mirror of
https://github.com/lux-apps/papers.git
synced 2026-10-30 18:15:21 +00:00
zeekay
abf99d05fc
Restructured from flat .tex/.pdf files into per-paper subdirectories. Added Antje Worring as co-author on all papers. Synced PDFs to pdfs/ for download site.
46 lines
2.0 KiB
TeX
46 lines
2.0 KiB
TeX
\section{Discussion}
|
|
|
|
\label{sec:discussion}
|
|
%% ============================================================================
|
|
|
|
\subsection{When Quasar is the Right Choice}
|
|
|
|
Quasar is optimal for applications requiring:
|
|
\begin{itemize}
|
|
\item Sub-second deterministic finality with post-quantum guarantees.
|
|
\item Moderate validator sets (21--200 validators) where O($n$) message complexity
|
|
dominates O($n^2$) alternatives.
|
|
\item Bridge and cross-chain operations where finality speed is the bottleneck.
|
|
\item Regulatory environments that will require post-quantum cryptography (CNSA~2.0
|
|
timeline: 2030--2035).
|
|
\end{itemize}
|
|
|
|
\subsection{When Other Protocols Win}
|
|
|
|
\begin{itemize}
|
|
\item \textbf{Maximum raw throughput}: Narwhal+Tusk or Solana, if finality latency
|
|
$>$1\,s is acceptable.
|
|
\item \textbf{Maximum decentralization ($n > 1{,}000$)}: Snow-family protocols, which
|
|
degrade gracefully with validator count at the cost of probabilistic finality.
|
|
\item \textbf{Existing ecosystem}: Ethereum PoS for applications where 12.8\,min
|
|
finality is acceptable and the validator set ($>$1M stakers) provides economic
|
|
security beyond BFT thresholds.
|
|
\end{itemize}
|
|
|
|
\subsection{Threats to Validity}
|
|
|
|
\begin{enumerate}
|
|
\item \textbf{Execution excluded}: Real-world throughput is lower than consensus-layer
|
|
TPS due to EVM/VM execution overhead. Quasar's DEX precompiles mitigate this
|
|
for financial operations.
|
|
\item \textbf{Adversarial conditions}: All benchmarks used honest validators. Byzantine
|
|
behavior (equivocation, withholding) would increase latency for all protocols.
|
|
\item \textbf{Network conditions}: Fixed inter-DC RTT does not capture real-world jitter,
|
|
packet loss, or BGP route changes.
|
|
\item \textbf{Software maturity}: CometBFT and Ethereum clients have years of production
|
|
optimization. Quasar's implementation is newer and may have optimization
|
|
headroom.
|
|
\end{enumerate}
|
|
|
|
%% ============================================================================
|