v0.9.3 — Source Variorum (GPL-3.0-or-later)

Source Variorum is a side-by-side textual collation workbench and the first instrument in the Computational Hermeneutics family. It brings the apparatus of textual criticism (witness, siglum, copy-text, apparatus criticus, transposition) to computational close reading of both source code and prose.

This is the first tagged release, and the first under the GPL-3.0-or-later licence. Source Variorum was previously released under MIT; as sole copyright holder, David M. Berry has relicensed the project, including all prior versions, under GPL-3.0-or-later.

Features

• Two engines: CX-Engine (code, line/token-aware) and TX-Engine (text, sentence/word-aware) with witness sigla.
• Three-column braid of Bezier ribbons showing matches, additions, deletions, substitutions, and block transpositions; braids default off at rest (panel tints and the overview strip still show every change).
• Editorial workflow: Auto | User modes, editable braid (approve/doubt/tentative/swap/delete) persisted as an editorial layer, word-level token refinement, and manual hand-braiding with precise drag-select boundaries (CollateX-style).
• Syntax highlighting: modern languages via Lezer (JS/TS, Python, HTML, CSS, C/C++, Java, Rust, Go, PHP, SQL, JSON, XML, Markdown) and historical dialects (PDP-1, AGC, MAD, FORTRAN, IPL-V, BASIC).
• Analysis: stemma / distance tree (UPGMA), dotplot, change-overview histogram, VVV-style vertical overview strip, witness statistics, deep-dive metrics.
• Export: TEI P5 parallel-segmentation, plus Markdown / PDF / JSON and the native .svar project format.
• Samples: full 25-source Spacewar! corpus, three-edition Othello, and the Phaedrus in two public-domain translations.

Licence

GNU General Public License v3.0 or later. See LICENSE.