Re: [Sml] ld+json schemas, We've been here for quite a while

[Hans-Joerg Happel <[e-mail]>]

I think we are sort of back to square one here, which is actually good.

Some points for discussion are included in the documents referenced in 
the BoF proposal. Those are certainly not perfect, perhaps partly too 
broad and more clarity is need - for what this discussion (and perhaps 
the BoF) are helpful (appreciating your input!).

I'll try to rephrase some points in different words:

* There's roughly about four ships of different sizes by now (i.e., 
Gmail, 1&1, Yahoo, ...)
* Those ships can effectively dock only in their dedicated ports (e.g., 
Gmail tools; rare/incomplete support by general clients)
* In order to onboard, passengers need to do some research, slightly 
different for each ship (i.e., study supported markup; setup guidelines)
* Current environment makes it hard for people to run their own small ships
* Existing ships are basically sailing in the bay (i.e., not yet ready 
for the ocean? - thus also far from boiling the ocean ;-)
* It's unclear if the current status quo will enable the discovery of 
new continents (i.e., I'd see more potential than sailing through the bay)

In a rough summary: I'd argue the current practice basically works for a 
"large sender to large ISPs" case.

As written in an earlier mail, the BoF does not intend to re-invent the 
wheel. Describing/slightly extending the status quo could be a 
worthwhile approach already. Most of that will probably be detail - such 
as the full/partial representation discussion, or to define under which 
circumstances clients should "accept" markup - but I think those may be 
worthwhile discussions in order to achieve a higher level of 
interoperability/hopefully adoption.

In that sense, yes, I think the BoF should probably recapitulate part of 
this discussion and decide if there are aspects worth to work on.

I also think that there are a number of things that could follow upon 
this later on (some of them mentioned in the BoF-related docs, such as 
discovery).

Best,
Hans-Joerg

On 09.03.23 21:04, John C Klensin wrote:
> +1
> John and I traditionally have different views about topics like
> whether ships have sailed sufficiently to justify abandoning
> discussions of other approaches,
> but I'm guessing he is probably correct in this case.
>
> I guess the next question is what you intend to do with the BOF?
> Reprise this conversation?  Something else?
>
>    john
>
>
> --On Thursday, March 9, 2023 13:03 -0500 John R Levine
> <[e-mail]> wrote:
>
>>> Regarding your suggestion, I think this looks quite similar
>>> to he one I made  in my mail from Feb 17th, suggesting, e.g.:
>>> ...
>> I think multipart/related makes more sense but I have a more
>> basic question.
>>
>> I had forgotten about all the schema.org schemas.  There are a
>> lot of them, and they are indeed used to embed info about
>> plane reservations and invoices into the HTML in mail
>> messages, as in Gmail Highlights.
>>
>> While I agree that in principle it would make more sense to
>> break the actions out into separate MIME elements rather than
>> hiding it in JSON chunks inside script elements, that ship has
>> sailed.  Unless we see interest from the people using ld+json
>> to put them into MIME parts, I don't see the point.
>>
>> I also note that schema.org is a W3C community group so I
>> would definitely talk to them to avoid a needless tug of war
>> with W3C.
>>
>> Regards,
>> John Levine, [e-mail], Taughannock Networks,
>> Trumansburg NY
>> Please consider the environment before reading this e-mail.
>> https://jl.ly
>