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
Peer restart-drain coordination controller: ready / restart commit · case 01
While ready, event restart_commit produces running/resume_old instead of stopped/close_old_session.
Peer restart-drain coordination controller: ready / restart commit · case 02
While ready, event restart_commit produces running/resume_old instead of stopped/close_old_session.
Peer restart-drain coordination controller: ready / restart commit · case 03
While ready, event restart_commit produces running/resume_old instead of stopped/close_old_session.
Peer restart-drain coordination controller: ready / restart commit · case 04
While ready, event restart_commit produces running/resume_old instead of stopped/close_old_session.
Peer restart-drain coordination controller: ready / restart commit · case 05
While ready, event restart_commit produces running/resume_old instead of stopped/close_old_session.
Peer restart-drain coordination controller: ready / restart cancel · case 01
While ready, event restart_cancel produces stopped/terminate instead of running/resume_admission.
Peer restart-drain coordination controller: ready / restart cancel · case 02
While ready, event restart_cancel produces stopped/terminate instead of running/resume_admission.
Peer restart-drain coordination controller: ready / restart cancel · case 03
While ready, event restart_cancel produces stopped/terminate instead of running/resume_admission.
Peer restart-drain coordination controller: ready / restart cancel · case 04
While ready, event restart_cancel produces stopped/terminate instead of running/resume_admission.
Peer restart-drain coordination controller: ready / restart cancel · case 05
While ready, event restart_cancel produces stopped/terminate instead of running/resume_admission.
Peer restart-drain coordination controller: stopped / new session · case 01
While stopped, event new_session produces running/resume_admission instead of synchronizing/exchange_generation.
Peer restart-drain coordination controller: stopped / new session · case 02
While stopped, event new_session produces running/resume_admission instead of synchronizing/exchange_generation.
Peer restart-drain coordination controller: stopped / new session · case 03
While stopped, event new_session produces running/resume_admission instead of synchronizing/exchange_generation.
Peer restart-drain coordination controller: stopped / new session · case 04
While stopped, event new_session produces running/resume_admission instead of synchronizing/exchange_generation.
Peer restart-drain coordination controller: stopped / new session · case 05
While stopped, event new_session produces running/resume_admission instead of synchronizing/exchange_generation.
Peer restart-drain coordination controller: synchronizing / generation agreed · case 01
While synchronizing, event generation_agreed produces synchronizing/wait instead of running/resume_admission.
Peer restart-drain coordination controller: synchronizing / generation agreed · case 02
While synchronizing, event generation_agreed produces synchronizing/wait instead of running/resume_admission.
Peer restart-drain coordination controller: synchronizing / generation agreed · case 03
While synchronizing, event generation_agreed produces synchronizing/wait instead of running/resume_admission.
Peer restart-drain coordination controller: synchronizing / generation agreed · case 04
While synchronizing, event generation_agreed produces synchronizing/wait instead of running/resume_admission.
Peer restart-drain coordination controller: synchronizing / generation agreed · case 05
While synchronizing, event generation_agreed produces synchronizing/wait instead of running/resume_admission.
Peer restart-drain coordination controller: draining / new work · case 01
While draining, event new_work produces draining/accept_work instead of draining/redirect_work.
Peer restart-drain coordination controller: draining / new work · case 02
While draining, event new_work produces draining/accept_work instead of draining/redirect_work.
Peer restart-drain coordination controller: draining / new work · case 03
While draining, event new_work produces draining/accept_work instead of draining/redirect_work.
Peer restart-drain coordination controller: draining / new work · case 04
While draining, event new_work produces draining/accept_work instead of draining/redirect_work.
Peer restart-drain coordination controller: draining / new work · case 05
While draining, event new_work produces draining/accept_work instead of draining/redirect_work.
Peer restart-drain coordination controller: running / restart commit · case 01
While running, event restart_commit produces stopped/terminate instead of running/reject_unprepared.
Peer restart-drain coordination controller: running / restart commit · case 02
While running, event restart_commit produces stopped/terminate instead of running/reject_unprepared.
Peer restart-drain coordination controller: running / restart commit · case 03
While running, event restart_commit produces stopped/terminate instead of running/reject_unprepared.
Peer restart-drain coordination controller: running / restart commit · case 04
While running, event restart_commit produces stopped/terminate instead of running/reject_unprepared.
Peer restart-drain coordination controller: running / restart commit · case 05
While running, event restart_commit produces stopped/terminate instead of running/reject_unprepared.
Peer restart-drain coordination controller: synchronizing / old session frame · case 01
While synchronizing, event old_session_frame produces running/deliver instead of synchronizing/reject_old_generation.
Peer restart-drain coordination controller: synchronizing / old session frame · case 02
While synchronizing, event old_session_frame produces running/deliver instead of synchronizing/reject_old_generation.
Peer restart-drain coordination controller: synchronizing / old session frame · case 03
While synchronizing, event old_session_frame produces running/deliver instead of synchronizing/reject_old_generation.
Peer restart-drain coordination controller: synchronizing / old session frame · case 04
While synchronizing, event old_session_frame produces running/deliver instead of synchronizing/reject_old_generation.
Peer restart-drain coordination controller: synchronizing / old session frame · case 05
While synchronizing, event old_session_frame produces running/deliver instead of synchronizing/reject_old_generation.
Peer restart-drain coordination controller: stopped / duplicate commit · case 01
While stopped, event duplicate_commit produces running/resume_old instead of stopped/ack_stopped.
Peer restart-drain coordination controller: stopped / duplicate commit · case 02
While stopped, event duplicate_commit produces running/resume_old instead of stopped/ack_stopped.
Peer restart-drain coordination controller: stopped / duplicate commit · case 03
While stopped, event duplicate_commit produces running/resume_old instead of stopped/ack_stopped.
Peer restart-drain coordination controller: stopped / duplicate commit · case 04
While stopped, event duplicate_commit produces running/resume_old instead of stopped/ack_stopped.
Peer restart-drain coordination controller: stopped / duplicate commit · case 05
While stopped, event duplicate_commit produces running/resume_old instead of stopped/ack_stopped.
Asynchronous command cancellation controller: idle / submit · case 01
While idle, event submit produces running/execute instead of queued/enqueue.
Asynchronous command cancellation controller: idle / submit · case 02
While idle, event submit produces running/execute instead of queued/enqueue.
Asynchronous command cancellation controller: idle / submit · case 03
While idle, event submit produces running/execute instead of queued/enqueue.
Asynchronous command cancellation controller: idle / submit · case 04
While idle, event submit produces running/execute instead of queued/enqueue.
Asynchronous command cancellation controller: idle / submit · case 05
While idle, event submit produces running/execute instead of queued/enqueue.
Asynchronous command cancellation controller: queued / dispatch · case 01
While queued, event dispatch produces done/report_success instead of running/start_command.
Asynchronous command cancellation controller: queued / dispatch · case 02
While queued, event dispatch produces done/report_success instead of running/start_command.
Asynchronous command cancellation controller: queued / dispatch · case 03
While queued, event dispatch produces done/report_success instead of running/start_command.
Asynchronous command cancellation controller: queued / dispatch · case 04
While queued, event dispatch produces done/report_success instead of running/start_command.
Asynchronous command cancellation controller: queued / dispatch · case 05
While queued, event dispatch produces done/report_success instead of running/start_command.
Asynchronous command cancellation controller: queued / cancel · case 01
While queued, event cancel produces running/start_command instead of cancelled/remove_and_ack.
Asynchronous command cancellation controller: queued / cancel · case 02
While queued, event cancel produces running/start_command instead of cancelled/remove_and_ack.
Asynchronous command cancellation controller: queued / cancel · case 03
While queued, event cancel produces running/start_command instead of cancelled/remove_and_ack.
Asynchronous command cancellation controller: queued / cancel · case 04
While queued, event cancel produces running/start_command instead of cancelled/remove_and_ack.
Asynchronous command cancellation controller: queued / cancel · case 05
While queued, event cancel produces running/start_command instead of cancelled/remove_and_ack.
Asynchronous command cancellation controller: running / cancel · case 01
While running, event cancel produces cancelled/ack_cancel instead of cancelling/request_stop.
Asynchronous command cancellation controller: running / cancel · case 02
While running, event cancel produces cancelled/ack_cancel instead of cancelling/request_stop.
Asynchronous command cancellation controller: running / cancel · case 03
While running, event cancel produces cancelled/ack_cancel instead of cancelling/request_stop.
Asynchronous command cancellation controller: running / cancel · case 04
While running, event cancel produces cancelled/ack_cancel instead of cancelling/request_stop.
Asynchronous command cancellation controller: running / cancel · case 05
While running, event cancel produces cancelled/ack_cancel instead of cancelling/request_stop.
Asynchronous command cancellation controller: cancelling / stopped · case 01
While cancelling, event stopped produces done/report_success instead of cancelled/ack_cancel.
Asynchronous command cancellation controller: cancelling / stopped · case 02
While cancelling, event stopped produces done/report_success instead of cancelled/ack_cancel.
Asynchronous command cancellation controller: cancelling / stopped · case 03
While cancelling, event stopped produces done/report_success instead of cancelled/ack_cancel.
Asynchronous command cancellation controller: cancelling / stopped · case 04
While cancelling, event stopped produces done/report_success instead of cancelled/ack_cancel.
Asynchronous command cancellation controller: cancelling / stopped · case 05
While cancelling, event stopped produces done/report_success instead of cancelled/ack_cancel.
Asynchronous command cancellation controller: cancelling / completed first · case 01
While cancelling, event completed_first produces cancelled/ack_cancel instead of done/report_result.
Asynchronous command cancellation controller: cancelling / completed first · case 02
While cancelling, event completed_first produces cancelled/ack_cancel instead of done/report_result.
Asynchronous command cancellation controller: cancelling / completed first · case 03
While cancelling, event completed_first produces cancelled/ack_cancel instead of done/report_result.
Asynchronous command cancellation controller: cancelling / completed first · case 04
While cancelling, event completed_first produces cancelled/ack_cancel instead of done/report_result.
Asynchronous command cancellation controller: cancelling / completed first · case 05
While cancelling, event completed_first produces cancelled/ack_cancel instead of done/report_result.
Asynchronous command cancellation controller: running / completed · case 01
While running, event completed produces running/wait instead of done/report_result.
Asynchronous command cancellation controller: running / completed · case 02
While running, event completed produces running/wait instead of done/report_result.
Asynchronous command cancellation controller: running / completed · case 03
While running, event completed produces running/wait instead of done/report_result.
Asynchronous command cancellation controller: running / completed · case 04
While running, event completed produces running/wait instead of done/report_result.
Asynchronous command cancellation controller: running / completed · case 05
While running, event completed produces running/wait instead of done/report_result.
Asynchronous command cancellation controller: done / cancel · case 01
While done, event cancel produces cancelled/ack_cancel instead of done/report_already_done.
Asynchronous command cancellation controller: done / cancel · case 02
While done, event cancel produces cancelled/ack_cancel instead of done/report_already_done.
Asynchronous command cancellation controller: done / cancel · case 03
While done, event cancel produces cancelled/ack_cancel instead of done/report_already_done.
Asynchronous command cancellation controller: done / cancel · case 04
While done, event cancel produces cancelled/ack_cancel instead of done/report_already_done.
Asynchronous command cancellation controller: done / cancel · case 05
While done, event cancel produces cancelled/ack_cancel instead of done/report_already_done.
Asynchronous command cancellation controller: cancelled / late completion · case 01
While cancelled, event late_completion produces done/report_result instead of cancelled/discard_late_result.
Asynchronous command cancellation controller: cancelled / late completion · case 02
While cancelled, event late_completion produces done/report_result instead of cancelled/discard_late_result.
Asynchronous command cancellation controller: cancelled / late completion · case 03
While cancelled, event late_completion produces done/report_result instead of cancelled/discard_late_result.
Asynchronous command cancellation controller: cancelled / late completion · case 04
While cancelled, event late_completion produces done/report_result instead of cancelled/discard_late_result.
Asynchronous command cancellation controller: cancelled / late completion · case 05
While cancelled, event late_completion produces done/report_result instead of cancelled/discard_late_result.
Asynchronous command cancellation controller: idle / cancel · case 01
While idle, event cancel produces cancelled/ack_cancel instead of idle/report_unknown.
Asynchronous command cancellation controller: idle / cancel · case 02
While idle, event cancel produces cancelled/ack_cancel instead of idle/report_unknown.
Asynchronous command cancellation controller: idle / cancel · case 03
While idle, event cancel produces cancelled/ack_cancel instead of idle/report_unknown.
Asynchronous command cancellation controller: idle / cancel · case 04
While idle, event cancel produces cancelled/ack_cancel instead of idle/report_unknown.
Asynchronous command cancellation controller: idle / cancel · case 05
While idle, event cancel produces cancelled/ack_cancel instead of idle/report_unknown.
Delimited batch command controller: idle / begin · case 01
While idle, event begin produces committed/apply instead of collecting/create_batch.
Delimited batch command controller: idle / begin · case 02
While idle, event begin produces committed/apply instead of collecting/create_batch.
Delimited batch command controller: idle / begin · case 03
While idle, event begin produces committed/apply instead of collecting/create_batch.
Delimited batch command controller: idle / begin · case 04
While idle, event begin produces committed/apply instead of collecting/create_batch.
Delimited batch command controller: idle / begin · case 05
While idle, event begin produces committed/apply instead of collecting/create_batch.
Delimited batch command controller: collecting / operation · case 01
While collecting, event operation produces collecting/apply_operation instead of collecting/stage_operation.
Delimited batch command controller: collecting / operation · case 02
While collecting, event operation produces collecting/apply_operation instead of collecting/stage_operation.
Delimited batch command controller: collecting / operation · case 03
While collecting, event operation produces collecting/apply_operation instead of collecting/stage_operation.
Delimited batch command controller: collecting / operation · case 04
While collecting, event operation produces collecting/apply_operation instead of collecting/stage_operation.
Delimited batch command controller: collecting / operation · case 05
While collecting, event operation produces collecting/apply_operation instead of collecting/stage_operation.
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 ↗