pending does not tell me whether the FR10 verified this bracket

MeiBaker0516 · 18 Aug 2026, 11:49 UTC

Reply to discussion
ME
MeiBaker0516
Our application crashed after sending a fixture-verification request and before saving the result. FR10 may have continued while the local file stayed pending, then startup offered to resend it. I'm separating never-submitted from outcome-unknown, but the restart decision needs evidence behind those words, not just a better label.

14 replies

JO
JoAli0244
Replying to MeiBaker0516

What durable identity survives for this particular request on both sides?

6 points
ME
MeiBaker0516
Replying to JoAli0244

The application kept a unique request identity before sending, and the receiver retains a result under that identity. Our startup helper ignored both and only checked pending. I've held its automatic resend while we review reconciliation.

3 points
KA
KaiBrooks0802
Replying to MeiBaker0516

Check what the retained result means under the supported interface agreement before treating an identity match as completed verification.

10 points
JA
JamieBrooks0817
Replying to MeiBaker0516

And make the held startup state visible while that lookup is running. Starting reconciliation should not briefly make the bracket appear ready for another request.

-2 points
HA
HazelChan1093
Replying to MeiBaker0516

Could the same identifier be reused after reinstalling or restoring the application? I've seen a familiar number look much more convincing than its record deserved. Worth checking the identity lifetime before trusting a retained result.

24 points
ME
MeiBaker0516
Replying to HazelChan1093

The two software owners checked identity lifetime and the retained response contract. This result belongs to the original request and records completed verification. We can reconcile this case without resending; uncertain or missing identities will remain held.

12 points
JO
JoAli0244
Replying to MeiBaker0516

Has that conclusion been saved durably, or is it only displayed after the lookup?

15 points
ME
MeiBaker0516
Replying to JoAli0244

Saved and linked to the retained response, with no new verification request sent. The developer is now testing failure while saving that reconciliation so a second startup cannot treat it as untouched work.

18 points
KA
KaiBrooks0802
Replying to MeiBaker0516

Keep the original interrupted local record in the audit history; reconciliation explains it rather than making the crash disappear.

12 points
JA
JamieBrooks0817
Replying to MeiBaker0516

Does a failed reconciliation lookup leave a useful reason and escalation route, not just an indefinite spinner? The next shift shouldn't have to work out whether waiting longer is the approved response.

7 points
ME
MeiBaker0516
Replying to JamieBrooks0817

The offline tests now show held with a reason after failed lookup or failed save, and a saved reconciliation survives restart. The original request history is intact. Unknown cases have a named review route rather than a retry shortcut.

7 points
HA
HazelChan1093
Replying to MeiBaker0516

That answers my identity concern for this case. Is the normal startup launcher using this version too, or are you still exercising the development copy?

8 points
ME
MeiBaker0516
Replying to HazelChan1093

Still the development copy. Normal-launch verification and the covering-maintainer exercise are the remaining tasks. The old bracket result is reconciled, but I'm leaving the general startup change short of release until those checks are done.

3 points
JO
JoAli0244
Replying to MeiBaker0516

Please return with the normal-launch result. A recovered request and a released startup fix are different outcomes.

7 points

Add to the discussion

Welcome to Application Robot

Everyone can read the forum. Sign in or create an account to start a discussion, reply, or upload photos.

Forgot your password?

By creating an account, you agree to our Terms and Conditions and community guidelines. Read our Privacy Policy for how your information is handled.