Re: Mohamed Boucadair's Discuss on draft-ietf-httpbis-safe-method-w-body-13: (with DISCUSS and COMMENT)

On Mon, Nov 17, 2025 at 11:11:14AM +0100, Julian Reschke wrote:
> Hi Glenn.
> 
> Am 17.11.2025 um 10:06 schrieb Glenn Strauss:
> > ...
> > Related to Mohamed's note, please clarify that QUERY supports
> > HTTP Range requests.
> 
> They are as optional as in HTTP.
> > https://www.rfc-editor.org/rfc/rfc9110#name-range
> >     [...]
> >     A server MUST ignore a Range header field received with a request method that is unrecognized or for which range handling is not defined. For this specification, GET is the only method for which range handling is defined.
> >     [...]
> > 
> > https://datatracker.ietf.org/doc/draft-ietf-httpbis-safe-method-w-body/
> > Section 2.8. Range Requests
> >     The semantics of Range Requests for QUERY are identical to those for
> >     GET, as defined in Section 14 of [HTTP].  Byte Range requests (the
> >     only range unit defined at the time of writing), however, offer
> >     little value for the results of a QUERY request.
> > 
> >     Query formats often define their own way for limiting or paging
> >     through result sets, such as with "FETCH FIRST ... ROWS ONLY" in SQL.
> >     It is expected that these built-in features will be used instead of
> >     HTTP Range Requests.
> > 
> > If HTTP Range Request handling as defined in RFC9110 should not be used
> > with QUERY, then please be clear that HTTP Range Request 'Range' header
> > from RFC9110 is not defined for QUERY.  If HTTP Range Request handling
> 
> There is no reason for "SHOULD NOT" here. We say that byte-range requests
> are of little value though.
> 
> > could be used with QUERY, then please more explicitly state that HTTP
> > Range Request handling is defined for QUERY.
> 
> "The semantics of Range Requests for QUERY are identical to those for GET,
> as defined in Section 14 of [HTTP]."
> 
> > I could imagine a Range byte-range request as a coarse way for a
> > client to limit the size of the potential response data returned for
> > an initial request, leaving open the possibility of resubmitting the
> > same request for future retrieval of additional byte-ranges.
> 
> Yes.
> 
> > At the same time, I wonder if HTTP Range Requests should not be defined
> > for QUERY.  My server currently evaluates the Range request header only
> > for request method GET.
> They are in fact defined.
> 
> Do you want to suggest concrete text changes here?

I propose modifying the first sentence of Section 2.8, but only if you
also find the following clearer:

> > Section 2.8. Range Requests

-     The semantics of Range Requests for QUERY are identical to those for
-     GET, as defined in Section 14 of [HTTP].

+     HTTP Range Requests (Section 14 of [HTTP]) are defined for QUERY
+     and have semantics identical to those for GET.

Secondarily, "and have semantics identical to those for GET"
might be omitted.

"HTTP Range Requests (Section 14 of [HTTP]) are defined for QUERY."
might be sufficient.

Cheers, Glenn

Received on Monday, 17 November 2025 10:43:05 UTC