search

ehn-dcc-development / eu-dcc-schema

Schema for the ehn DCC payload
Apache License 2.0
164 stars 59 forks source link

Update country-2-codes.json #92

Closed kruzikh closed 3 years ago

ryanbnl commented 3 years ago

What is the context for the changes? The "why"? I would expect either an issue linked or something in the description :)

If I look at the other valuesets, they're not consistent: test-manf.json has a version with date, vaccine-prophylaxis.json a uri with a date.

It would be nice if they were consistent, or documentation was added to show others why.

Helps avoid Chesterton's Fence situations :-)

gabywh commented 3 years ago

It would be nice if they were consistent,

Hence standardize on e.g. semver - widely known / understood as a standard versioning mechanism

Helps avoid Chesterton's Fence situations :-)

Summarized as: "don't take down a fence until you know why it was put up"

Context known to author, just wasn't put in the description here

gabywh commented 3 years ago

As discussed 7 June 2021 for v 1.3.0: add non alpha-2 "pseudo" country codes UNHCR, WHO to this valueset

gabywh commented 3 years ago

I'm fine with correct process. Timescale even if points were made available in writing immediately?

Bear in mind: v1. 3.0 propsed for adoption tomorrow afternoon, complete with text and example around UNHCR and also stating value MUST come from the country-code valueset (as superset of ISO3166 alpha-2) as discussed / agreed yesterday in the meeting(s). If this value now not in the valueset for 1.3.0 then either doc broken or valueset broken (or both).

Hynek @.***> schrieb am Di., 8. Juni 2021, 12:28:

@.**** commented on this pull request.

In valuesets/country-2-codes.json https://github.com/ehn-dcc-development/ehn-dcc-schema/pull/92#discussion_r647314769 :

"valueSetValues": { "AD": { "display": "Andorra", "lang": "en", "active": true, "system": "urn:iso:std:iso:3166",

  • "valueSetId": "iso3166-1-2"
  • "version": "1.0"

I don't want to be too formal here, but we should in writing:

  1. Specify the business need, i.e. why (in which situation) we want to enable special entries in the value set.
  2. Specify which entries we want to add, i.e. code, display label, code system and version information. This we should send to SgS for acceptance.

— You are receiving this because your review was requested. Reply to this email directly, view it on GitHub https://github.com/ehn-dcc-development/ehn-dcc-schema/pull/92#discussion_r647314769, or unsubscribe https://github.com/notifications/unsubscribe-auth/ASHCPHKM27CDHR3JHOBND4DTRXWDBANCNFSM46CHB3GA .

martin-lindstrom commented 3 years ago

Let's not forget that the actual CWT as specified in https://github.com/ehn-dcc-development/hcert-spec/blob/main/hcert_spec.md also has an issuer, and the iss claim is still defined to be ISO 3166-1 alpha-2.

So, we should really update the hcert spec as well in order to handle the UNHCR, WHO or any other registered organization as an issuer.

And, what about the C RDN of the document signer certificate? Aren't people comparing the iss field with that?

kruzikh commented 3 years ago

Gaby, is there an urgent need to include UNHCR now (other then because it is included in documents)? Do we have any certificates or issuer that is planning to use it? I am just asking....

gabywh commented 3 years ago

Here is the text from v1.3.0:

Member State or third country in which the vaccine was administered Country expressed as a 2-letter ISO3166 code (RECOMMENDED) or a reference to an international organisation responsible for the vaccination event (such as UNHCR or WHO). This must be a coded value from the value set country-2-codes.json . The value set will be distributed from the EUDCC Gateway starting with the gateway version 1.1. Exactly 1 (one) field MUST be provided. Example: "co": "CZ" "co": "UNHCR"

Thus UNHCR and WHO shall be supported. Decision already made.

gabywh commented 3 years ago

Gaby, is there an urgent need to include UNHCR now (other then because it is included in documents)? Do we have any certificates or issuer that is planning to use it? I am just asking....

So yes there is an urgent need - as above:

If this value now not in the valueset for 1.3.0 then either doc broken or valueset broken (or both).

gabywh commented 3 years ago

Thus UNHCR and WHO shall be supported. Decision already made.

That said, I fully appreciate the preferred way of doing things is via CP to semantic sub-group: am happy to provide the text for the CP, but if the CP isn't going to go through and go through with a guaranteed yes (always a chance the CP is refused, of course, it is a proposal) by tomorrow afternoon then we have a problem.

Other option: we could of course take the text referring to UNCHR and WHO back out again, but including this text was already discussed and agreed en-masse on Monday, so the correct way to remove the text would be to schedule a new review of the v1.3.0 document and give people time to review it, so basically we need to re-review it now. I suspect that isn't going to happen either ;)

gabywh commented 3 years ago

Let's not forget that the actual CWT as specified in https://github.com/ehn-dcc-development/hcert-spec/blob/main/hcert_spec.md also has an issuer, and the iss claim is still defined to be ISO 3166-1 alpha-2.

So, we should really update the hcert spec as well in order to handle the UNHCR, WHO or any other registered organization as an issuer.

Yes, Issuer could well be UNHCR, WHO etc

And, what about the C RDN of the document signer certificate? Aren't people comparing the iss field with that?

I don't know what people are comparing here - verification business rules

gabywh commented 3 years ago

Gaby, is there an urgent need to include UNHCR now (other then because it is included in documents)?

So long as we stay with issuance within the Member States scheduled in the go-live events, then no: all with ISO3166 alpha-2 country code. However, it does mean docs and/or valueset are broken as it currently stands, which is surely reason enough to fix it?

gabywh commented 3 years ago

Let's not forget that the actual CWT as specified in https://github.com/ehn-dcc-development/hcert-spec/blob/main/hcert_spec.md also has an issuer, and the iss claim is still defined to be ISO 3166-1 alpha-2.

So, we should really update the hcert spec as well in order to handle the UNHCR, WHO or any other registered organization as an issuer.

In terms of keeping specs in sync: we also need to update the interop guidelines doc, specifically Annex 2 where all the useful implementation notes for UVCI somehow "dropped out" of the final version and which I had to pick up in my FAQ UVCI as a temporary measure.