Sync & Install protocols
Zack T. Smith
Can anyone tell me whether the Sync and, more important,
the Install serial protocols are publicly defined anywhere?
I'd like to write a program for Linux (the free Unix) that
would let Linux users download packages and do
syncronizations.
My expectaton is that Apple is being secretive about
the protocol, but if anyone has done any serial snooping
and discovered something interesting then I would surely
like to know about it. In fact, I bet everyone would :).
By the way, if no info whatsoever is to be availed at
present, would anyone know of any programs for the Mac
that would let me listen into serial port activity,
and maybe echo to a log file?
Thanks for any info,
--
Zack T. Smith
Creator of MacShell, the C Shell for the Mac
Email me if you'd like a demo copy.
James da Silva
>My expectaton is that Apple is being secretive about the protocol, but if
>anyone has done any serial snooping and discovered something interesting
>then I would surely like to know about it. In fact, I bet everyone would
>:).
I've had a look through a data line monitor at a package download and the
protocol doesn't seem too difficult, just strangely verbose. The only
likely hook would be what was used as the crc/checksum on the data packets.
If one of the standard algorithms wasn't used, it'd be a trick to figure
out.
The bigger question is, what would Apple's attitude to a public
implementation of the protocols be? The fine print on the NCK license has
the usual B.S. prohibitions against reverse engineering. Whether such
things are really binding is doubtful, but who would be able to stand up to
a nuisance suit for some free software?
Given that the current NCK sucks and thus adversely affects Newton sales,
you'd think Apple would be happy to see some other implementations. At
least, that would be rational. :-)
Anyway, I'd like to see some free software to replace the connection kit,
probably using its own protocol and newton-side program (ala slurpee), Just
In Case. It probably wouldn't be too hard to be much faster than NCK.
We'd still need to speak the package download protocol to bootstrap the
alternative newton-side.
Anyone working on anything like this? Anyone interested in working on it?
I'm newtonscript-impaired and $800 short of an NTK, but I can do some work
on the Unix end.
Jaime
............................................................................
: Stand on my shoulders, : j...@cs.umd.edu : James da Silva
: not on my toes. : uunet!mimsy!jds : http://www.cs.umd.edu/users/jds
Charles Jordan, Charlotte, NC
> In article <zackCvu...@netcom.com>, Zack T. Smith <za...@netcom.com> wrote:
> >My expectaton is that Apple is being secretive about the protocol, but if
> >anyone has done any serial snooping and discovered something interesting
> >then I would surely like to know about it. In fact, I bet everyone would
> >:).
>
Color me very interested.
> I've had a look through a data line monitor at a package download and the
> protocol doesn't seem too difficult, just strangely verbose. The only
> likely hook would be what was used as the crc/checksum on the data packets.
> If one of the standard algorithms wasn't used, it'd be a trick to figure
> out.
>
I read somewhere that they used MNP 5. I've been wanting to examine the
data stream via a line monitor, but don't (yet) have a Windows machine to
run the NCK on.
> The bigger question is, what would Apple's attitude to a public
> implementation of the protocols be? The fine print on the NCK license has
> the usual B.S. prohibitions against reverse engineering. Whether such
> things are really binding is doubtful, but who would be able to stand up to
> a nuisance suit for some free software?
>
Someone married to a litigator. Seriously, my wife *loves* to go to court,
dislikes Apple and understands computers. At any rate, I think it's a moot
point. There are already a couple of other implementations of the install
protocols and, so far, Apple hasn't raised any flags.
> Given that the current NCK sucks and thus adversely affects Newton sales,
> you'd think Apple would be happy to see some other implementations. At
> least, that would be rational. :-)
>
Personally, I think that Apple Computer and rational thoughts have little
in common.
> Anyway, I'd like to see some free software to replace the connection kit,
> probably using its own protocol and newton-side program (ala slurpee), Just
> In Case. It probably wouldn't be too hard to be much faster than NCK.
> We'd still need to speak the package download protocol to bootstrap the
> alternative newton-side.
>
If you need to speak the package download protocol anyway, why re-invent the
wheel? I'd like to have a UNIX implementation of the package downloader
that didn't require any additional sw on the newton. It just seems like the
simplest solution.
What I'd really like is an X-Windows application that interfaced to the standard
apps on the Newton so that I could use the keyboard and screen on my desk to
enter data into the Newton. I've used Typomatic, and it's a good start, but
you have to read the Newton's screen and use the Newton's pen to move to another
field. I'd like to have everything displayed on my workstation and use my mouse,
or even the tab key, to change fields.
> Anyone working on anything like this? Anyone interested in working on it?
> I'm newtonscript-impaired and $800 short of an NTK, but I can do some work
> on the Unix end.
>
I too am newtonscript-impaired, but I'm not sure that's a liability. At least
as far as implementing a UNIX version of the download protocol. I'm thinking
about purchasing a 486 laptop. If I make the plunge, then I'll have something
to run NCK and can take a stab at decoding the download protocol.