Re: [media-types] [Anima] do we need +jose?

[Orie Steele <[e-mail]>]

This draft might change those requirements, it could use more reviews:

https://datatracker.ietf.org/doc/draft-ietf-mediaman-suffixes/

I personally think we should not assume every structured suffix will also
be a registered sub type.

I think we should:

1. Move the normative statements to a document and out of the registry.
2. Update the registry policy to have specification required.
3. Detail the processing guidelines from right to left, in a way that does
not create exceptions.
4. If 4 is impossible, document the exception and take steps to avoid them
in the future.



On Thu, May 11, 2023, 5:07 AM Esko Dijk <[e-mail]> wrote:

> Update to my previous email: as I learnt now the registration of a +suffix
> in the SSS registry doesn't require that the registered name is an existing
> media type name. (Example: +der)
> So +jws could be registered with the registration fields pointing to RFC
> 7515 "application/jose+json" as the reference.
>
> It could also be named +josejson or +jose-json then ? Not as nice as +jws
> but at least more relatable to the original media type name.
>
> Esko
>
> -----Original Message-----
> From: Esko Dijk
> Sent: Wednesday, May 10, 2023 16:38
> To: Michael Richardson <[e-mail]>; [e-mail];
> [e-mail]; [e-mail]
> Subject: RE: [Anima] do we need +jose?
>
> > should really be doing:
> >        application/voucher+jws
>
> Because "application/jws" does not seem to be an existing media type, it
> would be strange to use "+jws".
> Looking at draft-ietf-anima-jws-voucher-06: what it really uses is the
> "JWS JSON Serialization" which has the "application/jose+json" media type.
> This is not the "application/jose" type, so it would be strange to use
> "+jose" as your subject suggests.
> Now given that we shouldn't use multiple structured syntax suffixes in
> concatenation at this moment, the only option for the suffix media type at
> this moment looks to be "+json".
>
> (Or alternatively we would need a new spec that defines the
> "application/jws" media type - not advisable it seems, adds to confusion.)
>
> So we can have names like e.g.:
>
>      application/voucher-jose+json
>      application/voucher-jws+json
>
> In the cases above the "+json" at the end isn't wrong, because it actually
> is JSON.  (For the earlier case of "application/voucher-cms+json" it was
> wrong as you say, because the CMS envelope isn't actually JSON.)
>
> Esko
>
>
> IoTconsultancy.nl  |  Email/Teams: [e-mail]
>
>
> -----Original Message-----
> From: Anima <[e-mail]> On Behalf Of Michael Richardson
> Sent: Tuesday, May 9, 2023 20:51
> To: [e-mail]; [e-mail]; [e-mail]
> Subject: [Anima] do we need +jose?
>
>
> Hi, https://datatracker.ietf.org/doc/draft-ietf-anima-jws-voucher/
> is in WGLC, and
> https://datatracker.ietf.org/doc/draft-ietf-anima-brski-prm/ depends upon
> it.
>
> In anima-jws-voucher, we defined:
>
> https://www.ietf.org/archive/id/draft-ietf-anima-jws-voucher-06.html#name-application-voucher-jwsjson
>
>         Type name:  application
>         Subtype name:  voucher-jws+json
>
> which is in alignment with
> https://www.rfc-editor.org/rfc/rfc8366.html#section-8.3
> where we defined:
>       Type name:  application
>       Subtype name:  voucher-cms+json
>
> probably this was a mistake!  (JSON in a CMS envelope)
>
> I think, based upon discussion about +cose and our other documents, that we
> should really be doing:
>        application/voucher+jws
>
> While jwt is given as a structured suffix in the IANA registry, jws is not.
> I'm not entirely sure if this matters... we are dealing with JWS, not
> tokens...
>
> Please advise.  While we have lots of running code (since 2018) for
> voucher-jws, it's a
> change we could probably make via Postel Principal.
>
> --
> Michael Richardson <[e-mail]>   . o O ( IPv6 IøT consulting )
>            Sandelman Software Works Inc, Ottawa and Worldwide
> _______________________________________________
> media-types mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/media-types
>