[regext] Re: On bare identifiers in Extensions draft

[kowa...@denic.de]

Hi,

On 30.05.25 21:54, Hollenbeck, Scott wrote:

-----Original Message-----
From: Andrew (andy) Newton <a...@hxr.us>
Sent: Friday, May 30, 2025 3:26 PM
To: Pawel Kowalik <kowa...@denic.de>; Gould, James <jgo...@verisign.com>;
Hollenbeck, Scott <shollenb...@verisign.com>; regext@ietf.org
Subject: [EXTERNAL] Re: [regext] Re: On bare identifiers in Extensions draft

Caution: This email originated from outside the organization. Do not click links
or open attachments unless you recognize the sender and know the content is
safe.


On 5/22/25 02:08, Pawel Kowalik wrote:

I would also not put too much value to prefixes. Typically the frameworks

require full name anyway to access the object or to map it to a programming
language representation.

Actually I would be in favour of restoring the text of -04 2.4.5 with an 
explicit

update to RFC9083 allowing the bare identifier pattern (which also would solve
the issue of justification when SHOULD NOT could be broken and DE review
interpretation issue). Contrary to other cases brought up by Andy (I-JSON,
Unichars) I think this update would be OK in this draft, as it only affects
extensions.

I have written up a proposed compromise:

https://secure-web.cisco.com/18e-
FXFFrvru7_Ae6AK7p8rtjkZgn1wKIJulucJBH0gMF_pqcn_Z3P5XBbnRUhfr97Kg8jUe
hdWb8w9S-08AxVMrEJ-9-
GKwooCKhKE8j4j5gYXhRE3FgBEbbewwT227tVWzvXrBeZHC_fSYM_vZSApLm-
vRwTXQXtrjazBqI0nerukoVr3BW2g_Weleak82qTdWaSSeLK2kqq82G-
KUFkwURUKOiovtg2ElazRSk0KRcUy7joTxE9yR8auKQ_bHDBLf2OjTRNNh9t2CVF9
tvWDfndWm2yO7mwVnz4KfXXmQ/https%3A%2F%2Fgithub.com%2Fanewton19
98%2Fdraft-regext-rdap-extensions%2Fcompare%2Fmain...allow-bare

[SAH] "only when a technical solution cannot otherwise be defined". I still worry that 
we're going to have debates about what constitutes a "technical solution", but it sounds 
like the burden of proof will be on whomever is proposing to use a bare extension identifier.  I 
can live with that.

- when the extension is only adding a single JSON property

Kind Regards,

Pawel

[PK] I'm not happy with "technical solution cannot otherwise be defined" as this is a condition likely impossible to fulfil or proof, as there will be always some solution possible. This also does not express the prime motivation for bare identifiers use: - when identifier is generic and is a building block of RDAP protocol itself, backed with an IETF RFC -> analogy to standards tree media type registration

smime.p7s
Description: S/MIME Cryptographic Signature

_______________________________________________
regext mailing list -- regext@ietf.org
To unsubscribe send an email to regext-le...@ietf.org