updates

5 files changed, 37 insertions, 4 deletions

diff --git a/chapters/app-reverse-cdf.tex b/appendices/reverse-cdf.tex
index 85542e1..85542e1 100644
--- a/chapters/app-reverse-cdf.tex
+++ b/appendices/reverse-cdf.tex
diff --git a/chapters/tail-latency.tex b/chapters/tail-latency.tex
index a96a897..9094e26 100644
--- a/chapters/tail-latency.tex
+++ b/chapters/tail-latency.tex
@@ -1,3 +1,4 @@
 \chapter{Controlling Insertion Tail Latency}
+\label{chap:tail-latency}
 
 \section{Introduction}
diff --git a/paper.tex b/paper.tex
index 98b857c..1a8600f 100644
--- a/paper.tex
+++ b/paper.tex
@@ -332,7 +332,7 @@ of Engineering Science and Mechanics
 
 % Generates the Acknowledgments. The argument should point to
 % the file containing your Acknowledgments. 
-% \thesisacknowledgments{chapters/acknowledgments}
+\thesisacknowledgments{supplementary/acknowledgments}
 
 % Generates the Epigraph/Dedication. The first argument should
 % point to the file containing your Epigraph/Dedication and
@@ -394,7 +394,7 @@ of Engineering Science and Mechanics
 % lines that redefine the \thechapter and \thesection:
 %\renewcommand\thechapter{}
 %\renewcommand\thesection{\arabic{section}}
-%\include{chapters/app-reverse-cdf}
+\input{appendices/reverse-cdf.tex}
 % \include{Appendix-B/Appendix-B}
 % \include{Appendix-C/Appendix-C}
 % \include{Appendix-D/Appendix-D}
diff --git a/supplementary/acknowledgments.tex b/supplementary/acknowledgments.tex
new file mode 100644
index 0000000..6086310
--- /dev/null
+++ b/supplementary/acknowledgments.tex
@@ -0,0 +1,31 @@
+Academic work is never done in a vaccuum, and I owe thanks to several
+people whose advice and feedback have been invaluable over the course
+of the work described in this dissertation.
+
+First, and most obviously, I owe a lot to Dr. Dong Xie for serving as
+my advisor.  We haven't always seen eye to eye on what the appropriate
+priorities of this work should be, and he has been very patient in the
+face of my grumbling.  While this project branched out into something
+far larger than the mere journal extension to his SIRS paper that he
+had originally envisioned, the original idea that spawned this line of
+work was entirely his.  On that note, thanks are also owed to Dr. Jeff
+Philips for originally pointing us in the direction of the dynamization
+work by Bentley, Saxe, and Overmars.
+
+Other academic collegues that are deserving of mention here are
+Dr. Ryan Stutsman, whose advice on techniques for representing
+tail latency distributions was absolutely critical to the work
+described in Chapter~\ref{chap:tail-latency}, and Dr. Zhuoyue Zhao,
+who collaborated directly with us on the work that ultimately became
+Chapter~\ref{chap:framework}. Chapter~\ref{chap:design-space} has been
+influenced, in part, by conversations with Dr. Niv Dayan.
+
+I also owe thanks to several of my fellow PhD students.  The fact that
+this document is even vaguely coherent is due in large part to Julia
+Custatis's attention to detail. Maxwell Norfolk and Jinyu Liu (as well
+as Julia) all got to suffer through being used as sounding boards for
+various assortments of my harebrained ideas. Along those same lines,
+Dr. Jack Sampson was instrumental to the maintenance of my sanity over
+these past few years.
+
+Thanks everyone. I could have never done this without you.
diff --git a/supplementary/nomenclature.tex b/supplementary/nomenclature.tex
index 01e0d48..15565a9 100644
--- a/supplementary/nomenclature.tex
+++ b/supplementary/nomenclature.tex
@@ -5,10 +5,11 @@
 	$N_B$ & Capacity of the mutable buffer \\ \hline
 	$s$   & Scale factor \\ \hline
 	$B(n)$ & Data structure construction cost from unsorted records \\ \hline
-	$B_M(n)$ & Data structure construction cost from merging existing instances\\ \hline
+	$B_M(n, k)$ & Data structure construction cost from merging $k$ existing instances\\ \hline
 	$L(n)$ & Point-lookup cost \\ \hline
 	$\mathscr{Q}(n)$ & Worst-case query cost over entire structure  \\ \hline
-	$\mathscr{Q}_S(n)$ & Worst-csae query cost over a single block/shard \\ \hline
+	$\mathscr{Q}_S(n)$ & Worst-case query cost over a single block/shard \\ \hline
+	$\mathscr{Q}_B(n)$ & Best-case query cost over entire structure \\ \hline
 	$I_A(n)$ & Amortized insertion cost \\ \hline
 	$I(n)$ & Worst-case insertion cost \\ \hline
 	$I_B(n)$ & Best-case insertion cost \\ \hline