Content-Type: text/plain;
	charset=utf-8
Mime-Version: 1.0 (Mac OS X Mail 13.4 \(3608.120.23.2.7\))
From: Carsten Bormann <cabo@tzi.org>
In-Reply-To: 
 <177694623882.1212598.16313196910512372107@dt-datatracker-b45949c58-5szpr>
Date: Thu, 14 May 2026 16:54:43 +0200
Content-Transfer-Encoding: quoted-printable
Message-Id: <5C27064D-1F89-4E41-B940-4F8612542DCA@tzi.org>
References: 
 <177694623882.1212598.16313196910512372107@dt-datatracker-b45949c58-5szpr>
To: Valery Smyslov <valery@smyslov.net>
Message-ID-Hash: HOPM4YAKYD5FCW2EOHSNGDYQC2MTP76W
CC: art@ietf.org, draft-ietf-iotops-7228bis.all@ietf.org, iotops@ietf.org,
 last-call@ietf.org
Precedence: list
Subject: =?utf-8?q?=5BIotops=5D_Re=3A_=5Bart=5D_draft-ietf-iotops-7228bis-06_ietf_las?=
	=?utf-8?q?t_call_Artart_review?=
Archived-At: 
 <https://mailarchive.ietf.org/arch/msg/iotops/pogdUGLW32lSGkTi40suwDsvMvs>

Hi Valery,

thank you for this review!

We created a common PR clarifying our use of MSL, motivated by the =
comments we got on MSL:

https://github.com/lwig-wg/terminology/pull/19/changes

We also created a PR trying to cover the improvements you specifically =
suggested:

https://github.com/lwig-wg/terminology/pull/24/changes

On 2026-04-23, at 14:10, Valery Smyslov via Datatracker =
<noreply@ietf.org> wrote:
> [=E2=80=A6]
> Nits:
> 1. The purpose of the note below Table 10 (the text in parentheses: =
"'Sx'
> stands for 'Size x'") is not clear to me. [=E2=80=A6]

Indeed, it is better to explain the name in the text referencing the =
table, which we now do.

> 2. Section 5.3
> I do not think that reference to MSL in the context of B0 networks is =
relevant.
> MSL is only defined for TCP and I doubt if nodes connected to B0 =
network will
> use TCP. In any case, MSL in TCP is only concerned with the chance for =
SN
> repeat after host reboots. Referring to COAP's MAX_LATENCY value =
(discussed in
> I-D.gomez-tiptop-coap) is more appropriate in this context, but its =
default
> value is 100 (and not 120 as MSL).

We added a reference to CoAP=E2=80=99s MAX_LATENCY, but decided to =
continue TCP=E2=80=99s MSL as the main yardstick when defining our =
classes, clarifying that this usage is just an example for protocol or =
application behavior that makes some assumption on a maximum latency in =
this general vicinity.

> 3. Section 5.3
> I also think that 62.5 bytes calculated here must be rounded to the =
full number
> of bytes (I assume that all protocols operate on byte boundary and =
cannot send
> say 62.3 bytes to remain "responsive" in B1 networks).

Indeed.  We use =E2=89=A5 63 now (which is still a bit too exact for =
this metric, but that=E2=80=99s what the arithmetic gives us).

You can see the changes in the context of all the IETF last-call related =
changes at

=
https://author-tools.ietf.org/api/iddiff?doc_1=3Ddraft-ietf-iotops-7228bis=
&url_2=3Dhttps://lwig-wg.github.io/terminology/draft-ietf-iotops-7228bis.t=
xt

Of course, IESG processing continues, so any further comments are =
welcome.

Gr=C3=BC=C3=9Fe, Carsten