You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
REPLY_RAW format seemingly at odds with IEC 60839-11-5:2020 #319
Describe the bug
Looking at the code, I can see the reply format for RAW card reads.
However, this does not match IEC 60839-11-5:2020 which we are working against.
Expected behavior
For example, the IEC standard shows 2 bytes for the reply data size, whereas you only have 1.
I see that the SIA spec is 2.2.2 published in 2024.
So is the IEC standard now "out-of-date"?
If so, shouldn't some backward compatibility be required?
Hi @mpfj, Can you please post a screenshot of what you are looking at? I checked v2.2 and v2.1.5 (predates the IEC version) of the spec and both seems to agree (with code) that data can be arbitrary length.
Also "whereas you only have 1" -- where are you seeing this?
I now realise I'm not necessarily looking at the same thing but here's the IEC definition of the RAW card reading reply:
And here's the definition from osdp.h:
I guess the struct isn't a direct byte-for-byte mapping.
Regardless, my reply packets always seem to be missing the first 2 card data bytes and was just checking to see if I've missed something.
It's possible my testing tool ("OSDP Bench" from the Windows Store) might be wrong so was hoping I could just get confirmation that the IEC reply format is still valid.
Activity
sidcha commentedon Jul 6, 2026
Hi @mpfj, Can you please post a screenshot of what you are looking at? I checked v2.2 and v2.1.5 (predates the IEC version) of the spec and both seems to agree (with code) that data can be arbitrary length.
Also "whereas you only have 1" -- where are you seeing this?
mpfj commentedon Jul 6, 2026
I now realise I'm not necessarily looking at the same thing but here's the IEC definition of the RAW card reading reply:
And here's the definition from osdp.h:
I guess the struct isn't a direct byte-for-byte mapping.
Regardless, my reply packets always seem to be missing the first 2 card data bytes and was just checking to see if I've missed something.
It's possible my testing tool ("OSDP Bench" from the Windows Store) might be wrong so was hoping I could just get confirmation that the IEC reply format is still valid.
Cheers
sidcha commentedon Jul 7, 2026
LibOSDP exposed API (osdp.h) is not the same as the wire OSDP contract. Have you tried the LibOSDP python CP example https://doc.osdp.dev/python/getting-started#control-panel-mode to cross check the results?