[Arm-netbook] Patent-Left

[Jean Flamelle]

On 2/25/18, Luke Kenneth Casson Leighton <lkcl at lkcl.net> wrote:
>  ... *thinks*.... ok so let's think this through.
>
>  what do proprietary companies owning Certification Marks do.... o
> arse, you're right: they charge staggeringly-large amounts of money
> and place huge burdens on people.  FCC 2G/3G/LTE Certification (per
> firmware revision, per product, per *version* of product, per
> *company* e.g. AT&T... $50k *EACH*...), BLE Certification (USD $10,000
> per software release *per factory*), and so on.

I heard that's what keeps fairphone out of the US.

>  that's gonna get really old, really quickly.

Already has for me tbh xD

-

>  i know that the FSF, whom people are in effect "Authorised to
> Represent" when they received the RYF Endorsement Certification Mark,

That gets to an interesting idea of patent-left: what-if anyone
receiving certification to use a hypothetical "left"-patent, was as a
condition of the license allowed to certify anyone else irrevocably
unless their use or use of anyone directly certified or endorsed by
them, causes proven harm with their use to anyone emotionally,
physically, or socially---aside from game-theory-arbitrary conflict
that just happens to involve an instance of the patent's use through
no fault of the implementation. Sounds like MLM.

We could actually add that if the patent is used without license it's
only violated if said harm is done, except burden of proof them shifts
to the person using the patent to demonstrate no harm was done
(including to the licensor-network which may have been convinced waste
excessively taxing effort to ensure the implementation didn't cause
harm which could have been dedicated elsewhere protecting public trust
from damage that was indeed caused prove-ably because the community
was too focused on a false alarm which could have reasonably been
avoided).

I hope that idea is less dense than the economic one earlier.

"~
> simply trust people to "get it right".  they *don't* ask that people
> "re-certify", they said to me, basically, "you represent us and our
> good name, don't screw up by doing things like add proprietary
> firmware on it, or modify the description so people get confused and
> buy the wrong thing".
~"

I like this

>  they *didn't* say "You Must Re-Certify If You Make Even One Tiny Change".

*thumbs up*

>  so i believe there's room for both types of approaches... the
> question is: which approach could risk causing harm?  btw, just to be
> clear: anyone who *guarantees* full libre compliance (releases
> everything under libre licenses - casework, CAD, source, everything)
> zero charges *and* assistance in any way possible.

I think there is too, though I hope to make the ever more communal
option safer and easier to protect legally with this brainstorm.
That's still a longterm goal. We can always run the risk of being too
strict and loosen the way we do things later, so long as you, RMS, and
the entirety of the FSF stick to their guns (morals).

(which should that even be a question at this point?)
It's like asking if someone with endorsement from peta will ever start
killing animals for sport.
(maybe because somewhere down the line they started to blindly assume
humane population control strategies will never succeed)

Sure its happened before, but the question asks the probability of a
scenario so rare requiring so much conspiracy that it borders upon
irritating even the most pragmatic actor.

Thank you to Luke and everyone contributing to make this all happen btw.