Re: [homenet] number of devices in homenet

[Daniel Migault <[e-mail]>]

Hi Juliusz,

I believe the answer is similar to [1]. Regarding the number of devices we
expect, it seems hard to have exact prediction, however, if we do not make
it possible this number will likely remain zero. We need to make it easier
to have more devices.

Currently services mostly send private data to the cloud, and data is
accessed through a service hosted on that cloud. While service provider
have their own interest in doing so, it is also quite hard for individual
to set a service in the homenet and be able to access that service. It is
easy to export the video of your video camera on a public cloud, it is much
difficult to be able to access the vdeo camera from outside. This is
something we need to change.

Video camera, all sorts of sensors, light controls... there is no special
need these data goes outside your homenet.

I believe the number of devices can become quite large. I also believe that
2 would be sufficient to justify not configuring things more than once.

Yours
Daniel

[1]
https://mailarchive.ietf.org/arch/msg/homenet/NCbU4XHh0tk8HAowH5rcGEB3fOo

On Tue, Jun 11, 2019 at 9:11 PM Juliusz Chroboczek <[e-mail]> wrote:

> Dear Michael,
>
> >> Please see my unanswered e-mail of 21 November 2018.
>
> >>
> https://mailarchive.ietf.org/arch/msg/homenet/vz1kdCJISN6UPNZpj9ZD4e8EdwQ
>
> Thank you for your detailed reply.  I'm glad we're finally having
> a discussion about my objections to Daniel's proposal.
>
> > We strongly believe that the HNA needs to know the list of names in
> > order to be able to answer for those names when there is unstable (or
> > no) Internet connectivity.
>
> > Otherwise, applications and people have to know two different names for
> the
> > service. (A public one for when away, and the .local one)
>
> That's a good point.  While I happen to believe that it's reasonable to
> have a service known as "boombox.local" from home, and
> "boombox.jch.example.org" from the Internet, this might be inconvenient
> for e.g. smartphone users.
>
> >    o  the credentials for the dynamic DNS server need to be securely
> >       transferred to the hosts that wish to use it.  This is not a
> >       problem for a technical user to do with one or two hosts, but it
> >       does not scale to multiple hosts and becomes a problem for non-
> >       technical users.
>
> I think that's our main disagreement.
>
> For some reason, you guys seem to be assuming that the average user will
> want to publish hundreds of names in the global DNS.
>
> However, none of the end-user services that I know use incoming
> connections require a name in the global DNS to function (WebRTC, Skype,
> online games, BitTorrent, remote desktops, BTSync/Resilio, syncthing).
> Thus, my assumption is that the typical user will want to publish exactly
> 0 public names, and that only the extreme geek will publish up to 3 or 4
> (music server, NAS, game server, web server with family photographs).
>
> Richard, Daniel -- please be so kind as to explain why you think my
> assumption is wrong.  How many names do you envision wanting to publish in
> the public DNS, and for what purpose?
>
> -- Juliusz
>
> _______________________________________________
> homenet mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/homenet
>