OLPC Laptop - an open source substandard ?

[Jim Gettys]

Fundamentally, the issue is that which is faced by any change: learning
curve requires additional investment and time to make the change.

Unless that investment can be done in parallel, this change translates
into delay and cost.  Right now, there are only one set of people
familiar with the mesh code, working under a tight deadline.  We're not
about to tell them to delay the overall project by switching horses in
midstream.

So to break the conundrum there also have to be people willing to
actually do the work, as the primary firmware team is working on getting
us code we can use and deploy.

Remember, the code we're talking about here started life well before
OLPC: we were not in a position to suggest alternative choices in RTOS
systems as we didn't exist then...  

As in the base system BIOS/firmware situation, the right way to look at
this is we know the desired/needed end-state, and to take steps toward
that state as time and resources permit.  There, we're now at a fully
free software stack, having started life with a conventional BIOS.
                             Best Regards,
                                      - Jim


On Sat, 2006-12-30 at 10:53 -0500, Michail Bletsas wrote:
> 
> 
> >   eCOS can be configured to run with a full tcp/ip stack in less
> than 8k, so
> > I'm not sure what the issue is.
> > 
> The issue is that we need working code now and we have to rely on
> whatever RTOS people that can deliver code under hard deadlines feel
> comfortable working with. 
> 
> M.
> _______________________________________________
> Devel mailing list
> Devel at laptop.org
> http://mailman.laptop.org/mailman/listinfo/devel
-- 
Jim Gettys
One Laptop Per Child