Compare two projects?
Author: kaihendryCreated Apr 13, 2026Updated Apr 15, 2026
Use case is that given two projects, which can be objectively be said to be "lighter weight"?
gh repo clone openai/codex
gh repo clone badlogic/pi-monoThis output is way too noisy:
lightweight $ sloc pi-mono/packages/coding-agent/
───────────────────────────────────────────────────────────────────────────────
Language Files Lines Blanks Comments Code Complexity
───────────────────────────────────────────────────────────────────────────────
TypeScript 317 80,008 9,951 7,149 62,908 9,296
Markdown 57 14,732 3,781 0 10,951 0
JSON 13 833 12 0 821 0
JavaScript 4 2,930 212 203 2,515 339
JSONL 2 2,022 0 0 2,022 0
Shell 2 245 31 20 194 18
C 1 72 12 12 48 4
CSS 1 1,001 163 23 815 0
HTML 1 55 4 3 48 0
───────────────────────────────────────────────────────────────────────────────
Total 398 101,898 14,166 7,410 80,322 9,657
───────────────────────────────────────────────────────────────────────────────
Estimated Cost to Develop (organic) $2,701,914
Estimated Schedule Effort (organic) 20.07 months
Estimated People Required (organic) 11.96
───────────────────────────────────────────────────────────────────────────────
Processed 6922718 bytes, 6.923 megabytes (SI)
───────────────────────────────────────────────────────────────────────────────
lightweight $ scc codex/
───────────────────────────────────────────────────────────────────────────────
Language Files Lines Blanks Comments Code Complexity
───────────────────────────────────────────────────────────────────────────────
Rust 1,435 634,264 53,662 20,476 560,126 30,491
TypeScript 457 7,421 1,204 2,265 3,952 228
Plain Text 443 7,271 699 0 6,572 0
JSON 228 89,016 3 0 89,013 0
Markdown 154 14,783 3,575 0 11,208 0
TOML 111 4,137 450 258 3,429 19
Bazel 102 1,887 223 51 1,613 35
Python 82 22,017 3,568 418 18,031 2,180
YAML 40 4,751 492 203 4,056 0
Patch 29 2,702 146 0 2,556 0
Shell 29 3,151 428 280 2,443 357
SQL 26 273 21 0 252 0
Meson 7 385 36 6 343 30
C 6 5,945 955 645 4,345 1,525
JavaScript 6 2,169 207 112 1,850 319
C Header 5 297 39 56 202 7
SVG 5 17 0 0 17 0
License 4 718 112 0 606 0
Powershell 3 699 118 70 511 124
XML 3 663 13 0 650 0
Dockerfile 2 86 16 13 57 13
HTML 2 320 5 0 315 0
Nix 2 143 8 10 125 3
Smarty Template 2 112 16 0 96 35
Autoconf 1 1 0 1 0 0
Docker ignore 1 1 0 0 1 0
Emacs Lisp 1 1 0 0 1 0
JSONL 1 8,041 0 0 8,041 0
Jupyter 1 587 0 0 587 0
───────────────────────────────────────────────────────────────────────────────
Total 3,188 811,858 65,996 24,864 720,998 35,366
───────────────────────────────────────────────────────────────────────────────
Estimated Cost to Develop (organic) $27,066,125
Estimated Schedule Effort (organic) 48.17 months
Estimated People Required (organic) 49.92
───────────────────────────────────────────────────────────────────────────────
Processed 28626595 bytes, 28.627 megabytes (SI)
───────────────────────────────────────────────────────────────────────────────I also feel the cost to develop metric in light of the AI agent stuff makes no sense. --no-cocomo should be the default. Listing the languages also arguably doesn't matter.
Source: boyter/scc