Hi Pawel,

From: Pawel Kowalik <[email protected]>
Date: Wednesday, March 18, 2026 at 2:50 AM
To: Jasdip Singh <[email protected]>, regext <[email protected]>
Subject: [regext] Re: POST use in EoH versus DoH

On 17.03.26 18:54, Jasdip Singh wrote:
Just wanted to follow up on one point from yesterday’s EoH presentation in 
REGEXT. When comparing EoH with DoH, it was noted that both use GET and POST. 
But there is one important distinction when it comes to BCP 56. DoH correctly 
falls back to POST if an equivalent GET query gets too large, per BCP 56 (4th 
para in [1]). In contrast, EoH seems to overload the POST use for PUT and 
DELETE scenarios, violating BCP 56. IMHO, it does not help to compare EoH with 
DoH. Guess that’s why the HTTPDIR review [2] recommended HTTP CONNECT instead 
to avoid the BCP 56 violation.

I thing you missed that DoH allows for virtually any DNS packet, also UPDATE 
(which makes for Add/Update/Delete) - which would be still GET/POST http 
requests.

[JS] That’s a good insight. However, there seems to be a distinction between 
DoH and EoH when it comes to operations. A DNS UPDATE operation is a mixed bag 
of add’s and delete’s (along with prerequisites) [1], whereas EPP, and thereby 
EoH, entails separate <create>, <delete>, and <update> operations. It seems 
right that in DoH, a DNS message could plausibly be for a DNS UPDATE. However, 
DoH is primarily intended for DNS queries normally sent over UDP [2] and not 
DNS updates which are more likely over TCP (IIUC). In contrast, EoH would 
primarily be for separate writes through <create>, <delete>, <update>, etc. 
Though the GET-to-POST transition in DoH primarily happens when a DNS query 
gets too large (per BCP 56), even using POST for DNS UPDATE — a single binary 
blob of prerequisites, add’s, and delete’s — seems appropriate as a processing 
instruction. Given EPP has separate create/update/delete actions, using POST 
for all these operations seems less appropriate when PUT and DELETE could have 
been used if BCP 56 were followed.


Hoping this is clear as mud. ;)


Thanks,

Jasdip


[1] https://datatracker.ietf.org/doc/html/rfc2136#autoid-2

[2] https://datatracker.ietf.org/doc/html/rfc8484#autoid-26

_______________________________________________
regext mailing list -- [email protected]
To unsubscribe send an email to [email protected]

Reply via email to