ibutton pin grafting

Ryan Lackey (rdl@MIT.EDU)
Tue, 31 Mar 1998 21:38:29 EST

• Next message: Simon Spero: "RE: HOLOCOMM: Secrecy by Delocalization"
• Previous message: Bruce Schneier: "Re: HOLOCOMM: Secrecy by Delocalization"

To: David Honig <honig@otc.net>
Cc: ibuttonpunks@soldier-of-fortune.mit.edu
Reply-To: ibuttonpunks@soldier-of-fortune.mit.edu
Subject: Re: Java iButton from Dallas Semiconductor
In-Reply-To: Your message of "Tue, 31 Mar 1998 09:27:02 PST."
             <3.0.5.32.19980331092702.007a2200@otc.net>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Date: Tue, 31 Mar 1998 21:36:16 EST
From: Ryan Lackey <rdl@MIT.EDU>

David Honig wrote:
> At 03:44 PM 3/30/98 EST, Ryan Lackey wrote:
> >If only there were a simple way to graft an LED and microswitch onto them
> >to enter a PIN and reveal internal state without trusting the local
> >computer... Sounds like a fun hardware hacking project.
>
> I thought of this too, with just a photosensor port, but a silica window
> wouldn't be as resistant
> to severe abuse as the stainless. Unless I'm wrong.

I do not propose directly adding pin/display hardware to the ibutton, but
rather building a pen or wallet or whatever with this functionality. A
*ring* needs to be bulletproof, but a wallet doesn't. You could recess the
LED and cover it with sapphire, too. Pretty hardcore damage resistance that
way.

Simply using a trusted pilot or something would work, too.
>
> But shouldn't the device's analog-wire-to-digital-signal-converter be able
> to sense, e.g., the timing of
> a simple (momentary) doorbell switch attached to its "one wire" bus? Use
> a unary-coding
> of your data...a battery and a switch.. you can pulse-dial a telephone this
> way, too, BTW. This would be essentially all software with zero mods to
> the can,
> and you could plug the can into a modified holder containing the switch
> and a battery. I don't know if the APIs give access to this low level.

Yes, this is a good plan -- it's the kind of input I originally wanted to do.
It could be done with just a 5v battery, actually, if you slowed it down
the speed far enough -- use 5 wpm morse as a passphrase input mechanism.

Unfortunately, this totally fails the usability test. I think something
like a combination lock style rotating bezel or whatever makes the most
sense as an input device, or *perhaps* some kind of chording thing, with
a small microcontroller, and LED for status or LCD display. Make it
the size of a AA battery or slightly larger.
>
> Also, how hard is it to detect EM radiation coming from the cans? Very low
> power devices
> hermetically sealed with a battery in steel is pretty tight...
>
Can you say "faraday cage", boys and girls? :) (actually, I think it might
not be steel the whole way around, since the top surface is insulated). I
think it would pass TEMPEST tests, though.

Ibuttons just rule.

- --
Ryan Lackey
rdl@mit.edu
http://mit.edu/rdl/

------- End of Forwarded Message

-- 
Ryan Lackey
rdl@mit.edu
http://mit.edu/rdl/		

• Next message: Simon Spero: "RE: HOLOCOMM: Secrecy by Delocalization"
• Previous message: Bruce Schneier: "Re: HOLOCOMM: Secrecy by Delocalization"