Understand the failure.
Verify the repair.
Small, reproducible software failures. The broken implementation, the fix that didn’t work, and the one that passed—preserved together.
Explore the cases ↓How results are verified ↗WHAT THE ARCHIVE CONTAINS
100840 executable cases. 20168 are open.
Every case records the implementation that fails, the fix that did not work, and the repair that passed its checks—with recorded outputs and source hashes. This release adds 100840 cases across 20168 failure mechanisms and 254 domains.
The open tier gives you the failure and the unsuccessful fix for one case in every mechanism. The remaining 80672 cases, 5 variants per mechanism, are member-only: the verified repair, its recorded checks, and the full fixture suite are held in the member archive. Read the methodology ↗
A RECORD OF WHAT WENT WRONG
Browse the archive / 100840
A titlecase digraph is reduced to an uppercase digraph · case 01
A titlecase digraph is reduced to an uppercase digraph.
A titlecase digraph is reduced to an uppercase digraph · case 02
A titlecase digraph is reduced to an uppercase digraph.
A titlecase digraph is reduced to an uppercase digraph · case 03
A titlecase digraph is reduced to an uppercase digraph.
A titlecase digraph is reduced to an uppercase digraph · case 04
A titlecase digraph is reduced to an uppercase digraph.
A titlecase digraph is reduced to an uppercase digraph · case 05
A titlecase digraph is reduced to an uppercase digraph.
Capitalization leaves an uppercase tail · case 01
Capitalization leaves an uppercase tail.
Capitalization leaves an uppercase tail · case 02
Capitalization leaves an uppercase tail.
Capitalization leaves an uppercase tail · case 03
Capitalization leaves an uppercase tail.
Capitalization leaves an uppercase tail · case 04
Capitalization leaves an uppercase tail.
Capitalization leaves an uppercase tail · case 05
Capitalization leaves an uppercase tail.
An uppercase expansion is truncated during case swap · case 01
An uppercase expansion is truncated during case swap.
An uppercase expansion is truncated during case swap · case 02
An uppercase expansion is truncated during case swap.
An uppercase expansion is truncated during case swap · case 03
An uppercase expansion is truncated during case swap.
An uppercase expansion is truncated during case swap · case 04
An uppercase expansion is truncated during case swap.
An uppercase expansion is truncated during case swap · case 05
An uppercase expansion is truncated during case swap.
Alphabetic filtering admits combining letter numbers · case 01
Alphabetic filtering admits combining letter numbers.
Alphabetic filtering admits combining letter numbers · case 02
Alphabetic filtering admits combining letter numbers.
Alphabetic filtering admits combining letter numbers · case 03
Alphabetic filtering admits combining letter numbers.
Alphabetic filtering admits combining letter numbers · case 04
Alphabetic filtering admits combining letter numbers.
Alphabetic filtering admits combining letter numbers · case 05
Alphabetic filtering admits combining letter numbers.
A zero-class mark survives mark stripping · case 01
A zero-class mark survives mark stripping.
A zero-class mark survives mark stripping · case 02
A zero-class mark survives mark stripping.
A zero-class mark survives mark stripping · case 03
A zero-class mark survives mark stripping.
A zero-class mark survives mark stripping · case 04
A zero-class mark survives mark stripping.
A zero-class mark survives mark stripping · case 05
A zero-class mark survives mark stripping.
Invisible format controls survive control filtering · case 01
Invisible format controls survive control filtering.
Invisible format controls survive control filtering · case 02
Invisible format controls survive control filtering.
Invisible format controls survive control filtering · case 03
Invisible format controls survive control filtering.
Invisible format controls survive control filtering · case 04
Invisible format controls survive control filtering.
Invisible format controls survive control filtering · case 05
Invisible format controls survive control filtering.
Fullwidth letters occupy one display column · case 01
Fullwidth letters occupy one display column.
Fullwidth letters occupy one display column · case 02
Fullwidth letters occupy one display column.
Fullwidth letters occupy one display column · case 03
Fullwidth letters occupy one display column.
Fullwidth letters occupy one display column · case 04
Fullwidth letters occupy one display column.
Fullwidth letters occupy one display column · case 05
Fullwidth letters occupy one display column.
Character count underestimates an encoded field · case 01
Character count underestimates an encoded field.
Character count underestimates an encoded field · case 02
Character count underestimates an encoded field.
Character count underestimates an encoded field · case 03
Character count underestimates an encoded field.
Character count underestimates an encoded field · case 04
Character count underestimates an encoded field.
Character count underestimates an encoded field · case 05
Character count underestimates an encoded field.
An astral character is counted as one UTF-16 unit · case 01
An astral character is counted as one UTF-16 unit.
An astral character is counted as one UTF-16 unit · case 02
An astral character is counted as one UTF-16 unit.
An astral character is counted as one UTF-16 unit · case 03
An astral character is counted as one UTF-16 unit.
An astral character is counted as one UTF-16 unit · case 04
An astral character is counted as one UTF-16 unit.
An astral character is counted as one UTF-16 unit · case 05
An astral character is counted as one UTF-16 unit.
UTF-16 BOM is interpreted as ordinary text · case 01
UTF-16 BOM is interpreted as ordinary text.
UTF-16 BOM is interpreted as ordinary text · case 02
UTF-16 BOM is interpreted as ordinary text.
UTF-16 BOM is interpreted as ordinary text · case 03
UTF-16 BOM is interpreted as ordinary text.
UTF-16 BOM is interpreted as ordinary text · case 04
UTF-16 BOM is interpreted as ordinary text.
UTF-16 BOM is interpreted as ordinary text · case 05
UTF-16 BOM is interpreted as ordinary text.
Latin-1 bytes are decoded through a typographic code page · case 01
Latin-1 bytes are decoded through a typographic code page.
Latin-1 bytes are decoded through a typographic code page · case 02
Latin-1 bytes are decoded through a typographic code page.
Latin-1 bytes are decoded through a typographic code page · case 03
Latin-1 bytes are decoded through a typographic code page.
Latin-1 bytes are decoded through a typographic code page · case 04
Latin-1 bytes are decoded through a typographic code page.
Latin-1 bytes are decoded through a typographic code page · case 05
Latin-1 bytes are decoded through a typographic code page.
Windows punctuation decodes as controls · case 01
Windows punctuation decodes as controls.
Windows punctuation decodes as controls · case 02
Windows punctuation decodes as controls.
Windows punctuation decodes as controls · case 03
Windows punctuation decodes as controls.
Windows punctuation decodes as controls · case 04
Windows punctuation decodes as controls.
Windows punctuation decodes as controls · case 05
Windows punctuation decodes as controls.
Invalid UTF-8 bytes silently disappear · case 01
Invalid UTF-8 bytes silently disappear.
Invalid UTF-8 bytes silently disappear · case 02
Invalid UTF-8 bytes silently disappear.
Invalid UTF-8 bytes silently disappear · case 03
Invalid UTF-8 bytes silently disappear.
Invalid UTF-8 bytes silently disappear · case 04
Invalid UTF-8 bytes silently disappear.
Invalid UTF-8 bytes silently disappear · case 05
Invalid UTF-8 bytes silently disappear.
Undecodable bytes cannot be re-encoded losslessly · case 01
Undecodable bytes cannot be re-encoded losslessly.
Undecodable bytes cannot be re-encoded losslessly · case 02
Undecodable bytes cannot be re-encoded losslessly.
Undecodable bytes cannot be re-encoded losslessly · case 03
Undecodable bytes cannot be re-encoded losslessly.
Undecodable bytes cannot be re-encoded losslessly · case 04
Undecodable bytes cannot be re-encoded losslessly.
Undecodable bytes cannot be re-encoded losslessly · case 05
Undecodable bytes cannot be re-encoded losslessly.
Odd hex nybbles become a complete byte · case 01
Odd hex nybbles become a complete byte.
Odd hex nybbles become a complete byte · case 02
Odd hex nybbles become a complete byte.
Odd hex nybbles become a complete byte · case 03
Odd hex nybbles become a complete byte.
Odd hex nybbles become a complete byte · case 04
Odd hex nybbles become a complete byte.
Odd hex nybbles become a complete byte · case 05
Odd hex nybbles become a complete byte.
Removing a prefix eats a set of leading characters · case 01
Removing a prefix eats a set of leading characters.
Removing a prefix eats a set of leading characters · case 02
Removing a prefix eats a set of leading characters.
Removing a prefix eats a set of leading characters · case 03
Removing a prefix eats a set of leading characters.
Removing a prefix eats a set of leading characters · case 04
Removing a prefix eats a set of leading characters.
Removing a prefix eats a set of leading characters · case 05
Removing a prefix eats a set of leading characters.
Removing a suffix erases unrelated trailing characters · case 01
Removing a suffix erases unrelated trailing characters.
Removing a suffix erases unrelated trailing characters · case 02
Removing a suffix erases unrelated trailing characters.
Removing a suffix erases unrelated trailing characters · case 03
Removing a suffix erases unrelated trailing characters.
Removing a suffix erases unrelated trailing characters · case 04
Removing a suffix erases unrelated trailing characters.
Removing a suffix erases unrelated trailing characters · case 05
Removing a suffix erases unrelated trailing characters.
An assignment value loses embedded equals signs · case 01
An assignment value loses embedded equals signs.
An assignment value loses embedded equals signs · case 02
An assignment value loses embedded equals signs.
An assignment value loses embedded equals signs · case 03
An assignment value loses embedded equals signs.
An assignment value loses embedded equals signs · case 04
An assignment value loses embedded equals signs.
An assignment value loses embedded equals signs · case 05
An assignment value loses embedded equals signs.
A qualified label splits at its first colon · case 01
A qualified label splits at its first colon.
A qualified label splits at its first colon · case 02
A qualified label splits at its first colon.
A qualified label splits at its first colon · case 03
A qualified label splits at its first colon.
A qualified label splits at its first colon · case 04
A qualified label splits at its first colon.
A qualified label splits at its first colon · case 05
A qualified label splits at its first colon.
Whitespace tokenization creates empty tokens · case 01
Whitespace tokenization creates empty tokens.
Whitespace tokenization creates empty tokens · case 02
Whitespace tokenization creates empty tokens.
Whitespace tokenization creates empty tokens · case 03
Whitespace tokenization creates empty tokens.
Whitespace tokenization creates empty tokens · case 04
Whitespace tokenization creates empty tokens.
Whitespace tokenization creates empty tokens · case 05
Whitespace tokenization creates empty tokens.
INSPECTABLE BY DESIGN
Every result has a runnable source.
Runnable implementations with recorded outputs, source hashes, and explicit contracts. Related variants share a failure mechanism and belong together in evaluation splits.
Read the methodology ↗