From: Will Leingang (will.leingang_at_navpress.com)
Date: Sat Sep 14 2002 - 22:03:09 PDT
Funny you mention ITSY, most of linux development I've done in the past has
been that of getting an ITSY system built. I'm actually performing most of
this work with my limited knowledge of embedded systems from
running/building Linux on a few iPaq's (pocket pc) and an xBox.
To explain the kernel I "tried" to build for the Newton, would be like
trying to describe the process of making a peanut butter and jelly sandwich,
there was simply no science to it. I basically streamlined the Familiar
kernel source by removing hardware calls that newton hardware wouldn't be
able to facilitate, thereby allowing the system to operate in system rom
with nothing other than simple memory calls.
Getting it to the Newton was as much of a hack as anything. I hijacked
everything I needed from an iPaq 3600, by running lead wires from the
Newton's rom to my iPaqs, this was of course an amateurs mistake, as the
hardware was in no way similar. It was a rush approach, to a delicate task.
I figured since the Newton didn't allow access to the system ROM, I would
re-flash it according to Compaq engineering specs for the Intel chipset, I
failed to take into consideration the differences of the chipset altogether.
I think I overloaded the Sharp ROM by not properly securing one of the
leads.
Optimally, there would need to be a boot loader to facilitate any further
development. I'm as of yet unfamiliar with any interface on the Newton that
could be hacked to accommodate this. Because Apple has admittedly gone the
way of open-source, I would think that if enough sweet-talking could be done
we could squeeze schematics and engineering specs out of the developer
support archives. After all, if it is a "dead" product, Apple shouldn't be
afraid of assisting open-source Newton development. Wink Wink, Nod Nod...
It would be nice to put a few heads together on this, possibly even come up
with a newton linux project team. For now though, I'm going to continue to
research the possibility of hardware changes, and a Newt linux loader, maybe
even a little brown nosing with past newton project leads.
Will
-----Original Message-----
From: Jeff Sheldon [mailto:jeff_at_jurai.net]
No one has a clear enough understanding of Newton internals to approach
this. There's a theory that the Newton 2x00s can auto-boot from a PCMCIA
card and perhaps by-pass the ROM, but with the primary chipset closed and no
longer in production, it all pretty much dead-ends regardless. I for one
would embrace Linux in a heartbeat, since much of what we use could be
re-written as open source, though we'd lose Inkwell. We'd need some
discipline to keep applet design focued on a "common interface toolbox" or
some such so our Newtons still feel mostly like Newtons.
It would be great to have reliable IMAP, SSH, and various internet
applications ported to a Newton via Linux.
> I put some time into a kernel (modified Familiar build) last night,
> after a desperate attempt to build a simple boot loader I think I
> trashed my Newton's ROM.
Could you provide more specifics about this, preferably to this list?
Perhaps kernel details and how you set it up for install on the Newton for
use?
> Has anyone had success even trying to remotely access the system
> RAM/ROM? What steps have been taken in hardware documentation?
There's been some exploration into this posted at
http://www.kallisys.com/?lg=en
I'd be interested to see some formal schematics made of the MP2100 & eMate
as the singular source of all which remains as "MessagePad" and work around
that model for grassroots hardware & Linux development. Something in league
with Compaq's open Itsy PDA.
In the meantime, I'm going to rip open one of my Newtons and write down chip
make/model, check to see if the board is multi-layer, and such. I'd be
happy to consier a special-interest mailing list surrounding this if people
want to provide tips (ie. "You can use this smaller hi-res LCD planel in
place of a 2100 screen).
Hell, how much could one remove off an existing newton to shrink it and not
leave the system in a failure state? I may look into seeing if a card slot
and the expansion port could be disabled & removed. There may also be newer
chip products smaller and compatible with older newton chips. It would be
great to convert the remaining PCMCIA slot into a permanent Compact-Flash
compatible slot (avoiding a converter) and flat battery packs could replace
the bulk of the lower battery tray.
A formal schematic could lead to a good digital layout for re-design and
automatic board prodcution. This could seriously piss off the wrong people,
of course, but I can't imagine anyone persuring Newton re-design to this
point.
-- Read the List FAQ/Etiquette: http://www.newtontalk.net/faq.html Read the Newton FAQ: http://www.chuma.org/newton/faq/ This is the NewtonTalk mailing list - http://www.newtontalk.net/
This archive was generated by hypermail 2.1.5 : Mon May 26 2003 - 12:59:21 PDT