[Battlemesh] Linksys promises not to block free firmware

[Daniel Golle]

Hi Henning,

I've already posted this on prpl's FCC list, also posting it here,
appologies for the cross-post...

On Fri, May 13, 2016 at 07:58:45PM +0200, Henning Rogge wrote:
> Sounds interesting.
> 
> http://arstechnica.com/information-technology/2016/05/linksys-wrt-routers-wont-block-open-source-firmware-despite-fcc-rules/

The WiFi radios here are FullMACs, ie. the WiFi operations are carried
out by the internal firmware of the microcontroller built-into that
type of WiFi chips. While that obviously has the legal advantage of
having the opportunity to lock-down the radio firmware seperately from
the rest of the system, often times the firmware lacks significant
features relevant for less common roles (IBSS, Monitor/Injection, ...)
and usually do not function well e.g. for community mesh networks or
impose a vendor lock-in (remember the OLPC pre-draft 802.11s mesh?
really, read about it and understand the implications a very similar
community interaction carried out by the very same chip company had for
a particular charity project popular at the time...).

So all this is not surprising at all to me, I don't even consider that
news. The strategy of Marvell and Linksys of marketing high-priced
"open source ready" devices based on recent Marvell platforms has been
going on for a while, obviously either suggesting that other devices
would not be "open source ready" or that the vendors involved would
actively put significant effort and make complete datasheets publicly
available. I can't tell whether any of the companies involved are
actively pushing for either direction, ie. lobbying for other devices
to be banned from markets they want to exclusively cater to (which
would obviously be very evil) or if they are publicly supporting the
community by making datasheets publicly available and show support for
the effort of making the system 100% blob-free (which would be nice).
If anyone has been following related press-releases or had any
interaction with the parties involved it would be good to hear more
about what's going on beyond the front-lines of marketing.

So those devices are interesting if you need something which looks
like a customizable router, but only in case you are happy with the
WiFi features provided by the vendor's wifi firmware (which is a
proprietary binary). Ie. experimenting with or optimizing the MAC layer
for less popular use-cases (point-to-point, outdoor mesh, vehicles) is
simply not possible for unafiliated individuals or once the platform
is no longer actively being marketed.

This is a HUGE step back compared to the freedom mac80211-based drivers
on top of SoftMAC designs (such as ath9k or mt76) provide to
independent researches and developers, community mesh networks and
practically everyone in a couple of years from now when the official
product life-time of the chip has ended and thus the vendor is likely
to no longer update the wifi chip's firmware for features not yet
invented or implemented at the time of its making.

In short: In my opinion this is a misleading marketing campain which in
the current political situation may provide the wrong impression that
"open source friendly" is an attribute not in contradiction with the
newly introduced lock-down requirements of the ETSI and the FCC.
It's like saying that every Android-phone is "open source friendly"
only because you may modify certain aspects of the OS and run free open
source software on-top of it.


Cheers


Daniel


> 
> Henning Rogge
> _______________________________________________
> Battlemesh mailing list
> Battlemesh at ml.ninux.org
> http://ml.ninux.org/mailman/listinfo/battlemesh