From: arnt@gulbrandsen.priv.no
To: sml@ietf.org, Ben Bucksch <ben.bucksch@beonex.com>
Date: Tue, 28 Apr 2026 20:09:16 +0200
Message-Id: 
 <51SoKRXB4fsH5xlB23GyYy9B0ltqVy1mtnbcXJ8yAOk=.sha-256@antelope.email>
References: <87pl3k2l0o.fsf@libertango.gulbrandsen.priv.no>
 <7d018465-e97f-27f4-3be3-2404de50ef72@beonex.com>
Mime-Version: 1.0
Content-Type: multipart/alternative; boundary="https://antelope.email"
Message-ID-Hash: CW7QR6LODV2Y46NMNFHM5JOMQAGFISNN
Precedence: list
Subject: =?utf-8?q?=5BSml=5D_Re=3A_Suppressing_SML_fallbacks?=
Archived-At: 
 <https://mailarchive.ietf.org/arch/msg/sml/oqARvbPrnHU56L9tOBs_exExMAc>

--https://antelope.email
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Transfer-Encoding: quoted-printable

Hi,=C2=A0
Now that your say it, I remember that.
Perhaps the spec should be "render either this or your json-ld=20
thing, your choice, they're equivalent".
HJ? Can you add this to the draft, unless you disagree?
Arnt
On 2026-04-28 Ben Bucksch <ben.bucksch@beonex.com> wrote:
> Hey Arnt, SML group,
> Arnt Gulbrandsen wrote on 27.4.2026, 22:41:
> > SML is extremely minimal in some senses, there's no rendering=20
> > advice or anything. When I composed some mail on Friday, I=20
> > found that there was something I wanted: Suppose you send some=20
> > HTML mail and you attach some SML. Fine. You may want to add a=20
> > fallback of some sort. In actual fact I immediately =E2=80=A6
> Yes, there's definitely a need for that.
> We've discussed this about a year ago, in one of the IETF=20
> meetings. HJ suggested to mark this fallback HTML element with=20
> an attribute that references an ID in the SML. That attribute=20
> tells the renderer that the SML is equivalent to that HTML=20
> element, so the renderer can do its fancy programmatic rendering=20
> instead of the static HTML. We had this conceptually mostly=20
> nailed down, and were fairly concrete in the solution. We=20
> stopped just short of defining the exact HTML attribute name,=20
> value, and how IDs look like and how exactly they are found in=20
> the SML.
> Why important: E.g. the Polls usecase that I presented at FOSDEM=20
> 2026 <https://fosdem.org/2026/schedule/event/JQSRHP-parula/>.=20
> You want the responses from all recipients, not only those that=20
> happen to use Parula or HJ's version of K9Mail. So, my plan (not=20
> implemented yet) is to include a static preview of the options,=20
> and a link to a webpage, where the recipient can then answer the=20
> poll. Parula (and other SML mail readers that support this=20
> particular "meeting time poll" usecase) can render the calendar=20
> UI inline, and merge the poll with my own calendar, which=20
> supports me better, but non-Parula users can still respond. For=20
> this to work, Parula needs to know at which point in the HTML=20
> the fallback is, so that it can replace it.
> Ben
> =E2=80=8B--
> Beonex GmbH
> Taunusstein bei Wiesbaden, Germany
> CEO / Gesch=C3=A4ftsf=C3=BChrer: Ben Bucksch
> Handelsregister: Amtsgericht Wiesbaden, HRB 30065
> =E2=80=8B--
> Sml mailing list -- sml@ietf.org
> To unsubscribe send an email to sml-leave@ietf.org

--https://antelope.email
Content-Type: text/markdown; charset=utf-8
Content-Transfer-Encoding: quoted-printable

Hi,=C2=A0

Now that your say it, I remember that.

Perhaps the spec should be "render either this or your json-ld thing, =
your choice, they're equivalent".

HJ? Can you add this to the draft, unless you disagree?

Arnt

On 2026-04-28 Ben Bucksch <ben.bucksch@beonex.com> wrote:

> Hey Arnt, SML group,

> Arnt Gulbrandsen wrote on 27.4.2026, 22:41:

> > SML is extremely minimal in some senses, there's no rendering advice =
or anything. When I composed some mail on Friday, I found that there was =
something I wanted: Suppose you send some HTML mail and you attach some =
SML. Fine. You may want to add a fallback of some sort. In actual fact I =
immediately =E2=80=A6

> Yes, there's definitely a need for that.

> We've discussed this about a year ago, in one of the IETF meetings. HJ =
suggested to mark this fallback HTML element with an attribute that =
references an ID in the SML. That attribute tells the renderer that the =
SML is equivalent to that HTML element, so the renderer can do its fancy =
programmatic rendering instead of the static HTML. We had this conceptual=
ly mostly nailed down, and were fairly concrete in the solution. We =
stopped just short of defining the exact HTML attribute name, value, and =
how IDs look like and how exactly they are found in the SML.

> Why important: E.g. the Polls usecase that I [presented at FOSDEM =
2026](https://fosdem.org/2026/schedule/event/JQSRHP-parula/). You want =
the responses from all recipients, not only those that happen to use =
Parula or HJ's version of K9Mail. So, my plan (not implemented yet) is =
to include a static preview of the options, and a link to a webpage, =
where the recipient can then answer the poll. Parula (and other SML mail =
readers that support this particular "meeting time poll" usecase) can =
render the calendar UI inline, and merge the poll with my own calendar, =
which supports me better, but non-Parula users can still respond. For =
this to work, Parula needs to know at which point in the HTML the =
fallback is, so that it can replace it.

> Ben

> --
> Beonex GmbH
> Taunusstein bei Wiesbaden, Germany
> CEO / Gesch=C3=A4ftsf=C3=BChrer: Ben Bucksch
> Handelsregister: Amtsgericht Wiesbaden, HRB 30065

> --

> Sml mailing list -- sml@ietf.org

> To unsubscribe send an email to sml-leave@ietf.org

--https://antelope.email
Content-Type: text/html; charset=utf-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">Hi,&nbsp;
<div></div>

<div>Now that your say it, I remember that.</div>

<div></div>

<div>Perhaps the spec should be "render either this or your json-ld =
thing,
your choice, they're equivalent".</div>

<div></div>

<div>HJ? Can you add this to the draft, unless you disagree?</div>

<div></div>

<div>Arnt</div>

<div data-id=3D"quote-attribution:c7e67cdd-c7e0-422e-a085-7ff8ce1512aa">O=
n
2026-04-28 <span data-id=3D"person:5eebe042-f4fe-4844-b515-1a4cb59232be">=
Ben
Bucksch</span> &lt;<span
data-id=3D"person:5eebe042-f4fe-4844-b515-1a4cb59232be">ben.bucksch@beone=
x.com</span>&gt;
wrote:</div>

<blockquote data-id=3D"quoted-content:9122bd04-c73d-455e-8524-5173387309d=
2">
<div>
<div>
<p>Hey Arnt, SML group,</p>
</div>

<div>
<div>
<p>Arnt Gulbrandsen wrote on 27.4.2026, 22:41:</p>
</div>

<blockquote class=3D"truncated">SML is extremely minimal in some senses, =
there's
no rendering advice or anything. When I composed some mail on Friday, I =
found
that there was something I wanted: Suppose you send some HTML mail and =
you
attach some SML. Fine. You may want to add a fallback of some sort. In =
actual
fact I immediately =E2=80=A6</blockquote>
</div>

<div>
<p>Yes, there's definitely a need for that.</p>

<p>We've discussed this about a year ago, in one of the IETF meetings. HJ
suggested to mark this fallback HTML element with an attribute that =
references
an ID in the SML. That attribute tells the renderer that the SML is =
equivalent
to that HTML element, so the renderer can do its fancy programmatic =
rendering
instead of the static HTML. We had this conceptually mostly nailed down, =
and
were fairly concrete in the solution. We stopped just short of defining =
the
exact HTML attribute name, value, and how IDs look like and how exactly =
they
are found in the SML.</p>

<p>Why important: E.g. the Polls usecase that I <a
href=3D"https://fosdem.org/2026/schedule/event/JQSRHP-parula/"
title=3D"fosdem.org

https://fosdem.org/2026/schedule/event/JQSRHP-parula/">presented at =
FOSDEM
2026</a>. You want the responses from all recipients, not only those that
happen to use Parula or HJ's version of K9Mail. So, my plan (not =
implemented
yet) is to include a static preview of the options, and a link to a =
webpage,
where the recipient can then answer the poll. Parula (and other SML mail
readers that support this particular "meeting time poll" usecase) can =
render
the calendar UI inline, and merge the poll with my own calendar, which
supports me better, but non-Parula users can still respond. For this to =
work,
Parula needs to know at which point in the HTML the fallback is, so that =
it
can replace it.</p>

<p>Ben</p>

<p>--<br>Beonex GmbH<br>Taunusstein bei Wiesbaden, Germany<br>CEO /
Gesch=C3=A4ftsf=C3=BChrer: Ben Bucksch<br>Handelsregister: Amtsgericht =
Wiesbaden, HRB
30065</p>
</div>
</div>

<div>
<div>
<p>--</p>

<p>Sml mailing list -- sml@ietf.org</p>

<p>To unsubscribe send an email to sml-leave@ietf.org</p>
</div>
</div>
</blockquote>
<!--
{
  "@context": {
    "schema": "https:\/\/schema.org\/",
    "sml": "https:\/\/sml.draft.iana.org\/"
  },
  "@graph": [
    {
      "@id": "mid:7d018465-e97f-27f4-3be3-2404de50ef72@beonex.com",
      "@type": "schema:EmailMessage",
      "schema:sender": {
        "@id": "person:5eebe042-f4fe-4844-b515-1a4cb59232be"
      },
      "schema:dateSent": "2026-04-28T02:10:13Z"
    },
    {
      "@id": "person:5eebe042-f4fe-4844-b515-1a4cb59232be",
      "@type": "schema:Person",
      "schema:name": "Ben Bucksch",
      "schema:email": "ben.bucksch@beonex.com"
    },
    {
      "@id": "quote-attribution:c7e67cdd-c7e0-422e-a085-7ff8ce1512aa",
      "@type": "sml:QuoteAttribution",
      "sml:date": {
        "@id": "mid:7d018465-e97f-27f4-3be3-2404de50ef72@beonex.com"
      },
      "sml:from": {
        "@id": "person:5eebe042-f4fe-4844-b515-1a4cb59232be"
      }
    },
    {
      "@id": "quoted-content:9122bd04-c73d-455e-8524-5173387309d2",
      "@type": "sml:QuotedContent",
      "sml:message": {
        "@id": "mid:7d018465-e97f-27f4-3be3-2404de50ef72@beonex.com"
      },
      "sml:attribution": {
        "@id": "quote-attribution:c7e67cdd-c7e0-422e-a085-7ff8ce1512aa"
      },
      "sml:content": [
        {
          "@id": "text-content:3bf7c783-394d-4ce2-b6b8-3d99f734e4f4"
        }
      ]
    },
    {
      "@id": "text-content:3bf7c783-394d-4ce2-b6b8-3d99f734e4f4",
      "@type": "sml:TextContent"
    }
  ]
}
--></div>

--https://antelope.email--