq;†̈;C‡:roa‡;rqCX̆=

^̈¤4

†

̈

_»h^;

_̆‡^¤roa

y

^qx^C _:^‡C

^C_‰^y`^C

u‡a;x̆̈r qa§y;Ca‡r öaWr ;x̆r ‡̆ʻqr ʻC̈̆r C‡hyq;̈¤r Cqr ;x̆r

_hy

C

a

̈

C

Cq

`

88‰¤̈†WChrWah̆§YYrʻx̆̈̆rC‡oäW†;Ca‡rCqr‰̈̆q̆‡;̆hrC‡r†rq;̈C»;§¤r
Wä̆s;as§̆qqrCW‰ä;†‡;räh̆̈4rqar;x†;r†‡¤r‰̈̆oCLraor†‡r†̈;C»§̆r

¡L;̆‡qCa‡q

§

†

‡

C

§^»

§

^x‡fl⁼ §

_ʻx;

^̆W

^̆_;†

W†LCWCX̆qr;x̆r»a‡;̆‡;rD†§y̆roär;x†;r‡yW`̆̈raor»x†̈†»;̆̈q=r

flxCqr;̆»x‡Ciy̆r‰̈aD̆qryq̆oy§roärWaq;rha»yW̆‡;q4r`y;r‡a;r†§§r

‡

_‰»y_`^C§

x^†

;

;

h †

a

;

flxCqr †§§aʻqr ;x̆r ‰y`§Cqx̆̈r ;ar ‰̈C‡;r ̆L†»;§¤r †qr Wy»xr aor ;x̆r

^q̈‰

_C^̈‰

Ch̆†qr »†‡r `̆r ̆†qC§¤r ̆L‰̈̆qq̆hr C‡r §C‡̆†̈r oäW=r Rär C‡q;†‡»̆4r

§^C;_q^a_o;_̆x^†»_†h

‡

^̆ _‡C :^q_̆^‡; §

†̈;C»§̆r†qr»†‡roC;rC‡r;x̆rq‰†»̆r†D†C§†`§̆=rflxCqr;̆»x‡Ciy̆r†§qar

W†‡¤r†»†h̆WC»rha»yW̆‡;qr»a‡;†C‡r•ärʻay§hr»a‡;†C‡4rCor‡a;r

^;x

§

̆

̆

q

C

^̈:^ẍ

_;^qy

†

^;_C^†W _;^a;_x̆

qyoŏ̈qrḧ†ʻ`†»Kq4rW†C‡§¤r;x†;r;x̆r‡†;ÿ†§r;̆L;y†§ro§aʻr»†‡r

oär ;x̆r §CWC;†;Ca‡qr aor ;x̆C̈r ‰̈CWC;CD̆r ~5dGqr ;¤‰̆q̆;;C‡:r

̈

^̆_sW_C x

‡

h

̆

`̆r C‡;̆̈̈y‰;̆hr `¤r ;xCqr ‰̈CäC;CX†;Ca‡=r uhhC;Ca‡†§§¤4r ;x̆r
‰¤̈†WChrWah̆§rhăqr‡a;rʻ̆§§r†»»aWWah†;̆rW†C‡q;†¤qraor;x̆r †»†h̆WC»r Wah̆§r qy»xr
†qr oC:ÿ̆q4r »C;†;Ca‡q4r †‡hr ‰a§C;C»†§§¤r

†‰‰§C»†;Ca‡qcr oaa;‡a;̆q4r ʻC;xr oaa;‡a;̆qr †‡‡a;†;C‡:r ;xaq̆r
oaa;‡a;̆q4roaa;‡a;̆qr a‡r ;xaq̆r oaa;‡a;̆q4roaa;‡a;̆qr y‰a‡r ;xaq̆r oaa;‡a;̆q4r
†‡hr

C

C

̆

̈

̆

̆

D

`

¤

y‰a‡r;xaq̆roaa;‡a;̆qrWä̆roaa;‡a;̆q4rʻxC»xr;x̆Wq̆§D̆qrx†D̆r†rD†̈C̆;¤raoroaa;‡a;̆q4r†‡hr

h

»

†

§

‡̆»̆qq†̈¤r†»K‡aʻ§̆h:̆W̆‡;rq̆»;Ca‡q=

6ah̆̈‡r ;¤‰̆q̆;;C‡:r x†qr ̈̆‡h̆̈̆hr ;x̆r ‰¤̈†WChr Wah̆§r

C‡r ;xaq̆r D̆̈¤r §C;;§̆r oaa;‡a;̆qr qaW̆r Wä̆r ̆D̆‡r qW†§§̆̈r oaa;‡a;̆q4r ʻxC»xYqr
WC‡y;C†̆r Cqr †§qar †y:W̆‡;̆hr `¤r qy‰̆̈q»̈C‰;r ‡yW̆̈†§qr h̆‡a;C‡:r ‡a;̆qr †;r
;x̆r oaa;4rʻxar ;x̆WD̆§D̆qr †̈̆r h̆q»̈C`̆hr `¤r oaa;‡a;̆q4r a‡r ;a‰r aorʻxC»xr
oaa;‡a;̆r
h†::̆̈qr†‡hr†q;̆̈CqKqrx†D̆r`̆̆‡r§†Ch4r†‡hr;xaq̆r‡a;̆qr†§qar‡a;rCWWy‡̆r;aroaa;‡a;̆̈¤4r†‡hrC‡h̆̆hr†:†C‡r;x̆roaa;‡a;̆qr†̈̆r‡a;̆h4r‡a;r;arW̆‡;Ca‡r;xaq̆r
;x̆Wq̆§D̆qroaa;‡a;̆qrʻxC»xr†̈̆rqy‰‰§̆W̆‡;̆h4r†‡hr;xaq̆r‡a;̆qrqx†§§r‡a;r`̆r88§̆o;YYrʻC;xay;r88;ʻarŏ̆;YY4r̆;»=

R

†

»

q ‡

†

̆

;

a`qa§̆;̆=rflx̆rC‡q‰C̈†;Ca‡roär;xCqr;̆»x‡a§a:¤r»aW̆qröaWr;x̆r §̆;;̆̈§†¤C‡:r ̈̆:CW̆r
yq̆hr C‡r ;x̆r V;†̈r f†̈qr 6aDC̆r •my»†qr

?̆»†yq̆r ;x̆r †Way‡;r aor ;̆L;r C‡r qy»xr q¤q;̆Wqr »†‡r `̆r
y‡»ay‡;†`§¤rC‡oC‡C;̆4r†‡r̆L;̆‡qCa‡r;araÿrq¤q;̆WrCqr‡̆»̆qq†̈¤r

§

†

:

~5́́c=r „‡r ;x̆r a‰̆‡C‡:r q»̆‡̆r ;̆L;r Cqr hCq‰§†¤̆hr ;ar yqr †;r †‡r †‡:§̆4r
̈̆»̆hC‡:r ;ar ;x̆r D†‡CqxC‡:r ‰aC‡;=rf̆̈̆r aÿr ̆¤̆qC:x;r

;ar†§§aʻroär;x̆C̈r§†¤ay;=rflxCqr̆L;̆‡qCa‡rCqr`†q̆hra‡rmsq¤q;̆Wqr
•ƒ̆̆hr~55lc=r¥qC‡:4roär̆L†W‰§̆4r;x̆rVC̆̈‰C‡qKCr;̈C†‡:§̆4rʻ̆r »†‡r §†¤r ay;r †‡r
†̈`C;̈†̈¤r C‡oC‡C;̆r ‡̆q;C‡:r aor oaa;‡a;̆q4r

RC:ÿ̆r‚=rfl̆L;r†̆q;x̆;C»†§§¤r§†Chray;ra‡r† [C§`̆̈;r¦ÿD̆=

f̆r`̆:C‡r`¤rʻ̈C;C‡:raÿr;̆L;r§C‡̆r†qrC‡rRC:ÿ̆r~=ru;r;x̆r‰aC‡;r ʻ̆r ʻCqxr ;ar
C‡q̆̈;r †r oaa;‡a;̆4r ʻ̆r W†K̆r †r qW†§§r hC†:a‡†§r hCD̆̈qCa‡r öaWr ;x̆r§C‡̆4r
a‡rʻxC»xr ;x̆r oaa;‡a;̆rCqrhCq‰§†¤̆hr
•RC:ÿ̆rPc=rVy»xroaa;‡a;̆r§C‡̆qrqy‰‰ä;roaa;‡a;̆qr;x̆Wq̆§D̆qr •RC:ÿ̆r‹c4r̆;»=

@‡̆r ‰aqqC`§̆r ḧ†ʻ`†»Kr aor qy»xr §†¤ay;qr Cqr ;x†;r ʻxC§̆r ;x̆¤r
†»»aWWah†;̆r†‡rC‡oC‡C;̆r†Way‡;raor;̆L;4r;x̆¤ryq̆rX̆̈ar†̈̆†4r
;̆‡hC‡:r;aʻ†̈hqr†r‰†:̆r;x†;rCqr»aW‰§̆;̆§¤r`§†‡KrC‡r;x̆r§CWC;=r

[̆̈̆rCqr;x̆rW†C‡r`ah¤r;̆L;= u ^oa_a;‡a;^{̆ :}_ă^q_rx̆̈^{̆4 C‡}
q^W†§§̆^̈ q_C^X̆₌

[̆̈̆rCqr;x̆rW†C‡r`ah¤r;̆L;= u ^oaa^;_‡a;^{̆ :}_ă^q _x̆̈^̆4
C_{‡ qW}^†§^§̆^̈rqC^X̆=

uh†Wq4r¦x†̈§C̆ruh†Wq4r̂̆C;xruh†Wqr•uW̆̈C»†‡roaa;`†§§c4r̂̆§§¤ruh†Wq4r?†`†ruh†Wy4r@ẄCruŏK4reaq̆‰xruooy§4r6†LCWruoC‡a:̆‡aD4ruöa`4rv†X†‡C‡ruoqxC‡se†W4r¦†̈§aqru:†qqC4ru:̆‡;r64rv†̈a†ru:C̈̈̆4r
ey‡Cäru:a:a4ru§̆qq†‡ḧaru:aq;C‡C4r?ʼQ̈‡rQ:̈̆‡4r̃̆̈̈aru:y†¤a4rë=4r¦ẍCqru:yC§†4r6†̈DC‡ru:yq;Q‡4ruxW†hr[†qq†‡rV;†KaX†4r̃†`§aruCW†̈4rIa‡‡†ruC̈4rVay†hruQ;rV†§̆W4rĕ‰KäC̈ruC¤†`̆C4r@̈x†‡ruK4reCWWC̆r
QK̆qqa‡4ru`hy§†xru§r[†WC̈C4rR†ʻ†Xr†§sƒ†`̆C̆̆4rV†§̆xru§sV†ïC4r6yx†WW†hru§†hhC‡4ru§†‡r@qQ̈Carh†r¦aq;†rVC§D†4r¦†̈§aru§`†‡4r¦ẍCq;Cʼ†‡ru§`̆̈q4r6†̈Karu§`̆̈;4r¦ẍCqru§`̈C:x;4r[y:aru§»†‡;†̈†4ru§̆qq†‡ḧar

Räroy;ÿ̆rʻäK4rʻ̆r`̆§C̆D̆r§†¤ay;rq»x̆W̆qr»ay§hr`̆rh̆DCq̆hr

q

`†q̆hr a‡r q‰†»̆soC§§C‡:r »ÿD̆qr qy»xr †qr ;x̆r [C§`̆̈;r ¦ÿD̆r

̆

D

§

_{q ;x̆}W^q̆

̆

;

<sup>6äCrvy‡̆q4rI†‡ru§̆L†4rfi†̈¤ru§̆L†‡h̆̈r•oaa;`†§§̆̈c4rflahhru§̆L†‡h̆̈4r¡§CQX̆̈ru§oa‡Xa4r6y†X†Wru§C4r6yx†WW†hru§C4r̃̈C‡»̆raorV†Ch4rfl†;¤†‡†ru§C4rfl=re=ru§§†̈h4r¡§CX†`̆;xru‡‡̆ru§§̆‡4re†W̆qru§§̆‡r•uW̆̈C»†‡roaa;`†§§c4r¦ä`C‡ru§§̈̆h4rvC»a§†qru§‡ayhʼC4ru§̆ʼ†‡ḧaru§‰CX†̈4rVyWW̆̈ru§;C»̆4rfĭ‡‡†h¤ru§;W†‡4rƒ̆‡Qru§D†̈†ha4rV̆̈:Caru§D†̈̆Xr?ay§̆;4rI†WC†‡rüC̆§r,§D†̈̆X4rfla‡¤ru§D†̈̆X4ruḧC†‡aru§D̆q4rVxa§†ruW†4rVCWa‡†ruWQ‡†̈4ruW†‡h†rm†;a‡†4r¦†;̆§§aruW†̈†‡;̆4r¦ẍCq;C‡̆ruW̆̈;C§4ruWC̈ru`hy§[†WCh4r6̆xhCruWC̈†`†hC4ruxW̆hruW§†4rVCWa‡ruWä4rVCWa‡ruWq;̆§§4ru‡rV†‡:s6C4ru‡r¦xa¤ay‡:4r̂̈Cq;C‡†ru‡†‰†y4rI†WC̆‡ru‡h̆̈qa‡4re†qa‡ru‡h̆̈qa‡r•`†q̆`†§§r‰§†¤̆̈c4rv̆C§ru‡h̆̈qa‡4rV»a;;¤ru‡h̆̈qa‡4r"C‡‡C̆ru‡h̆̈qa‡4rea‡†qru‡h̆̈qqa‡r •†̈»x̆̈¤c4ru‡ḧQ4ru‡ḧ̆ʻrfx¤W̆‡;4rfiCyq̆‰‰̆ru‡ḧ̆ʻq4r6̆:x†‡ru‡ḧ̆ʻq4rfl†;¤†‡†ru‡ḧC†‡aD†4rvC»a§†qru‡̆§K†4r6†̈Kru‡̆§§C4r¦†q‰̆̈ru‡K̆̈:̈̆‡4rƒC»Kru‡KC̆§4rI†D̆ru‡‡†`§̆4r[̆‡̈¤ru‡;»xay̆;4r¦ẍCqru‡;xa‡¤4reax‡r¦=ru‡;a‡Ca4ru§̆Lru‡;ä4r[C̈aoyWCrü†C4r¡̈‡̆q;arü†K†KC4ru‡ha‡Crü†‡†:†4r̃†y§rü†‡†q4r6CK̆§rü†‡`ÿy4rI†‡C̆§rü†‡Xy`Q†4reä:̆rü»̆4r6̆§Cqq†rü»x̆̈4r¡hhC̆rü:aq4rĬ‡CqrüKxC‰aD4ru‡ḧ̆Crü§aDqKC4re†qa‡rüWq;̆†h4rĬ̈C»KrüWq;̈a‡:4rfl¤§̆̈rü‡†qa‡4rI†‡C̆§rü‡̆oʼäh4r?̈̆‡;rü‡̆§4reax†‡rü‡̆‡:4re†C̈arü̈C̆;†4rm̆‡C‡rü̈a¤a4rƒa‡rü;̆q;4rV†ChrV†Coruq††h4ruq†K†r̂y`a4r„̈†̆‡†ruqx̆̈4r6†‡h¤ruqxoäh4re†W̆§ruqx§̆¤4ru†̈a‡ruqxWä̆4rVx†ʻ‡ruqXxWä̆4r?̈†xCWruq§ayW4r6†̈KruqiyC;x4r¡X̆iyC̆§ruq;†»Ca4r6Da‡haru;†‡:†‡†4rfi†̈̈̆;;ru;KC‡q4r¦ẍCqru;KC‡qa‡4rV†Wru;ʻ̆§§4rƒa§†‡hruyh̆‡̈C̆;x4ru§̆L†‡ḧ̆ruyh̆;4r6C†ruyhC‡†4rI†‡ruy̆̈`†»x4rey§C̆ruy:yq;¤‡C†K4r</sup>
RQ`CaruÿQ§Ca4rvC»a§̆ruyq;C‡4r¡WC§C̆ruy;yW‡4r̂䆤ruD»Q4r?̈C†‡ruD̆̈¤4rfC§§C†WruD̆̈¤r•`†qK̆;`†§§c4r¦†§̆`ruDC§Qq4r"§†h†ruD̈†WaD4rĕ†‡̆;;̆ruʻ4r¡̈C»ruL̆§qa‡4r6†̈;C‡ruL̆‡̈a;4ru¤hC§:̆4ruKC‡ru¤ah̆§̆4rf†§;̆̈ru¤aDQ4rv̆C§ru¤̈̆q4rn†aruXC†ʻa‡ay4rfĭä:̆qr?†4rnayqq̆or?†`†4rmy»†qr?†`C‡4rfiQ§†r?†`§y†‡C4rI†DChr?†»†‡C4r6ä̆‡†r?†»»†̈C‡4r[aʻ†̈hr?†»x4rea‡ar?†»a‡4rf†C‡̆r?†»a‡4rV††ʼChr?†h†;4ruxW̆hr?†h†¤4r¦̆X†̈r?†•hC;4†•4r?†̆rIaa‡†4r6†W†hayr?†:†¤aKa4rfiyq;†D̆r?†xaK̆‡4rV̆̈̆W†C†r?†C4rR̆̈‡†‡har?†C†‡a4r¦äC‡‡̆r?†C§̆¤rƒ†̆4r?aqqr?†C§̆¤4rƒah‡̆¤r?†C§̆¤4rfl†‡C†r?†C§̆¤4rV†‡ḧC‡̆r?†C§§¤4r6†̈C†‡ar?†C‡a;;C4ru‡h¤r?†C̈h4ru§̆Kq†‡h†̈r?†ʼ̆DqKC4r?̈†‡CWC̈r?†ʼC»Y4r?̈†‡Cq§†Dr?†ʼC»Y4r7§†;†‡r?†ʼ̈†WaDC»Y4r¦ẍCqr?†K̆̈r•oaa;`†§§r‰§†¤̆̈c4rI†‡C̆§ru=r?†K̆̈4r6†;;r?†K̆̈r•oaa;`†§§̆̈c4r6†ÿC»̆r?†K̆̈4reaqQre†DC̆̈r?†̈K̆̈a4rV†̈†xr?†§†`†:†‡4rVX†`a§»qr?†§†ʼ»X†4rI†‡C̆§r?†•§†‡4re†̈aq§†Dr

»†‡ x†D̆ oaa;‡a;̆q uqr§a‡:r†qr;x̆ oaa;‡a;̆ †‰‰̆†̈qr~{‹ ä

‹

o

̆

P

{

¤

a

?†§†Q;QK4rQCDC§̆=r?†§»•Cys‡†C;̆=4r„D†‡ar?†§C»Y4r̃C‡Cr?†§C§C4r6C»x†̆§r?†§§r•oaa;`†§§̆̈c4ru§̆qq†‡ḧar?†§§†‡

;

x

ʻ

†

̆

;

x

̈

a

y

_̆ o^a_a;‡^a

=

:

x

r

;

•RC:ÿ̆rZc>roärC‡oC‡C;̆r†Way‡;qraor;̆L;rqy»xr»ÿD̆qr‰̈ahy»̆r

x

;

̆

L

;

x

fl

†̆q;x̆;C»†§§¤r‰§̆†qC‡:r»aW‰§̆;̆§¤r`§†»Kr‰†:̆qr•RC:ÿ̆r‚c=

¦a‡»§yqCa‡q

¥‡̆‡»yW`̆̈̆hr `¤r ;x̆r »a‡q;̈†C‡;qr aor x†DC‡:r ;ar ‰̈ahy»̆r †r »a‡»Cq̆r
qyWW†̈¤r aor ;xCqr ʻäK4r ʻ̆r C‡q;̆†hr ̈̆‰̆†;r C;r C‡r C;qr
̆‡;C̈̆;¤rqar†qr‡a;r;ar§̆†D̆r†‡¤;xC‡:ray;=rrrrrrrrrr

[̆̈̆rCqr;x̆rW†C‡r`ah¤r;̆L;=

RC:ÿ̆qr~4rP4r‹=rR̈†»;†§r§†¤ay;raoroaa;‡a;̆qryqC‡: ;x̆rVC̆̈‰C‡qKCr¦ÿD̆=

?C`§Ca:̈†‰x¤

u»K‡aʻ§̆h:̆W̆‡;q

flx̆r †y;xär ʻCqx̆qr ;ar †»K‡aʻ§̆h:̆r †§§r ‰̆̈qa‡qr `ä‡rC‡r ;x̆r ¤̆†̈r ~5́54r
C‡r †§‰x†`̆;C»†§r äh̆̈tr "C;†§Cʼr u†`4r u†§C¤†x4r ¦†̈§;a‡r
u†̈a‡4r6aa‡CK†ru†D†4r6=r[†̈aa‡ru``†qrJ†W†̈4r̂x†§Chru`hy§§†xr•uW̆̈C»†‡roaa;`†§§r
‰§†¤̆̈c4ru`hy§§†xru`hy§i†hC̈†KxyW4rQ̈C»ru`Ch†§4rR†`̈C»̆ru`̈C̆§4r[a»C‡̆ru»xCay4r@§CD̆̈ru»K§†‡h4ruW¤ruh†Wqr•qC‡:̆̈c4r?§y̆r

~=r 6ÿ‰x¤4r flaW4r "„„>r 88fCKC‰§C†tr flx̆r R̈̆̆r ̃̈a:̈†WWC‡:r m†‡:y†:̆r
;x†;r u‡¤a‡̆r»†‡r¡hC;YY4rV„fi?@"„̂rPGǴ4rPGǴ=

P=rmy»†q4rfĭä:̆>r88V;†̈rf†̈qr6aDC̆YY4rflx̆r6aDC̆rflx̆†;̈̆4r~5́́
‹=rƒ̆̆h4re†qa‡>r88fiIr†‡hrmsq¤q;̆WqYY4rflx̆r̃̆̈§reaÿ‡†§4r"a§yW̆r‹4rvyW`̆̈r
~r•F5c4rV‰̈C‡:r~55l=

OUR RECENT PUB LABRICATIONS

Boosta Gwuposty McFoo, Ralph Boogyjive Wilkerson-Roo, and SLP
Wolverine

Disclaimer: Any resemblance to the STOC 2006 accepted papers list is
purely coincidental. (People are always trying to steal our work.)

Detractors for a Constant Barrage of Incremental Minimum
Independently Publishable Papers

Total-Ignorance against Deep Space Attacks

Narrow Minds May Be Spacious: Separating Space Between Ears in
Conflict Resolution Distance Trisecting for Curvaceous Lab
Assistants

[Boosta McFoo is sole author of this publication]

Dizziness Among Equally Inebriated Problem Solvers

The Simplicity of Producing a Cash Deficiencizium

The PCP Theorem By Hallucinogenic Mind-Amplification

(in memoriam of our mentor, Dr. Timothy Leary)

Unconditional Stiffness of Nearly and Really Actually Rather
Wantonly Coloring Yourself

Lab Assistant Problems and Norma-Jean Embeddings

Spare Time No Brain-Pan Divisions of Labor for Semi-Algorithmic
Implications of E.coli Superficial Untestable Audits of Profit Graphs:
It\'s All About The Benjamins, Baby Philosophic Phlegmatic
Pre-deconstructionist Printers

On Aquatic Performance Measures for Ping-Pong

Line-Dancing Exterminators and the Inability to Rep a Boot-Camp
Clique, Win Gangstaz, Influence Hustlaz, Determine Chromatic Numbers,
You Know, The Usual

On fertilizing random bonfires when utility functions are
subadditive Stopping Dynamite and Bomb Threats in a Space of Bounded
Jack Bauer Dimension Time-Wasting Strategies Including Powernapping

Fast Spider Detection and Squashing Protocols Especially Effective
in the Absence of High Heels

Convoluted Paths Leading to Nowhere but with Guaranteed Stomach
Digestion

A Algorithm for Finding Words When You Don\'t Want to Talk Too Much
but You Can\'t Help Yourself

Funding Small Unbalanced Dictators

Borderline bathing algorithms and bubbly soap foaming

Random mistakes in driver licensing and exponential communication
boundaries between genders

Getting up early is way too NP-hard

Gravity challenged midgets and diameter testing

Earthlings in the distance and minions laughing in 0-gravity

Logic is Hard, Learning is Hard, School Sucks So Let\'s Go Home

Explicit stomach-capacity-achieving bliss-inducible cakes

Novel gigantic bounds for rolling dice and the rolling eyes of the
Schroedinger cat A PQ-OMG-WTF-BBQ-TAS for Unspeakably Alphabetic Soup
Lines on Flow Graphs Teaching a Child by Injecting Values Via
Hardwood-Posterior Integration Simultaneous use of multiple commodity
shampoos can cause hair splitting Probably Tropical Socially-Biased
Algorithms for Stochastically Inverted Arubaizations

Minimizing the average growl time of unrelated monkeys

Uniformly influencing probabilistic corn planting

Time-Space Send-Offs for Missing Elderly Persons

Approximation of Colorful Liquor Stores, Closed to a Class of Odd
Minors Grocery Bidding Strategies for Discount Club Managers

Optimal unfairness: greedy resource concealing with undocumented
objectives The use of an electronic voting machine during blackouts

A Really Old Quantum Lower Bound, for Rejected Produce Theorems and
Time-Space Ripoffs Advances in Mattress and Bedding Theory

On the Perplexity of Ultra-Rapid Wine Sampling

Disinformationally Insecure Bosses and Loss of Composure under
Protocol Wardrop Equilibria: When Negotations Break Down, Wardrop
Bombs

On the Difficulty of Christmas Presents Delivery in Andromeda

Price-Optimal Sharing Of Hotel Rooms Among Multiple Graduate
Students of Various Genders

Catching epsilon-fish with epsilon-nets

Neutrality Gaps for Switzerland and Minimum Face Rearrangement
Problems Efficient Randomized Strategies to Climb down a tree

Public Humiliation and Search Problems at the TG

Cut Hardness Among Ginsu World Class Cleavers

Log Hardness and Directed Minimization of Chainsaw Motion

Deficient Proofs of Complete Ignorance

On Fourier tails, Plancherel paws, and Parseval arses

Somewhat Far-Out Neighbors and the Fast Sobering Transform

Theoretical Robitussin for Decongesting Sinus Games

Near-Optimal Algorithms for Very Special One-of-a-Kind
Non-Judgmental Games Pseudorealistic Waddles By Regulars at Dino\'s
Pub

Cost-Stealing Schemes for Multiple Publish-or-Perish Problems With
Monastic Steiner Trees

Quantum Closets Cannot Contain Extremely-Similar Outfits

Two-Timing Dispersers of n^{o(1)}*2^{sqrt{log n}}/polylog(n)
Entropy, and the Ghosts of Jon-Benet Ramsey, Frank Sinatra, and
Woodrow Wilson

A Quasi-Ethical Approximation Scheme for Minimum Weight Spousination
and Its Pitter Pattering Consequences

On the Potentially Unimportant Potency of Idempotent Impotence

A Randomized Hyphenated-Time Simplexified Algorithm for Mind
Deprogramming New Cloroxination Guarantee for the Platinum Dome
Problem

Philistine Disagreement: for as many rounds as you like and possibly
more in the future

On Facing One-Way Traffic in NP-Darkness

Finding a Maximum Weight Spouse in Really No Time at All, With
Applications

On the Resolution-Respace-Time Riemannian Regeometry of Randomly
Restrained Resatisficingicaction

Tooth Extraction On Small Face Sources

A Polyamorous Quantum Adventure, Starring the Amazing Indiana Jones
Polynomial A subset of spayed non-planar dogs approximate subset TSP

On Yinz\'s Local Ignorance

Hyperdesign of Hyperstructured Hyperlinks, Hyperoptimization, and
Hypertheoretical Hyperquestions