> This was made worse by there being at least four hardware revisions (despite the company saying there were only two) with only some parts compatible across revisions.
To be completely honest, this sounds like a classic case of conflating hardware revision and configuration as it pertains to a managed configuration item. That you didn't have access to approved internal schematics, and technical knowledge of the black box was constrained to a defined external interface boundary prior to independent reverse engineering efforts is certainly consistent with that.
Class II ECPs consistent with MIL-HDBK-61/SAE EIA-649-1 and the system's documented objective maintenance plan are a thing. Surely the SMR code of this controller (more specifically, its maintenance repair and recoverability disposition indices as specified by AR 700-82 [1] that logistics activity would have assigned it) in the context of reported anecdote would be sufficient to infer an interesting story.
I certainly don't doubt that you encountered 4 different harware revisions, but it strikes me as much more likely that the maintenance plan for said system never intended internal hardware abstraction layers to be serviced by activity at your level, instead structured to be independently managed by contract activity driven by performance-based logistics incentives with the objective of ensuring that, from a black box behavioral perspective, what they reindoctrinated back into the supply system is tested to be operationally indistinguishable ("form, fit, function" as it were) and thus compliant with overarching performance specifications when integrated at the next higher assembly level.
To be completely honest, this sounds like a classic case of conflating hardware revision and configuration as it pertains to a managed configuration item. That you didn't have access to approved internal schematics, and technical knowledge of the black box was constrained to a defined external interface boundary prior to independent reverse engineering efforts is certainly consistent with that.
Class II ECPs consistent with MIL-HDBK-61/SAE EIA-649-1 and the system's documented objective maintenance plan are a thing. Surely the SMR code of this controller (more specifically, its maintenance repair and recoverability disposition indices as specified by AR 700-82 [1] that logistics activity would have assigned it) in the context of reported anecdote would be sufficient to infer an interesting story.
I certainly don't doubt that you encountered 4 different harware revisions, but it strikes me as much more likely that the maintenance plan for said system never intended internal hardware abstraction layers to be serviced by activity at your level, instead structured to be independently managed by contract activity driven by performance-based logistics incentives with the objective of ensuring that, from a black box behavioral perspective, what they reindoctrinated back into the supply system is tested to be operationally indistinguishable ("form, fit, function" as it were) and thus compliant with overarching performance specifications when integrated at the next higher assembly level.
[1] https://armypubs.army.mil/epubs/DR_pubs/DR_a/ARN30303-AR_700...