Nanonote like clamshell based on AM335x processor

Yann Sionneau yann at minet.net
Sun Aug 5 09:17:30 EDT 2012


Le 05/08/12 13:37, Werner Almesberger a écrit :
> Wolfgang Spraul wrote:
>> Werner himself and his unstoppable drive and inspiration. So I
> 
> Heh, thanks, but that's not entirely fair to a number of
> people, including yourself. What we can say with certainty is
> that activity and enthusiasm have greatly diminished.
> 
> Qi-Hw has changed from a place where people go because something
> exciting is happening (or where they at least expect it to
> happen any moment) to a place they go because it has become a
> habit.
> 
>> The TI idea would lead me away from all of those,
> 
> I quite strongly disagree with this view. If the last few years
> of Qi-Hw have shown anything, it's that you can't thrive on a
> starvation diet. It's pretty irrelevant if you're all set up to
> slay the dragon and bed the princess next week if you die of
> hunger today.
> 
> So if a company was willing to finance a Nanonote successor
> project around chips that are reasonably open, and thus breathe
> new life into our effort, I'd be more than willing to join the
> effort.
> 
> You've expressed a dislike for TI for a long time, but I still
> fail to see the reason. Maybe it's a legacy from the good times
> we had at Openmoko ?
> 
> PowerVR is evil, of course, and it's sad that TI still throw
> their support behind it, but if you don't make your product
> depend on this sort of acceleration, it's only a relatively
> minor nuisance to have it around, turned off and ignored by us.
[...]

I would tend to agree with Werner's point of view.
Chips may have proprietary IP in it, let's just not use what we don't
want to use.
If the chip, after discarding unusable IPs, seems to be good enough to
do what we want to do (good screen resolution, 802.11, etc) let's do it :)

It's always easy to point fingers at big companies saying they are evil
because they don't do open source hardware and that they don't care
about it.
And when they try to get in touch with you, you just look down on them
and say "your chips are not totally clean, I see dirt in them, sorry no.
Bye" ?

I don't see that as a way forward toward open source hardware expansion :)

This being said, I obviously am in favour of using a totally documented,
with open source drivers, chip.
But we would need to find one :) (let's put apart Milkymist SoC since
there is no ASIC yet and therefore no decent CPU frequency in it for a
general purpose product).

Is the TI AM335x documentation freely available ?
That's a deal breaker, we cannot do anything without any datasheet.
And I don't think Qi work would be interesting if only Qi volunteers
sign a NDA to have access to the datasheet.
At least anyone who buys the "nanonote v2" should receive a datasheet of
the product including the datasheet of the SoC.
But I would prefer the datasheet to be freely available to anybody.

All of this to sum up as : +1 Werner.

-- 
Yann Sionneau




More information about the discussion mailing list


interactive