[NTLK] Newton Mp2x00 Extract the ROM, request for help

Sylvain Pilet sylvain at pilet.net
Fri Feb 6 05:05:05 EST 2015

Hi Simon, Paul, Matthias, and all,

Awesome! Thank you, really thank you Simon.

I extracted a ROM of a Watson version 2.2.00-0, I'll send it to Paul for him to dissect. This ROM has been extracted from a Watson ROM introduced into a MessagePad 2000. Because this morning NCX did not communicate directly with the Watson (via The Watson-Dock), I will watch and correct this, a cable problem I think.

And, I will extract the various ROM Watson, from MessagePad 2x00 but also directly Watson. Because I noticed something;
- When Schlumberger electronic components are present in the Watson, ROM says:
Infos sur la mémoire
Interne____356 Ko______3167 Ko

Version ROM F2.1 (F1-037)-1
RAM installée 924ko

- When the electronic components Schlumberger are NOT present in the Watson, the SAME ROM says:
Infos sur la mémoire
Nom_____ Utilisés____Disponibles
Interne____27Ko_______3496 Ko

Version ROM 2.2.00-0
RAM installée 948ko

This means that Schlumberger electronics just add code in ROM when present in the machine. It may be useful to see how this is done.

That makes two versions of ROM (with the same ROM we all agree).
But I have also another Watson reacts to it differently. It indicates a 2.2.00-0 ROM, RAM 948 KB, while having many electronics installed in it. This seems to be a Watson following gereration, because the software is: WatsonPro 1.5, unlike others that do not have this mention.

I would send all files to Paul that he looks in detail.

Now it will be very interesting to see (and especially to extract) the Watson-ROM that Andreas has his home. The own seems especially, since accepts a standard packages be installed in it.

Oh, I have not thought of timed dump time of the ROM on One building, I would do for the next, but in fact you have to be patient!

This may perhaps interest you, there are two physically very different versions of the Watson ROM, I would send pictures in the first place ...

See you soon,

Sylvain Pilet

- A P P L E  N E W T O N -
- Apple Newton NPDS Server: http://www.message-pad.no-ip.org
- Apple Newton NPDS Tracker: http://www.message-pad.no-ip.org:3680

Le 5 févr. 2015 à 15:59, Paul Guyot <pguyot at kallisys.net> a écrit :

>> Those running Mac OS X 10.9 or higher might be interested in a preview version of NCX that allows you to extract the tethered Newton’s ROM <http://newtonresearch.org/connection21/index.html>. It takes about 33 minutes over a 115k serial connection.
>> Should work for Watson too, unless the dock protocol has been doctored. Thanks to Paul for the suggestion.
> Dear Simon,
> Excellent work! Thank you very much for this!
> I've been considering adding the feature to Einstein directly, however I no longer have any USB-Serial adapter as I gave mine to another Newton user. Then I looked at fixing serial emulation directly in Einstein, but this seems yet another challenge (BTW, could NCX use a regular unix pipe ?). I guess this release of NCX can also extract the ROM from a TCP connection.
> The download time could probably be reduced by compressing ROM on the Newton. Probably not worth the development effort, though.
> Also, I am confused about the 33 minutes delay, though. In 2000, I did extract a ROM through a 38400 regular connection within about 30 minutes, at least according to my notes.
> Watson definitely runs at 38400 since you cannot (normally) install any Dock extension to run at 115k.
> Sylvain has been kind enough to help me figure out about the Watson "digital signature" of packages. More than a decade after I first looked at the problem, I guess the solution is very near. Watson-compatible packages indeed have a 32 bytes special "raw" part, but this is no cryptographic digital signature unlike what I was told years ago, just a dummy structure with Schlumberger. It is the same for every package. They also have an undocumented flag set (we knew about it in 2003-2004). They eventually have reserved2 flags set to something that I haven't figured out yet (regular packages have 0 there).
> This isn't strong anyway as DoNotInstallScript get executed even when packages are refused (typically Avi's Backdrop offer to install as the backdrop application does appear before Watson says the package is not compatible).
> It also seems that (using a specific device ?) Andreas has been able to install packages on an internal flash and then get these packages accepted by a Watson ROM, producing screenshots as he posted here earlier this week. The issue with Watson is indeed that very few units were produced and therefore some have prototype ROM boards as well as specific hardware, and his success might be limited to a single device (I suspect regular Watson ROM boards will just erase the internal flash that was initialized by a MP2x00 US ROM board, just like other N2 ROMs do).
> Paul
> -- 
> Semiocast            http://semiocast.com/
> +33.183627948 - 20 rue Lacaze, 75014 Paris

More information about the NewtonTalk mailing list