Files
zeekay abf99d05fc papers: reorganize into per-paper dirs, add Antje Worring as co-author
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.
2026-04-18 21:50:07 -07:00

34 lines
1.7 KiB
TeX

\section{Related Work}
\label{sec:related}
Local-first software was articulated as a design philosophy by Kleppmann
et al.\ \cite{kleppmann2019localfirst}, who identified seven properties
of local-first applications: instant response, multi-device sync, offline
support, collaboration, data longevity, privacy, and user control. The
Lux capsule model satisfies all seven properties and adds cryptographic
enforcement (encryption, threshold access control, chain receipts) that
the original local-first paper described as desirable but did not
formalize.
CRDTs for collaborative editing have been implemented in Automerge
\cite{kleppmann2017automerge}, Yjs \cite{nicolaescu2015yjs}, and
Diamond Types \cite{josephg2023diamond}. The capsule data layer can
use any of these implementations, with the addition of encryption
for CRDT operations before broadcast.
Encrypted databases have been explored in CryptDB \cite{popa2011cryptdb}
(which supports queries over encrypted data via onion encryption) and
Mylar \cite{popa2014mylar} (which provides end-to-end encryption for web
applications). The capsule model takes a different approach: data is
decrypted locally and queries run over plaintext on the user's device,
avoiding the performance overhead and query limitations of encrypted
database schemes.
The Solid project \cite{bernerslee2016solid} proposed user-controlled
data pods as an alternative to platform-controlled data silos. Solid
uses Linked Data and RDF; the capsule model uses encrypted SQLite and
CRDTs. The key architectural difference is that Solid pods are
server-hosted (the pod provider can access data), while capsules are
encrypted and local-first (no provider can access plaintext data).