Re: [NTLK] New Newton possability (not a whine or hoax)

From: Jonas (jonas184_at_hotmail.com)
Date: Tue Oct 12 2004 - 14:30:19 PDT


----- Original Message -----
From: "Jim Witte" <jswitte_at_bloomington.in.us>
To: <newtontalk_at_newtontalk.net>
Cc: <newtontalk_at_newtontalk.net>
Sent: Tuesday, October 12, 2004 12:24 PM
Subject: Re: [NTLK] New Newton possability (not a whine or hoax)

>
> > This machine would most likely feature a faster CPU (in the area of
> > 233 MHz) without overclocking the sound, serial or IR ports.
>
> This I think would be possible, by putting an independent clock
> signal into those parts, or by changing the OS.

Or by forcing the multiplier higher rather then the bus speed.

> > The storage and system memory would both be expanded, though to what
> > degree I do not yet know. Most likely doubled.
>
> How? According to Paul, the address bus of the Voyager is already
> maxed out with 4MB. You could interface more system ram into MMU as
> virtual page space (up to 4GB if I remember right), but to get the
> system to access that as system RAM, would again require modification
> of the OS.

I'm still researching that.

>
> > This would have the added bonus of setting minds at ease in the memory
> > failure department.
>
> I assume you're refeferring to the -61100 bug. Paul is already
> working on that, as stated on NTLK.

That, and there was a recent thread asking if the Newton's internal memory
was likely to die of old age in the near future.

>
> > Along with this would go an LCD with both a modern touch screen and
> > backlight.
>
> Certainly possible. Rewrite TScreenDriver and TTabletDriver.
> getting the electrical interface to work without the Voyager specs,
> however..

I actually ment keeping the stock LCD for compatability while replacing the
touch screen and backlight with modern ones. Both of these parts are
available online.

>
> > Plus the Interconnect port on the back would be replaced by standard
> > serial and audio in/out.
>
> Certainly possible if you know how to do hand surface mount work, or
> plan to make your own 4-6 layer PCB without the electrical schematics.

Something like that.

>
> > Finally, there is the possability that a compact flash slot would be
> > added WITHOUT disrupting the 2 PCMCIA slots.
>
> Unlikely I'd think, but not impossible. The original concept Newton
> was supposed to support 4 PCMCIA slots. I don't know if there is
> support in the OS for that though. Better, would be to bring the
> system up to a CardBus spec, but that would require basically replacing
> the Voyager, as it's 32-bit not 16-bit, and can (with some hosts)
> support direct-card-DMA.

Replacing the chipset in this machine (I've been calling it the Newton 2900)
wouldnt really be possable.
Ask me again about a new chipset in a year or so...

>
> > Being basicly a newton 2100, it would still keep software
> > compatability while giving the machine some more life through nearly
> > all new components and the expanded hardware.
>
> This wouldn't by any chance be connected with the Einstein project :)

I'm not familier with that.

>
> > I have tried to round up all the common (and uncommon) modifications
> > and upgrades to implement them as best as possable in one package.
> > Would this be worth the money and downtime to you, the users of
> > Newtons?
>
> Certainly. 'Downtime' implies modifications to the system itself,
> which I'd think would be quite difficult, considering that the Voyager
> chips might have to be replaced, the NIC would have to be removed via
> hand surface-mounting, etc.

that it does, though not quite how you imagine it.

>
> My advice would be to forget about hardware modifications - learn
> minimal MacOSX Cocoa programming, learn to hack the system-level NOS
> (writing p-classes, etc), and approach Paul to help coding the Einstein
> project for a POSIX OS. That *is* the future, whether we like it or
> not (and why *shouldn't* we like it. It was a pipe-dream, but we've
> all scene it work.

I am not a coder, nor do I plan to be. I simply want to expand the hardware
as far as it will go.
Besides, wouldnt a faster but compatable newton be be better for a new OS?

>
> 220 Mhz-there-abouts CPU? Please - I want a 400MHz
> ARM/intel/whatever (ARM would be easier, from a recoding/emulation
> perspective since it has conditional instruction opcodes that don't
> have to be expanded into three or four opcodes for intel and maybe PPC
> [don't think PPC supports conditional ops, but could be wrong]).
> Preferably with either an SIMD processor like Altivec, accessible from
> the host-OS or from the emulation layer (harder, but possible), or a
> DSP, so that voice-recognition could become a real possibility. Bigger
> screen. External USB connections, Firewire, CardBus, the list goes on.
> This is all possible with Einstein, at minimal hardware cost once the
> software is done and the host-OS is chosen (or written - perhaps a very
> spare Mach kernel.)

Unfortunatly DEC/Intel only made up to a 233 MHz CPU.
The Intel 80200 that was ment to replace the StrongARM line is nowheres near
a drop in replacement.
As for the rest, ask me again in a year...

>
> Except for the possibility of an army of lawyers, which might not even
> come to pass..

I've tried to avoid legal trouble by stating that this is an *Upgrade* to
current messagepads, not a new unit being sold under the newton name.
And while there would be hardly any original components left, the all
important ROM chip would be one of them.

Finally, I'm using my old hotmail accounts as my server has gone offline
until the 18th along with my ADSL connection...

Does this answer some of your questions?

-- 
This is the NewtonTalk list - http://www.newtontalk.net/ for all inquiries
Official Newton FAQ: http://www.chuma.org/newton/faq/
WikiWikiNewt for all kinds of articles: http://tools.unna.org/wikiwikinewt/


This archive was generated by hypermail 2.1.5 : Tue Oct 12 2004 - 15:00:03 PDT