Re: [Masque] Éric Vyncke's No Objection on draft-ietf-masque-h3-datagram-10: (with COMMENT)

[David Schinazi <[e-mail]>]

Hi Eric, thanks for your review!
Responses inline.
David

### Abstract

s/This document describes HTTP Datagrams/This document specifies HTTP
> Datagrams and the Capsule Protocol/ ?


I personally prefer "describes" here.

### Section 1
> For consistency with other reasons, should a reference be added to
> `unreliable version of the CONNECT method` ?


We could add a reference to Section 9.3.6 of RFC 9110, but I think that
this would make the sentence harder to read. There's already a reference to
RFC 9110 in that paragraph. If this is just about consistency with the
WebSocket, I'd favor readability or consistency.

### Section 2
> ```
>    HTTP extensions can
>    override these requirements by defining a negotiation mechanism and
>    semantics for HTTP Datagrams.
> ```
> s/can/MAY/ ? (but I am not a native English speaker) I also noticed that
> 'MAY' is used at the end of section 2.1.


Agreed, fixed via this commit:
https://github.com/ietf-wg-masque/draft-ietf-masque-h3-datagram/commit/d7d06b1ca504617a8121ba206fb526a06b1db780

### Section 2.2

I wonder whether this small section is useful. The text should rather go
> into the intro of section 3.


This small subsection provides continuity, it exists in the HTTP Datagrams
section and points to how to encode HTTP Datagrams over capsules.

### Section 3.4

While I am not too familiar with HTTP & QUIC, I wonder when reading
> `Endpoints indicate that the Capsule Protocol is in use on a data stream by
> sending a Capsule-Protocol header field with a true value.` Until now, the
> Capsule Header is not yet defined even less how to send a true value (or
> did I misread part of the I-D)


The header is defined in the paragraph right above the one you quoted. It
references the Structured Fields RFC and that explains how to send a true
value.

### Section 3.5
> ```
> Since QUIC DATAGRAM frames are required to fit within a QUIC packet,
>    implementations that reencode DATAGRAM capsules into QUIC DATAGRAM
>    frames might be tempted to accumulate the entire capsule before
>    reencoding it.  This can cause flow control problems; see
>    Section 3.2.
> ```
> The problem is clear, but what is the recommendation for implementers ?


Agreed, fixed in this commit:
https://github.com/ietf-wg-masque/draft-ietf-masque-h3-datagram/commit/0de86dbe59612a4c006f89a51b878f1d5acdb325