When should you use it?
- You have old source code you would like to run again
- The original tools are gone
- You can test in an emulator
Not for: Commercial code you do not own; Code with no source
What do you need, and what do you get?
You need
- Old source files
- A compatible modern assembler
- An emulator
You get
- Compiling utilities and program
- Notes on fixes made
- Bugs found from comments
How does it work, step by step?
- Put it in gitOld code often has no version control; start one now.
- Pick toolsChoose an assembler compatible with the original and an emulator.
- Build small firstGet utilities compiling before the main program.
- Fix clashesAsk the agent to look for names that later became reserved words.
- Hunt old bugsPoint it at comments that describe bugs and ask where they could be.
The prompt
Our own prompt, inspired by TacticalCoder on Hacker News and written to work in Claude Code. Values in lime are placeholders.
Fill in the blanks (optional; the copied prompt uses your values)
Help me compile this old {language} code with {assembler} and run it in {emulator}. First, make a git repo and do not modify the originals folder. Build the small utilities first. When builds fail, check for names that later became reserved words. Then read comments that mention bugs and point to where each bug might be.
Inspired by TacticalCoder on Hacker News: Comment in Revision Demoparty 2026: Razor1911 [video]. The prompt above is written by playbooks.bot in our own words; nothing from the original is quoted.
What can go wrong?
- Huge single files strain context; work file by file.
- Old floppy copies may be damaged; back them up first.
- Keep original files untouched in their own folder.
Our take
TacticalCoder's project shows why old code is such good territory for coding agents. The models have absorbed decades of manuals, forum posts and assembler quirks that most of us have forgotten, so an error that would once cost an evening, like a macro named incbin colliding with a keyword added years later, is identified at once. Order of work matters here. Get the small utilities compiling and running first, as the author did with the old image and sprite utilities, because each success confirms the toolchain and emulator settings before you tackle the big program. Commit the original files to git before changing anything; in 1991, version control usually meant copying folders, and you want an untouched baseline to compare against when something behaves strangely. Old comments describing known bugs are a surprisingly good task list, so feed them to the agent and ask for a fix with an explanation. Keep each change small and test it in the emulator before moving on, since assembly errors rarely fail politely. Above all, enjoy it; reviving lost work is one of the most rewarding uses of these tools.
Where the time goes back: Old projects come back to life in hours instead of weekends of reading manuals.
FAQ
What tool did it use?
What was the tricky bug?
Source and credit
Inspired by TacticalCoder on Hacker News. We describe the idea in our own words, link to the original, and wrote the prompt ourselves; nothing is quoted.
- Inspired by
- Comment in Revision Demoparty 2026: Razor1911 [video]
- Creator
- TacticalCoder on Hacker News
- Published on
- Hacker News
- Original date
- Apr 8, 2026
- License
- Not openly licensed (the poster keeps all rights), so we quote nothing and link to the original
- Originally for
- Claude Code
- AI dry run
- Oct 23, 2026: passed. Dry-run tested by our AI test agent on sample input. PASS: Gates held on desk check; no live sends, bookings or purchases.
AI dry run on sample data; not run with live connected accounts. The playbooks.bot test agent filled the prompt with synthetic inputs, produced the output and checked it against the page's stated outputs and guardrails. - Verified
- Oct 23, 2026: source link resolves to the original, the credited creator published it, our excerpt or prompt matches it, and the license allows our use.
Are you the creator? Claim or correct this page.