Re: [homenet] How many people have installed the homenet code?

Tim Coote <> Thu, 21 April 2016 15:30 UTC

Return-Path: <>
Received: from localhost (localhost []) by (Postfix) with ESMTP id 4317212EC10 for <>; Thu, 21 Apr 2016 08:30:14 -0700 (PDT)
X-Virus-Scanned: amavisd-new at
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9] autolearn=ham autolearn_force=no
Received: from ([]) by localhost ( []) (amavisd-new, port 10024) with ESMTP id fca6gGfOFhkA for <>; Thu, 21 Apr 2016 08:30:13 -0700 (PDT)
Received: from ( []) by (Postfix) with ESMTP id BC39012E9FF for <>; Thu, 21 Apr 2016 08:30:12 -0700 (PDT)
Received: by (Postfix, from userid 1000) id 56E7516C129F; Thu, 21 Apr 2016 16:30:11 +0100 (BST)
Received: from [] (localhost [IPv6:::1]) by (Postfix) with ESMTP id AF96816C1183; Thu, 21 Apr 2016 16:30:10 +0100 (BST)
Content-Type: text/plain; charset=utf-8
Mime-Version: 1.0 (Mac OS X Mail 8.2 \(2104\))
From: Tim Coote <>
In-Reply-To: <>
Date: Thu, 21 Apr 2016 16:30:10 +0100
Content-Transfer-Encoding: quoted-printable
Message-Id: <>
References: <> <>
To: Juliusz Chroboczek <>
X-Mailer: Apple Mail (2.2104)
X-Spambayes-Classification: ham; 0.00
Archived-At: <>
Subject: Re: [homenet] How many people have installed the homenet code?
X-Mailman-Version: 2.1.17
Precedence: list
List-Id: IETF Homenet WG mailing list <>
List-Unsubscribe: <>, <>
List-Archive: <>
List-Post: <>
List-Help: <>
List-Subscribe: <>, <>
X-List-Received-Date: Thu, 21 Apr 2016 15:30:14 -0000

> On 21 Apr 2016, at 15:00, Juliusz Chroboczek <> wrote:
> Hnetd is designed to integrate into the OpenWRT ecosystem, which it does
> remarkably well.  Because of that, it depends on a large part of the
> OpenWRT stack, such as their DHCP and RA clients and servers.  While it is
> possible to run hnetd
> This was the second most important reason for writing shncpd: to make an
> implementation of Homenet that is easy to port to any Unix system.
> I think I've been reasonably successful with that, but shncpd needs some
> more work [1].
> [1]
>    has a list of known limitations.
It’s the IPv6 stuff that I’m most interested in - IoT with ipv4 (NAT) is too expensive to support at scale. At least for the retail market. My sense was that the api to the openwrt tools was a likely stumbling block.  I did get the babeld stuff working (thanks for what you’ve done there), but gave up with tools like dibbler for PD as the challenge is the coordination across nodes.
> Even if you're just testing plain babeld, even without hnetd, you must
> kill network manager.  If you don't, thinkgs may appear to work at first,
> but NM will take revenge later.
I hated NM. But I think that it’s getting better. It does provide an initial overall visualisation of what’s going on. Something will be required to show the overall picture across all digital entities involved in a system, including ‘phones with both wifi and LTE connections. For the consumer there needs to be at least a green/red light (cf a CD pipeline status), but finer granularity would also have value.  I would hope to encourage the NM community to at least take cognisance of the Homenet protocols.

At the complexity that is emerging in homes, a configuration defined by a gui will not scale: too hard to get to a known state, especially with >1 node.

Maybe I need to create a separate openwrt vm if I cannot run the hnetd stack on a more complete distro.
> — Juliusz