I entered an issue in the QUIC Multipath repo to discuss this, with my
initial suggestions: https://github.com/quicwg/multipath/issues/625
On 12/10/2025 2:50 PM, Martin Thomson wrote:
It would be helpful if the authors suggested codepoints, rather than have IANA
choose. It looks to me like some of these frame types in particular would
benefit from some having one byte encodings, others two bytes (4 or 8 might
even make some sense for some, but I don't think that's necessary for standards
track RFCs).
On Thu, Dec 11, 2025, at 09:01, David Dong via RT wrote:
Dear Jana Iyengar, Ian Swett, Martin Thomson (cc: quic WG),
As the designated experts for the QUIC registries, can you review the
registration proposals in draft-ietf-quic-multipath-18 for us? These
were already previously allocated as provisional registrations in -15
and some in a prior version of this document (I've copied the thread
below). Please see:
https://datatracker.ietf.org/doc/draft-ietf-quic-multipath/
The due date is December 24th.
If this is OK, we'll update the assignment at:
https://www.iana.org/assignments/quic/
Unless you ask us to wait for the other reviewers, we’ll act on the
first response we receive.
With thanks,
David Dong
IANA Services Sr. Specialist
--
On Thu Jul 10 00:52:51 2025, [email protected] wrote:
Thanks David.
On Thu, Jul 10, 2025, at 09:33, David Dong via RT wrote:
Hi Martin,
Thank you for your review and clarifications. I've updated/added the
below, and will pass your message to the authors as well regarding
shorter codepoints for permanent registration:
QUIC Transport Parameters:
0x0f739bbc1b666d0d initial_max_path_id provisional
[draft-ietf-quic-multipath-15,
Section 2.1] 2025-07-09 Yanmei Liu [Yanmei_Liu]
QUIC Frame Types:
0x15228c00-0x15228c01 PATH_ACK provisional [draft-ietf-
quic-multipath-15,
Section 4.1] 2023-07-26 Yanmei Liu [Yanmei_Liu]
0x15228c05 PATH_ABANDON provisional [draft-ietf-quic-
multipath-15,
Section 4.2] 2023-07-26 Yanmei Liu [Yanmei_Liu]
...
0x15228c07 PATH_STATUS_BACKUP provisional [draft-ietf-
quic-multipath-15,
Section 4.3] 2023-11-03 Yanmei Liu [Yanmei_Liu]
0x15228c08 PATH_STATUS_AVAILABLE provisional [draft-ietf-
quic-multipath-15,
Section 4.3] 2023-11-03 Yanmei Liu [Yanmei_Liu]
0x15228c09 PATH_NEW_CONNECTION_ID provisional [draft-ietf-
quic-multipath-15,
Section 4.4] 2025-07-09 Yanmei Liu [Yanmei_Liu]
0x15228c0a PATH_RETIRE_CONNECTION_ID provisional
[draft-ietf-quic-multipath-15,
Section 4.5] 2025-07-09 Yanmei Liu [Yanmei_Liu]
0x15228c0c MAX_PATH_ID provisional [draft-ietf-quic-
multipath-15,
Section 4.6] 2025-07-09 Yanmei Liu [Yanmei_Liu]
0x15228c0d PATHS_BLOCKED provisional [draft-ietf-quic-
multipath-15,
Section 4.7] 2025-07-09 Yanmei Liu [Yanmei_Liu]
0x15228c0e PATH_CIDS_BLOCKED provisional [draft-ietf-
quic-multipath-15,
Section 4.7] 2025-07-09 Yanmei Liu [Yanmei_Liu]
QUIC Transport Error Codes:
0x004150504142414e APPLICATION_ABANDON_PATH Path abandoned
at the
application's request provisional [draft-ietf-quic-multipath-15,
Section 4.2.1] 2025-07-09 Yanmei Liu
[Yanmei_Liu]
0x004e4f5f4349445f NO_CID_AVAILABLE_FOR_PATH Path abandoned
due to no
available connection IDs for the
path provisional [draft-ietf-quic-multipath-15, Section
4.2.1] 2025-07-09 Yanmei Liu [Yanmei_Liu]
0x0052534c494d4954 PATH_RESOURCE_LIMIT_REACHED Path abandoned
due to
resource limitations in the
transport provisional [draft-ietf-quic-multipath-15, Section
4.2.1] 2025-07-09 Yanmei Liu [Yanmei_Liu]
0x00554e5f494e5446 APPLICATION_ABANDON_PATH Path abandoned
due to
unstable interfaces provisional [draft-ietf-quic-multipath-15,
Section
4.2.1] 2025-07-09 Yanmei Liu [Yanmei_Liu]
Registry:
https://www.iana.org/assignments/quic/
Best regards,
David Dong
IANA Services Sr. Specialist
On Wed Jul 09 02:14:13 2025, [email protected] wrote:
Hi David,
Anything already registered as provisional can stay there, maybe
with
updated references or names for those listed in this new request.
My
understanding is that these are simply label changes, not semantic
changes. That only makes it slightly more annoying to match
registry
to older versions of the draft, but there is no interoperability
hazard here. (Note that QUIC does not depend on frames being
globally
unique, but we set the registry up that way because it wasn't worth
the headaches of multiple registrations for the same codepoint.)
The old registrations for transport parameters cannot be overridden
in
that way. That registry genuinely does need to be unique. So those
old registrations need to stay.
Any new registrations would then be added, with the new reference.
On Wed, Jul 9, 2025, at 12:04, David Dong via RT wrote:
Hi Martin,
Some of the QUIC Frame Types codepoints in this request were
requested
by a previous version of this document and they were assigned as
provisional as well (it looks like there are some changes to the
existing entries in the registries. Others also don't appear in
the
current version of the draft as well; I can ask the authors about
them):
QUIC Transport Parameters
---
0x0f739bbc1b666d05 enable_multipath provisional
[draft-
ietf-quic-multipath-05,
Section 3] 2023-07-26 Yanmei Liu [Yanmei_Liu]
0x0f739bbc1b666d06 enable_multipath(-06) provisional
[draft-
ietf-quic-multipath-06,
Section 3] 2023-11-03 Yanmei Liu [Yanmei_Liu]
QUIC Frame Types
---
0x15228c00-0x15228c01 ACK_MP provisional [draft-ietf-quic-
multipath-05,
Section 8.1] 2023-07-26 Yanmei Liu [Yanmei_Liu]
0x15228c05 PATH_ABANDON provisional [draft-ietf-quic-
multipath-05,
Section 8.2] 2023-07-26 Yanmei Liu [Yanmei_Liu]
0x15228c06 PATH_STATUS provisional [draft-ietf-quic-
multipath-05,
Section 8.3] 2023-07-26 Yanmei Liu [Yanmei_Liu]
0x15228c07 PATH_STANDBY provisional [draft-ietf-quic-
multipath-06,
Section 8.3] 2023-11-03 Yanmei Liu [Yanmei_Liu]
0x15228c08 PATH_AVAILABLE provisional [draft-ietf-quic-
multipath-06,
Section 8.4] 2023-11-03 Yanmei Liu [Yanmei_Liu]
QUIC Transport Error Codes
---
0x1001d76d3ded42f3 MP_PROTOCOL_VIOLATION Multipath protocol
violation provisional [draft-ietf-quic-multipath-05,
Section
9] 2023-07-26 Yanmei Liu [Yanmei_Liu]
Just to confirm, do we have approval to assign these new
registrations
as requested with the status field set to "Provisional"? These
three
existing provisional codepoints will have the Frame Type Name
updated:
0x15228c00-0x15228c01
From "ACK_MP" to "PATH_ACK"
0x15228c07
From "PATH_STATUS_BACKUP" to "PATH_STANDBY"
0x15228c08
From "PATH_STATUS_AVAILABLE" to "PATH_AVAILABLE"
Thank you.
Best regards,
David Dong
IANA Services Sr. Specialist
On Tue Jul 08 22:05:58 2025, [email protected] wrote:
These seem best as provisional registrations. If I were to
advise
the
authors, I would suggest that they look for shorter codepoints in
the
long term, so any permanent registration should wait for that.
It's probably good to get these on the books if people are going
to
deploy code based on these codepoints, but provisional
registration
is
ideal, especially if they might change.
On Wed, Jul 9, 2025, at 06:25, David Dong via RT wrote:
Dear Jana Iyengar, Ian Swett, Martin Thomson,
As the designated experts for the QUIC registries, can you
review
the
registration proposal below? If it's OK, we'll make the
assignment
at:
https://www.iana.org/assignments/quic/
These all appear to fall within the Specification Required
ranges;
could you let us know if these should be made as permanent
registrations, or should we treat this as an RFC 7120 early
allocation
for draft-ietf-quic-multipath-15 (in which case we will need
the
approval of one of the quic WG Chairs and the AD)?
The IESG has asked us to request that reviews be returned
within
two
weeks, which in this case would make the due date July 22nd.
With thanks,
David Dong
IANA Services Sr. Specialist
On Tue Jul 08 04:08:53 2025, [email protected] wrote:
Contact Name:
Yanmei Liu
Contact Email:
[email protected]
Type of Assignment:
We want to add 1 transport parameter, 10 frame types, and 4
transport
error codes for Multipath Extension for QUIC (draft-15).
1. transport parameter which need to be added to the "QUIC
Transport
Parameters" registry under the "QUIC Protocol" heading:
initial_max_path_id (0x0f739bbc1b666d0d)
2. 10 new frame types which need to be added to the "QUIC
Frame
Types"
registry under the "QUIC Protocol" heading:
- PATH_ACK (0x15228c00-0x15228c01)
- PATH_ABANDON (0x15228c05)
- PATH_STATUS_BACKUP (0x15228c07)
- PATH_STATUS_AVAILABLE (0x15228c08)
- PATH_NEW_CONNECTION_ID (0x15228c09)
- PATH_RETIRE_CONNECTION_ID (0x15228c0a)
- MAX_PATH_ID (0x15228c0c)
- PATHS_BLOCKED (0x15228c0d)
- PATH_CIDS_BLOCKED (0x15228c0e)
3. 4 Transport Error Codes for multipath extension are to be
added
to
the "QUIC Transport Error Codes" registry under the "QUIC
Protocol"
heading:
- APPLICATION_ABANDON_PATH (0x004150504142414e)
- PATH_RESOURCE_LIMIT_REACHED (0x0052534c494d4954)
- PATH_UNSTABLE_INTERFACE (0x00554e5f494e5446)
- NO_CID_AVAILABLE_FOR_PATH (0x004e4f5f4349445f)
Details of these parameters and frame types can be found in
document:
https://datatracker.ietf.org/doc/html/draft-ietf-quic-
multipath-
15#name-iana-considerations
Registry:
QUIC IANA name space:
https://www.iana.org/assignments/quic/quic.xhtml
Description:
Multipath Extension for
QUIC(https://datatracker.ietf.org/doc/html/draft-ietf-quic-
multipath-
15) has been in the last call of IETF QUIC Working Group.
Registration
for code points of the latest version is necessary for interop
tests.
As draft-15 has been reviewed by most people, we believe it's
very
close to the RFC version.
Additional Info:
1. Multipath extension for QUIC:
https://datatracker.ietf.org/doc/html/draft-ietf-quic-
multipath-
15#name-iana-considerations
2. QUIC IANA name space:
https://www.iana.org/assignments/quic/quic.xhtml
