File comparison glossary

Diff

A diff is the set of differences between two files or two versions of the same file.

The name comes from the Unix diff utility (1974). Modern diff tools go beyond line-by-line text comparison to understand document and data structure.

Redline

A redline is a marked-up view of a document showing every insertion and deletion between two versions, traditionally in red.

Common in legal and contract review: two drafts go in, one annotated document comes out, so a reviewer can read the change instead of hunting for it.

Semantic diff

A semantic diff compares two files by their meaning rather than their raw text, ignoring differences that don't change what the file says.

For JSON, reordered keys or changed indentation are not real changes; a semantic diff reports only actual value and structure changes.

Structural comparison

Structural comparison matches up the parts of two documents — paragraphs, tables, cells, or elements — and reports differences part by part.

For a Word document this means comparing paragraphs, tables, and formatting as such, instead of flattening everything to plain text first.

Side-by-side view

A side-by-side view shows both file versions in two synchronized panes with the differences highlighted in place.

Unified view

A unified view interleaves both versions in a single pane, marking removed and added content inline.

Tracked changes

Tracked changes are edits recorded inside a Word document itself, each attributed to an author and awaiting acceptance or rejection.

Tracked changes are stored in the file; a diff is computed between two files. A comparison tool can produce a redline even when nobody remembered to turn tracking on.

Checksum / hash

A checksum (or hash) is a short fingerprint computed from a file's content; two files with different checksums are guaranteed to differ.

Folder comparison tools use content hashes to verify whether same-named files are truly identical without comparing every byte twice.

Perceptual image comparison

Perceptual image comparison judges whether two images look the same to a human, rather than whether their bytes are identical.

A re-saved or resized photo has completely different bytes but looks identical; perceptual methods (and image embeddings) catch that.

Folder comparison

Folder comparison walks two directory trees and reports which files were added, removed, or changed between them.

Three-way merge

A three-way merge combines two changed versions of a file using their common ancestor to decide which changes to keep.

A concept from version control. Comparison tools like Diffesq report differences between two files; they don't merge or modify them.

Client-side processing

Client-side processing means the work happens on your own device, in your browser, instead of on a remote server.

This is how Diffesq handles every comparison — see About for what that means in practice.

Put the terms to work

Ready to see a diff instead of reading about one? Compare two files — free for text and Markdown, no account.