1485 Need to supercede Tissue 1398 to clarify SCT behavior
- Created
- 22 Mar 2016
- Status
- Not ApplicableClosed: the issue falls outside the scope of the standard. No further comments are taken. Show the workflow
- Part
- Part 6 (2009-12; Edition 2)
- Links
- #1450 originalSclXxx computation rules
- Page
- Clause
- Paragraph
- Category
- Issue may impact interoperability of implementations of Edition 2
Issue
Tissue 1398 specifies that an SCT might set the IED Original SCL attributes based upon the information of Version and Revision at the SCL level of an ICD.
Tissue 1450 (now green) specifies that an ICT shall now provide these attributes at the IED level. This is an Interop issue.
In order to fully resolve the interoperabiity issue, 1398 needs to be superseded and the SCT needs to be required to set the IED Original SCL attributes, based upon the ICT's SCL Version and Revision attributes, if the IED level attributes are not present in the ICD.
Proposal
Make it mandatory that the SCT set the IED level Original SCL attributes if the ICD does not provide those attributes. The SCT shall set the IED attribute level values based upon the ICD's SCL Version and Revision attributes if the ICD does not provide the IED level attributes.
| Discussion | Created |
|---|---|
| Not Applicable This remaining IntOp issues between versions will be handled in an annex for mixed version handling | 14 Nov 16 |
|
Not Applicable
The interop issue remains as there are many Ed2 IxD files in use that do not contain the originalSclXxx attributes. In this case, the default value of 2003A is wrong and known to be wrong.
Because the schema does not require these attributes, it is reasonable to expect this to continue to be a problem. Additionally, as originalSclXxx are defined to be the SCL version of the IxD, they are redundant in that file. Any tool or device importing an IxD should interpret the IED as having these attributes set to the SCL version of the file - not necessarily 2003A. As there is continued disagreement please set this tissue to yellow for referral to WG10. |
29 Mar 16 |
| Not Applicable By giving the orgxxx attributes a default value (Tissue 1450), a value always exists. This means that also ICD/IID files having another SCL version as 2003A SHALL have these attributes with the appropriate version. Therefore for input from conforming (ICT) tools there is no need for an SCT to determine the version otherwise, i.e. Tissue 1398 is superfluous respectively might in some form be used to find non conforming ICTs - this is a tool specification issue. However, if the SCL header says something different than the org attributes, who is right? So, an SCT either prefers one (e.g.orgxxx), OR refuses the input. | 24 Mar 16 |
|
Discussion (red)
I think the missleading word in TISSUE #1398 is the "might" in the final proposal ("...the system configurator tool might determine it from the SCL version..."). In the UCA testing committee meeting we were in the opinion that this should be a "shall".
The TISSUE #1450 tries to fix it but the sentence "Syntactically both attributes stay optional but get the default values "2003" for the version and "A" for the revision." might be confusing. For ICD/CID/IID they are optional, but for an SCD they should be mandatory, right? So they should be conditional based on the SCL file type (if this is the correct wording ...?). |
22 Mar 16 |
| Not Applicable Tissue 1450 makes it already mandatory that the original SCL version attributes are filled by ICT and SCT. Any syntactically missing attribute defaults to version 2003 and revision A. | 22 Mar 16 |