On Thu, 2006-04-06 at 11:01 -0700, Kevin Purcell wrote:
ttp://news.com.com/Gates+sizes+up+the+Webs+next+generation+-+page
> +3/2008-7345_3-6051890-3.html
>
> > But rather than try to quibble with whose memory or writing is
> > correct, let this third point make it clear:
> >
> > Third: We've always said, from day one, the machine is a fully *open*
> > platform, hardware (and from our end, software). This is why the specs
> > of the machine are up on the web in good detail even before the design
> > of the hardware is complete. How could our machine not be an open
> > platform, if we want kids to be able to understand how computers
> > really
> > work?????
> >
> > We've stated the hardware platform is open, consistently, from day
> > one,
> > to everyone, up to and including Bill Gates and other senior
> > executives
> > at Microsoft and everywhere else. We are not in the discrimination
> > business, or out to get anyone; we are here to build a machine that
> > can
> > meet our goals of enabling kids all over the world to learn.
>
> I don't disagree with that.
>
> But I also made a distinction that the hardware design won't be 100%
> "open" (whatever that means -- I think of "open" as in GPLed or Share-
> and-share-alike CC license) unless Quanta will release any IPR
> considerations they have. I think the latter is unlikely if it
> enables their competitors to make them because that is their specific
> competency.
The key IPR is in fact OLPC's; the novel display is a big deal, as it
makes a huge difference in cost.
I don't want to go into the precise details of the Quanta relationship;
but explicit from the beginning has been the understanding that there
would be competition, and in finite time. Otherwise, we'd have no good
way of ensuring the price would drop as we would like to see.
However, we do need to get to true volume to make the economics work, so
for the first year, we are trying to ensure that the business go in the
one direction. After that, the game changes.
>
> There are degrees of openness and as you make clear I'm sure you are
> going to make it as open as you can (more open than the iPaq, for
> example). So long as the "model" is open so a user knows what the
> peripheral chips are and where they live in the memory map or on the
> IO ports then that should suffice to bing any OS to the hardware
> platform.
Yup. That's what we're doing. Not that everything is quite in concrete
yet.
I expect at a minimum to get to where we were on the original iPAQ 36xx
models, where we were able to release pretty complete specs (some of the
later iPAQ's were complete buyouts where Compaq/HP did not have the
rights to the specifications). Whether we go further to schematics or
beyond, etc, I don't honestly know right now.
Understand that the display is unique, and to get the price point we
need we need to be able to fill a fab (actually, an old, already
amortized one).
It isn't as though you can go to a random LCD vendor and get one of
these. So it isn't as though you can just take the design trivially and
replicate it.
>
> I wouldn't be surprised if someone else came to you and Quanta and
> said "We want Quanta to make 5 million of these for us to sell in
> schools in the US with 30% of profits going to OLPC" everyone would
> jump at the chance.
We are certainly looking at these sorts of cross subsidies where
feasible. There are parts of the world where even $100 is more than
they can afford.
As for your other conspiracy theories, I'll leave them alone.
Regards,
- Jim
--
Jim Gettys
One Laptop Per Child
--
olpc-software mailing list
[email protected]
https://www.redhat.com/mailman/listinfo/olpc-software
