Loading ...
Sorry, an error occurred while loading the content.

RE: [tracker2] Standby mode

Expand Messages
  • scott@opentrac.org
    ... I figured it d be better to do it once the cops had the car in sight. My Accord has a 15 amp fuse on the fuel pump, so I think the simplest way to set it
    Message 1 of 9 , Jul 7, 2006
    • 0 Attachment
      > Besides if the vehicle is indeed moving I'd sooner have the
      > cops stop it
      > themselves and arrest the bad guys. Rather than disabling it
      > so they'd
      > abandon the vehicle. Of course if they start running then

      I figured it'd be better to do it once the cops had the car in sight. My
      Accord has a 15 amp fuse on the fuel pump, so I think the simplest way to
      set it up would be to power the tracker through that circuit (the high-side
      switch can't switch a separate supply), making sure it's got adequately
      heavy cabling, and just short the output leads of the switch. When the
      switch is turned on, it'll switch up to 20 amps before the current limiting
      cuts in, but that's enough to blow the fuse. You have to get out of the car
      and open an access panel to replace it.

      A simple relay might be better, though. You could also kill the ECU, but
      I'd be afraid of causing damage to the engine. Killing the fuel supply
      shouldn't do anything more dramatic than starving the engine and making the
      car coast to a stop.

      Before anyone tries to implement this, I should point out that my auto shop
      grades were... well, not quite on par with my CS/EE grades, anyway. =]

      > Wow, very nice to see just how much flexibility you're
      > putting into the
      > device.

      It'll be able to do a lot more if I ever get a chance to implement the
      scripting language I've been thinking about. That'll have to wait until
      after the first production firmware release, though.

      Scott
    • scott@opentrac.org
      ... That could be done in the hypothetical scripting language. I think it d be fine with the authentication techniques I m working on. Right now it s all
      Message 2 of 9 , Jul 7, 2006
      • 0 Attachment
        > In addition to the existing remote operation;
        >
        > A code word, recognized from, say, incoming messages, which
        > would just indirect into a pre-configured command. no acks no nothing.

        That could be done in the hypothetical scripting language. I think it'd be
        fine with the authentication techniques I'm working on. Right now it's all
        callsign-based. Another one will be something you can calculate in your
        head or keep on paper, and the third will be a true cryptographic MAC
        that'll have to be generated by a computer.

        > The command(s) intended to be idempotent, would not cause
        > harm if Eve the
        > eavesdropper replayed it (the fuelpump fuse would remain blown ;) Nor
        > would Eve know what it did, if it was not obvious from
        > subsequent tracker2
        > behavior.

        Idempotence is actually a major concern in the command interface and
        scripting language design. It's very easy to get the same command multiple
        times due to retries and network topology, and the flash memory has finite
        write endurance. That's an issue in scripting if you say 'execute this
        configuration command when contact A is closed' - it'd be easy to erase and
        rewrite the same page in flash very quickly and eventually destroy it.

        My initial idea for a scripting language resembled a PLC's ladder logic,
        rather than a procedural language. I'm still leaning toward that sort of
        setup, but it introduces some complexity in dealing with output events that
        have to be edge-triggered.

        Scott
      • juha.nurmela@quicknet.inet.fi
        ... This code word would not require authentication, it would be a less-privileged thing. Instead of keys to the machine room, you d have a cord to pull, to
        Message 3 of 9 , Jul 7, 2006
        • 0 Attachment
          On Fri, 7 Jul 2006 scott@... wrote:

          > That could be done in the hypothetical scripting language. I think it'd be
          > fine with the authentication techniques I'm working on.

          This 'code word' would not require authentication, it would be
          a less-privileged thing. Instead of keys to the machine room,
          you'd have a cord to pull, to reset the airconditioning.

          > Idempotence is actually a major concern in the command interface and
          > scripting language design. It's very easy to get the same command multiple

          You know the cure ;) Message identifiers.

          > My initial idea for a scripting language resembled a PLC's ladder logic,

          I guess a Forth-like language is out of the question. Easy to implement,
          but, kind of crazy to usher people to code in RPN something, what
          they could write by themselves and recompile, in the first place.




          Juha
        • scott@opentrac.org
          ... Responding to commands and initiating pre-programmed actions is a big part of what I want the scripting language to be able to do. ... Yeah, not a lot of
          Message 4 of 9 , Jul 7, 2006
          • 0 Attachment
            > This 'code word' would not require authentication, it would be
            > a less-privileged thing. Instead of keys to the machine room,
            > you'd have a cord to pull, to reset the airconditioning.

            Responding to commands and initiating pre-programmed actions is a big part
            of what I want the scripting language to be able to do.

            > I guess a Forth-like language is out of the question. Easy to
            > implement,
            > but, kind of crazy to usher people to code in RPN something, what
            > they could write by themselves and recompile, in the first place.

            Yeah, not a lot of room.. but you're not the first to suggest Forth. I've
            got a book on designing 'little languages' and their interpreters... I'm
            hoping to get some insipration from that. It needs to be line-editable
            though, so you can make code changes remotely.

            Scott
          • Curt, WE7U
            ... I once wrote a Forth interpreter for a 6803 processor in under 2k. -- Curt, WE7U. APRS Client Comparisons: http://www.eskimo.com/~archer Lotto: A tax
            Message 5 of 9 , Jul 10, 2006
            • 0 Attachment
              On Fri, 7 Jul 2006 scott@... wrote:

              > > I guess a Forth-like language is out of the question. Easy to
              > > implement,
              > > but, kind of crazy to usher people to code in RPN something, what
              > > they could write by themselves and recompile, in the first place.
              >
              > Yeah, not a lot of room.. but you're not the first to suggest Forth. I've
              > got a book on designing 'little languages' and their interpreters... I'm
              > hoping to get some insipration from that. It needs to be line-editable
              > though, so you can make code changes remotely.

              I once wrote a Forth interpreter for a 6803 processor in under 2k.

              --
              Curt, WE7U. APRS Client Comparisons: http://www.eskimo.com/~archer
              "Lotto: A tax on people who are bad at math." -- unknown
              "Windows: Microsoft's tax on computer illiterates." -- WE7U
              "The world DOES revolve around me: I picked the coordinate system!"
            Your message has been successfully submitted and would be delivered to recipients shortly.