Test Procedure Working Groups » IEC 61850 TPWG » Test Procedure Issues » Server

"or missing completely" means that datset="" is not allowed in SCL. Just the datset attribute shall be missing. Not the complete GSEControl.

About missing communication section, this is on purpose when the datset is not configured. I agree when a datset is configured also GSE shall be present.

TPWG: When an IED has GSEsetting name=fix the ICT will accept GOCB with missing dataset. When IED has GSEsetting name=conf the IED/ICT may refuse the complete configuration with a GOCB without dataset. In that case it's expected that the SCT tool removes the not-used GOCBs. when necessary the SCT can add a new GoCB again.

It's not clearly defined in part 6 if such configuration refusal is allowed or not.

As mentioned by Richard, in case of fix property, the SCT can not add or remove the control block, while in case of conf, the SCT is responsible for adding / removing the cntrol block.
According to 8-1, the absence of the GSE resp. SMV communication element from the <ConnectedAP > indicate the ICT that the control block is disabled within the SCD.

fix control blocks are the only control block where an empty datSet attribute makes sense. Part 6 considers complete / consistent configuration - those are captured in the SCT conformance tests. Removing a ressource is the way to deal with flexible (conf) control blocks when ther are no longer used, and not half / inconsistent configurations.

The linked tissue has nothing to do with the discussed issue of missing datSet attribute in the control block.

Based on Richard and Thierry's explanation, I think step 1 should be clarified that the DUT may refuse the configuration when datSet is missing ONLY if GSESettings cbName="Conf". If cbName="fix", then basically the DUT created the extra control blocks (not the SCT). It would be ridiculous for it to refuse a configuration that it caused. I found at least 2 cases like this in the 2019 IOP scd files.

This topic is straying from the original purpose.

This has absolutely nothing to do with what people think is ridiculous or whether there are other ways disable GOOSE transmission. It only concerns what is specified in the standards.

sGop9 clearly says "DUT is configured with a GSEControl element without the datSet" with one possible conformant response "DUT refuses the entire configuration".
But 61850-7-2:2010+AMD1:2020 Table 87 says (approximately) "if GOCB.DatSet is NULL then SetGoCBValues(GoEna,true) shall fail" and "the value of GoEna at IED startup depends on the IED
configuration." To me, this appears to be (at least one) proscribed method to inhibit GOOSE publication.

I interpret this as "if GSEControl@datSet is missing then GoEna at IED startup is false".

But sGop9 allows the publisher to refuser the configuration and not just "use the configuration but do not publish that specific GOOSE".
My suggestion is to change expected result sGop9 from "DUT (including IED tool) either refuses the entire configuration or it ignores parts of the new configuration or it accepts the
configuration" to "DUT (including IED tool) ignores parts of the new configuration or it accepts the configuration"