[MLS] Re: Group Context Extensions in Leaf Node Capabilities

Sean Turner <sean@sn3rd.com> Tue, 03 February 2026 19:00 UTC

From: Sean Turner <sean@sn3rd.com>
Message-Id: <7BB5F7E8-D720-411A-AF94-A93C5BCB34A6@sn3rd.com>
Content-Type: multipart/alternative; boundary="Apple-Mail=_D92FF96E-0B53-4C34-AA1F-BB743CC9A76F"
Mime-Version: 1.0 (Mac OS X Mail 16.0 \(3826.700.81.1.4\))
Date: Tue, 03 Feb 2026 13:59:41 -0500
In-Reply-To: <CAL02cgRyvAj-OA1Qy+fN0gwXip6WfnENjVewXyJMssZm0eeTNw@mail.gmail.com>
To: Jan Winkelmann <jan@cryspen.com>
References: <CAH0jtghrRL9bk+gxkQ2oU_VNgQ9giKgHWH+O235Fd0xnd+QzDA@mail.gmail.com> <CAL02cgRyvAj-OA1Qy+fN0gwXip6WfnENjVewXyJMssZm0eeTNw@mail.gmail.com>
Message-ID-Hash: JJDJFBHSIPUGBSHAOKOZSL67TVKTLEVE
CC: MLS List <mls@ietf.org>, Richard Barnes <rlb@ipv.sx>
Precedence: list
Subject: [MLS] Re: Group Context Extensions in Leaf Node Capabilities
Archived-At: <https://mailarchive.ietf.org/arch/msg/mls/ggqdywWfkhfeTB9sqN2yC2Z7ij8>

Jan,

If have a chance could you submit a errata via this link:
https://www.rfc-editor.org/errata.php
Nick or I can then as our AD to verify the errata.

spt

> On Feb 3, 2026, at 12:27, Richard Barnes <rlb@ipv.sx> wrote:
> 
> Hi Jan,
> 
> It seems to me you are correct.  Despite the compatibility checks on join, an incompatibility could creep in if a member removes capabilities in an Update / Commit.  The bounds established in RequiredCapabilities would still hold, though.
> 
> --Richard
> 
> 
> On Mon, Jan 26, 2026 at 5:12 AM Jan Winkelmann <jan@cryspen.com <mailto:jan@cryspen.com>> wrote:
>> Hi everyone,
>> 
>> I think I may have spotted a mistake in the RFC and would appreciate the thoughts, so I know whether to file an erratum.
>> 
>> I noticed it says on the topic of validating LeafNode capabilities:
>> 
>> > Note that the latter two requirements mean that all
>> > MLS GroupContext extensions are mandatory, in the
>> > sense that an extension in use by the group MUST be
>> > supported by all members of the group.
>> > --- https://www.rfc-editor.org/rfc/rfc9420.html#section-13.4-6
>> 
>> To that end, it requires that we check that the LeafNodes in KeyPackages that are added support all extensions in the group context[^1]. However, it doesn't seem to require that the same check is mandated for LeafNodes in Update proposals or update paths.
>> 
>> To me that seems like an oversight in the RFC, is that correct?
>> 
>> Let me know what you think!
>> 
>> Jan
>> 
>> [^1]:  From the RFC:
>> > A client adding a new member to a group MUST verify
>> > that the LeafNode for the new member is compatible
>> > with the group's extensions. The capabilities field
>> > MUST indicate support for each extension in the
>> > GroupContext.
>> > --- https://www.rfc-editor.org/rfc/rfc9420.html#section-13.4-5.5
>> and
>> > A client joining a group MUST verify that it supports
>> > every extension in the GroupContext for the group.
>> > Otherwise, it MUST treat the enclosing GroupInfo
>> > message as invalid and not join the group.
>> > --- https://www.rfc-editor.org/rfc/rfc9420.html#section-13.4-5.6
>> _______________________________________________
>> MLS mailing list -- mls@ietf.org <mailto:mls@ietf.org>
>> To unsubscribe send an email to mls-leave@ietf.org <mailto:mls-leave@ietf.org>
> _______________________________________________
> MLS mailing list -- mls@ietf.org
> To unsubscribe send an email to mls-leave@ietf.org

[MLS] Group Context Extensions in Leaf Node Capab… Jan Winkelmann
[MLS] Re: Group Context Extensions in Leaf Node C… Sean Turner
[MLS] Re: Group Context Extensions in Leaf Node C… Richard Barnes