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
Blackhole route terminates lookup instead of falling back · case 01
Blackhole route terminates lookup instead of falling back.
Blackhole route terminates lookup instead of falling back · case 02
Blackhole route terminates lookup instead of falling back.
Blackhole route terminates lookup instead of falling back · case 03
Blackhole route terminates lookup instead of falling back.
Blackhole route terminates lookup instead of falling back · case 04
Blackhole route terminates lookup instead of falling back.
Blackhole route terminates lookup instead of falling back · case 05
Blackhole route terminates lookup instead of falling back.
Send window is minimum of peer and congestion windows · case 01
Send window is minimum of peer and congestion windows.
Send window is minimum of peer and congestion windows · case 02
Send window is minimum of peer and congestion windows.
Send window is minimum of peer and congestion windows · case 03
Send window is minimum of peer and congestion windows.
Send window is minimum of peer and congestion windows · case 04
Send window is minimum of peer and congestion windows.
Send window is minimum of peer and congestion windows · case 05
Send window is minimum of peer and congestion windows.
Acknowledged byte count excludes already acknowledged prefix · case 01
Acknowledged byte count excludes already acknowledged prefix.
Acknowledged byte count excludes already acknowledged prefix · case 02
Acknowledged byte count excludes already acknowledged prefix.
Acknowledged byte count excludes already acknowledged prefix · case 03
Acknowledged byte count excludes already acknowledged prefix.
Acknowledged byte count excludes already acknowledged prefix · case 04
Acknowledged byte count excludes already acknowledged prefix.
Acknowledged byte count excludes already acknowledged prefix · case 05
Acknowledged byte count excludes already acknowledged prefix.
Receive window overlap clips both segment ends · case 01
Receive window overlap clips both segment ends.
Receive window overlap clips both segment ends · case 02
Receive window overlap clips both segment ends.
Receive window overlap clips both segment ends · case 03
Receive window overlap clips both segment ends.
Receive window overlap clips both segment ends · case 04
Receive window overlap clips both segment ends.
Receive window overlap clips both segment ends · case 05
Receive window overlap clips both segment ends.
Fin consumes one sequence number · case 01
Fin consumes one sequence number.
Fin consumes one sequence number · case 02
Fin consumes one sequence number.
Fin consumes one sequence number · case 03
Fin consumes one sequence number.
Fin consumes one sequence number · case 04
Fin consumes one sequence number.
Fin consumes one sequence number · case 05
Fin consumes one sequence number.
Syn sequence space is independent of payload · case 01
Syn sequence space is independent of payload.
Syn sequence space is independent of payload · case 02
Syn sequence space is independent of payload.
Syn sequence space is independent of payload · case 03
Syn sequence space is independent of payload.
Syn sequence space is independent of payload · case 04
Syn sequence space is independent of payload.
Syn sequence space is independent of payload · case 05
Syn sequence space is independent of payload.
Sequence ordering handles modular wraparound · case 01
Sequence ordering handles modular wraparound.
Sequence ordering handles modular wraparound · case 02
Sequence ordering handles modular wraparound.
Sequence ordering handles modular wraparound · case 03
Sequence ordering handles modular wraparound.
Sequence ordering handles modular wraparound · case 04
Sequence ordering handles modular wraparound.
Sequence ordering handles modular wraparound · case 05
Sequence ordering handles modular wraparound.
Serial distance is forward modular distance · case 01
Serial distance is forward modular distance.
Serial distance is forward modular distance · case 02
Serial distance is forward modular distance.
Serial distance is forward modular distance · case 03
Serial distance is forward modular distance.
Serial distance is forward modular distance · case 04
Serial distance is forward modular distance.
Serial distance is forward modular distance · case 05
Serial distance is forward modular distance.
Duplicate ack counter resets on forward progress · case 01
Duplicate ack counter resets on forward progress.
Duplicate ack counter resets on forward progress · case 02
Duplicate ack counter resets on forward progress.
Duplicate ack counter resets on forward progress · case 03
Duplicate ack counter resets on forward progress.
Duplicate ack counter resets on forward progress · case 04
Duplicate ack counter resets on forward progress.
Duplicate ack counter resets on forward progress · case 05
Duplicate ack counter resets on forward progress.
Fast retransmit fires on threshold crossing once · case 01
Fast retransmit fires on threshold crossing once.
Fast retransmit fires on threshold crossing once · case 02
Fast retransmit fires on threshold crossing once.
Fast retransmit fires on threshold crossing once · case 03
Fast retransmit fires on threshold crossing once.
Fast retransmit fires on threshold crossing once · case 04
Fast retransmit fires on threshold crossing once.
Fast retransmit fires on threshold crossing once · case 05
Fast retransmit fires on threshold crossing once.
Retransmitted segments do not supply ambiguous rtt samples · case 01
Retransmitted segments do not supply ambiguous rtt samples.
Retransmitted segments do not supply ambiguous rtt samples · case 02
Retransmitted segments do not supply ambiguous rtt samples.
Retransmitted segments do not supply ambiguous rtt samples · case 03
Retransmitted segments do not supply ambiguous rtt samples.
Retransmitted segments do not supply ambiguous rtt samples · case 04
Retransmitted segments do not supply ambiguous rtt samples.
Retransmitted segments do not supply ambiguous rtt samples · case 05
Retransmitted segments do not supply ambiguous rtt samples.
Rtt estimator weights old and new samples · case 01
Rtt estimator weights old and new samples.
Rtt estimator weights old and new samples · case 02
Rtt estimator weights old and new samples.
Rtt estimator weights old and new samples · case 03
Rtt estimator weights old and new samples.
Rtt estimator weights old and new samples · case 04
Rtt estimator weights old and new samples.
Rtt estimator weights old and new samples · case 05
Rtt estimator weights old and new samples.
Rto includes variation and clock granularity · case 01
Rto includes variation and clock granularity.
Rto includes variation and clock granularity · case 02
Rto includes variation and clock granularity.
Rto includes variation and clock granularity · case 03
Rto includes variation and clock granularity.
Rto includes variation and clock granularity · case 04
Rto includes variation and clock granularity.
Rto includes variation and clock granularity · case 05
Rto includes variation and clock granularity.
Retransmission timeout doubles with upper cap · case 01
Retransmission timeout doubles with upper cap.
Retransmission timeout doubles with upper cap · case 02
Retransmission timeout doubles with upper cap.
Retransmission timeout doubles with upper cap · case 03
Retransmission timeout doubles with upper cap.
Retransmission timeout doubles with upper cap · case 04
Retransmission timeout doubles with upper cap.
Retransmission timeout doubles with upper cap · case 05
Retransmission timeout doubles with upper cap.
Congestion loss threshold has two segment floor · case 01
Congestion loss threshold has two segment floor.
Congestion loss threshold has two segment floor · case 02
Congestion loss threshold has two segment floor.
Congestion loss threshold has two segment floor · case 03
Congestion loss threshold has two segment floor.
Congestion loss threshold has two segment floor · case 04
Congestion loss threshold has two segment floor.
Congestion loss threshold has two segment floor · case 05
Congestion loss threshold has two segment floor.
Zero window probe remains permitted with no send credit · case 01
Zero window probe remains permitted with no send credit.
Zero window probe remains permitted with no send credit · case 02
Zero window probe remains permitted with no send credit.
Zero window probe remains permitted with no send credit · case 03
Zero window probe remains permitted with no send credit.
Zero window probe remains permitted with no send credit · case 04
Zero window probe remains permitted with no send credit.
Zero window probe remains permitted with no send credit · case 05
Zero window probe remains permitted with no send credit.
Nagle buffering applies only with outstanding data and small write · case 01
Nagle buffering applies only with outstanding data and small write.
Nagle buffering applies only with outstanding data and small write · case 02
Nagle buffering applies only with outstanding data and small write.
Nagle buffering applies only with outstanding data and small write · case 03
Nagle buffering applies only with outstanding data and small write.
Nagle buffering applies only with outstanding data and small write · case 04
Nagle buffering applies only with outstanding data and small write.
Nagle buffering applies only with outstanding data and small write · case 05
Nagle buffering applies only with outstanding data and small write.
Delayed ack is forced by second full segment or timer · case 01
Delayed ack is forced by second full segment or timer.
Delayed ack is forced by second full segment or timer · case 02
Delayed ack is forced by second full segment or timer.
Delayed ack is forced by second full segment or timer · case 03
Delayed ack is forced by second full segment or timer.
Delayed ack is forced by second full segment or timer · case 04
Delayed ack is forced by second full segment or timer.
Delayed ack is forced by second full segment or timer · case 05
Delayed ack is forced by second full segment or timer.
Send buffer removes only cumulatively acknowledged segments · case 01
Send buffer removes only cumulatively acknowledged segments.
Send buffer removes only cumulatively acknowledged segments · case 02
Send buffer removes only cumulatively acknowledged segments.
Send buffer removes only cumulatively acknowledged segments · case 03
Send buffer removes only cumulatively acknowledged segments.
Send buffer removes only cumulatively acknowledged segments · case 04
Send buffer removes only cumulatively acknowledged segments.
Send buffer removes only cumulatively acknowledged segments · case 05
Send buffer removes only cumulatively acknowledged segments.
Out of order receive does not advance cumulative ack over gap · case 01
Out of order receive does not advance cumulative ack over gap.
Out of order receive does not advance cumulative ack over gap · case 02
Out of order receive does not advance cumulative ack over gap.
Out of order receive does not advance cumulative ack over gap · case 03
Out of order receive does not advance cumulative ack over gap.
Out of order receive does not advance cumulative ack over gap · case 04
Out of order receive does not advance cumulative ack over gap.
Out of order receive does not advance cumulative ack over gap · case 05
Out of order receive does not advance cumulative ack over gap.
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 ↗