[T2TRG] Overrides vs. default (Re: [Asdf] T2TRG/WISHI/ASDF/CoRE hackathon @ IETF111: Tuesday 20th 1400Z)

[Carsten Bormann <[e-mail]>]

[narrowed to t2trg/asdf]

When ingesting and processing SDF, the processing may need the resolved result of using sdfRef.
Beside the namespace juggling, there also is the possibility that the named qualities map that contains the sdfRef also has overrides [1].

[1]: https://www.ietf.org/archive/id/draft-ietf-asdf-sdf-07.html#section-4.4-5

I ran a quick script against the playground to find out what those overrides might be.
I found a lot of places that have “label” overrides that are totally redundant with the name of the named quality the sdfRef is used in.
I think we may want to reconsider if this redundant information should be in those references; we’d have mostly clean (non-overridden) references otherwise.  I think I’d rather get rid of “label” than have these redundant entries.

The other recurring pattern is providing a “default” override.  Now that is very similar to your typical constructor parameter:  This essentially provides an initial value to the instance.  So far, this is used in only two models:

DEFAULT: 0 Zigbee Alliance ./sdfObject/sdfobject-level.sdf.json
DEFAULT: 0 Zigbee Alliance ./sdfObject/sdfobject-level.sdf.json
DEFAULT: [] Zigbee Alliance ./sdfObject/sdfobject-level.sdf.json
DEFAULT: 0 Zigbee Alliance ./sdfObject/sdfobject-onoff.sdf.json
DEFAULT: 0 Zigbee Alliance ./sdfObject/sdfobject-onoff.sdf.json

Maybe we don’t want to lump instantiation parameters like this in the general override mechanism?

It is also a bit unclear to me what “default” actually means; the spec is a bit terse here, only saying “specifies the default value for initialization”.  Here we are again, “initialization”!

All the “default” qualities are on named sdfProperties; as these are essentially instance variables of the containing sdfObject, the analogy to “initialization” actually does make sense.  It is less clear what the default quality would mean for, say, sdfOutputData, or, worse, nested in compound data types or choices buried in dataqualities.

Grüße, Carsten


> On 2021-07-19, at 18:39, Carsten Bormann <[e-mail]> wrote:
> 
> Today, we scheduled the next daily meeting for
> 
> https://codimd.ietf.org/hackathon-2021-t2trg-111-tuesday
> start time: 2021-07-20 14:00:00 UTC max duration: 60
> (We needed 45 min today.)
> 
> https://www.timeanddate.com/worldclock/fixedtime.html?msg=IETF111+hackathon+T2TRG-WISHI-ASDF-CoRE&iso=20210720T14&ah=1
> 
> https://meetings.conf.meetecho.com/interim/?short=d49f0f3b-52d5-4a30-8a75-e2f2effec8b0
> 
> Notes from today’s meeting are at https://codimd.ietf.org/hackathon-2021-t2trg-111-monday
> 
> 
> Grüße, Carsten
> 
> 
>> On 2021-07-18, at 21:26, Carsten Bormann <[e-mail]> wrote:
>> 
>> For the IETF 111 hackathon, we are planning some activities that relate to T2TRG (and its WISHI activity), ASDF, and CoRE.  We are cognizant that many will be on vacation that week, but we still want to get some work done ahead of the IETF111 happening the following week.
>> 
>> We’ll run a kickoff at 1400Z on this Monday (today/tomorrow depending on your timezone: https://www.timeanddate.com/worldclock/fixedtime.html?msg=IETF111+hackathon+T2TRG-WISHI-ASDF-CoRE&iso=20210719T14&ah=1 ).
>> 
>> The main purpose is to see who is planning to do what and to set up collaboration points.
>> We also want to agree on a time (approximately, but not necessarily exactly 1400Z again) where we have a day sync call Tue to Fri; we’ll announce this here, so if Monday doesn’t work for you, you can chime in later.
>> 
>> You may want to pre-fill below’s CodiMD with any additional information that might be useful for the kickoff (or during the week).  You may also want to register for the hackathon at https://registration.ietf.org/111/new/hackathon/ .
>> 
>> Grüße, Carsten
>> 
>> 
>> # T2TRG/WISHI/ASDF/CoRE hackathon @ IETF111
>> 
>> Kickoff: Monday, 2021-07-19, 1400Z (60 min max)
>> 
>> Notes: https://codimd.ietf.org/hackathon-2021-t2trg-111-monday
>> 
>> Meetecho: https://meetings.conf.meetecho.com/interim/?short=5903310f-3a8d-4f3a-93c4-5dbb1c9cb768
>> 
>> Note that we will use meetecho for this meeting.
>> Please check whether your datatracker account works; if it doesn’t, reset its password or get a new one at:
>> 
>> https://datatracker.ietf.org/accounts/create/
>> 
>> (It is worth doing this ahead of time as there may need to be an interaction with the IETF secretariat.)
>> 
>> ## Areas of work
>> 
>> ### SDF related
>> 
>> * tools (e.g., sdf compact)
>> * definition(s) of equivalence
>> * converters (translations) from SDF to X, from X to SDF
>>   * YANG (sdf-yang-converter.org) (*)
>>   * WoT (sdf-wot-converter.org)
>>   * IPSO/LwM2M (https://github.com/EricssonResearch/ipso-odm)
>>   * OCF?
>>   * Azure DTDL (*)
>>       * curl --data-urlencode [e-mail] http://wishi.nomadiclab.com:8083/odm2dtdl
>> * CI pipeline to keep translations up to date?
>> 
>> ### CoRE related
>> 
>> * CoRE resource directory? (*)
>> * CoRAL, particularly -href (CRIs) (*)
>> 
>> ### Others?
>> 
>> _______________________________________________
>> T2TRG mailing list
>> [e-mail]
>> https://www.irtf.org/mailman/listinfo/t2trg
>