Announcement

Collapse
No announcement yet.

Predator Drones based on pirated, malfunctioning software?

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #16
    Re: Predator Drones based on pirated, malfunctioning software?

    Originally posted by ThePythonicCow View Post
    Your statement that one can reverse engineer any system, given the inputs and outputs, remains false.

    I think C1ue is using the term reverse engineering loosely. What he means is to come out with a formula that can produce approximately the same output given the same inputs.

    Comment


    • #17
      Re: Predator Drones based on pirated, malfunctioning software?

      Originally posted by TPC
      Now you've shifted your claim in another direction.

      Your initial claim stated "given the inputs and outputs".

      Now you amend that to "Given 100% of all possible inputs and 100% of all possible outputs".

      Your amended version is theoretically impossible to provide, outside of the singular stream of reality itself. "All possible inputs" includes things such as a high energy gamma ray from another galaxy flipping a memory bit, or the sun turning into a red giant, destroying planet earth, or your friendly local SWAT team firing a stray round into the device, or ...

      Apparently I made a false assumption when I read your initial claim. I presumed that you actually intended to state something meaningful.

      Silly me.
      Sorry, but you're the one who arbitrarily made the definition of reverse engineering as being the complete recreation of the original functionality and therefore impossible.

      Reverse engineering can be 1% of the original functionality, or it could be 110%.

      Since you chose to define it as 100% - then I merely redefined what would be needed to achieve that level.

      As my (multiple) examples showed, it is quite possible to achieve extremely high levels of reverse engineering - both to the point of improving past the original (from your rigid definition, not a complete recreation) as well as to near complete functionality (as in the Predator Drone targeting software which is close but apparently so far good enough) and onwards to completely identical function including reproduced bugs.

      Comment


      • #18
        Re: Predator Drones based on pirated, malfunctioning software?

        Old School

        This meant war


        New School

        WASHINGTON — The C.I.A. has drastically increased its bombing campaign in the mountains of Pakistan in recent weeks, American officials said. As part of its covert war in the region, the C.I.A. has launched 20 attacks with armed drone aircraft thus far in September, the most ever during a single month, and more than twice the number in a typical month.

        This means...something else

        Comment


        • #19
          Re: Predator Drones based on pirated, malfunctioning software?

          Originally posted by don View Post
          Old School

          This meant war


          New School

          WASHINGTON — The C.I.A. has drastically increased its bombing campaign in the mountains of Pakistan in recent weeks, American officials said. As part of its covert war in the region, the C.I.A. has launched 20 attacks with armed drone aircraft thus far in September, the most ever during a single month, and more than twice the number in a typical month.

          This means...something else

          Nice Stuka Don.

          Comment


          • #20
            Re: Predator Drones based on pirated, malfunctioning software?

            Originally posted by c1ue View Post
            As my (multiple) examples showed, it is quite possible to achieve extremely high levels of reverse engineering
            Agreed. Yes, such is quite possible, sometimes, but not always.

            I should know better by now than to ever attempt to disagree with anything you post. Doing so just frustrates me to no good end.
            Most folks are good; a few aren't.

            Comment


            • #21
              Re: Predator Drones based on pirated, malfunctioning software?

              Dang ... I am hopeless. Back to my crypto example.

              Let's say I inform you that I have a program that computes what is called a message digest (a fancy phrase for a checksum or hash.)

              Let's further say that I tell you exactly how my program works. It takes any input text, appends a specific random number that I have generated, one time, and then computes the SHA256 secure hash of the result. The only fact I refuse to divulge is that specific random number.

              I then allow you to present any number of input texts you want to my program, and I provide, as fast as can be computed on whatever computer that you provide me the resulting output SHA256 secure hash.

              You will never, in your lifetime or mine, be able to determine that random number by such efforts, and hence you will never be able to predict reliably what the output will be for any input that you have not already tested.

              I would consider that to be a failure to usefully reverse engineer my program.
              Most folks are good; a few aren't.

              Comment


              • #22
                Re: Predator Drones based on pirated, malfunctioning software?

                Somebody knock their heads together and lets move on!

                Comment


                • #23
                  Re: Predator Drones based on pirated, malfunctioning software?

                  Originally posted by TPC
                  Dang ... I am hopeless. Back to my crypto example.

                  Let's say I inform you that I have a program that computes what is called a message digest (a fancy phrase for a checksum or hash.)

                  Let's further say that I tell you exactly how my program works. It takes any input text, appends a specific random number that I have generated, one time, and then computes the SHA256 secure hash of the result. The only fact I refuse to divulge is that specific random number.

                  I then allow you to present any number of input texts you want to my program, and I provide, as fast as can be computed on whatever computer that you provide me the resulting output SHA256 secure hash.

                  You will never, in your lifetime or mine, be able to determine that random number by such efforts, and hence you will never be able to predict reliably what the output will be for any input that you have not already tested.

                  I would consider that to be a failure to usefully reverse engineer my program.
                  Your example is flawed, because you are not providing 100% of all possible inputs nor 100% of all possible outputs.

                  You are merely stating that I can enter any number of inputs and not be able to figure out the algorithm.

                  The two are not the same thing.

                  So, you first provide all possible inputs and outputs, then I'll figure out what's between

                  In the Predator drone, as well as all of the other examples, on the other hand, clearly reverse engineering IS possible and has been done.

                  Comment


                  • #24
                    Re: Predator Drones based on pirated, malfunctioning software?

                    Originally posted by c1ue View Post
                    Your example is flawed, because you are not providing 100% of all possible inputs nor 100% of all possible outputs.
                    I can't provide 100% of the inputs, nor can you receive them. That's the point. There are too many of them, (aleph null) I suspect, that would be required to determine the hidden random number and thus usefully reverse engineering the system.

                    That's why not all sufficiently complex systems (in this case a few lines of Python) can be usefully reverse engineered.

                    I have thus demonstrated the two points that I was making:
                    1. that you were wrong to claim that all systems can be usefully reverse engineered, and
                    2. that I was foolish to think I could ever get you to usefully admit that you were wrong.

                    Illud res quod erant futurus probo. (That's Latin for "Those things which were to be demonstrated.")
                    Most folks are good; a few aren't.

                    Comment


                    • #25
                      Re: Predator Drones based on pirated, malfunctioning software?

                      Originally posted by flintlock View Post
                      Somebody knock their heads together and lets move on!
                      What's the sound of two rocks being knocked together <grin>?
                      Most folks are good; a few aren't.

                      Comment


                      • #26
                        Re: Predator Drones based on pirated, malfunctioning software?

                        Originally posted by TPC
                        I can't provide 100% of the inputs, nor can you receive them. That's the point. There are too many of them, (aleph null) I suspect, that would be required to determine the hidden random number and thus usefully reverse engineering the system.

                        That's why not all sufficiently complex systems (in this case a few lines of Python) can be usefully reverse engineered.


                        I have thus demonstrated the two points that I was making:
                        1. that you were wrong to claim that all systems can be usefully reverse engineered, and
                        2. that I was foolish to think I could ever get you to usefully admit that you were wrong.
                        Illud res quod erant futurus probo. (That's Latin for "Those things which were to be demonstrated.")
                        Your argument is the equivalent of saying: "What am I thinking"

                        Since you cannot fully describe this either, neither can your brain be replicated by reverse engineering.

                        Again, you fail to understand the underlying principle: if a complete behavior can be described, then it can be reverse engineered.

                        If it cannot, neither then can reverse engineering occur - except for that portion which IS described.

                        Simply because you can write an open ended, undescribable phenomenon in a few lines of Python code is irrelevant.

                        You can as easily write a few lines to create a random number generator. This too cannot be reverse engineered, and for much the same reason - because there IS no behavior to copy.

                        Errors in reverse engineering or copying occur not in the described behavior - barring bad engineering - but in the lack of specification and/or test set.

                        Comment


                        • #27
                          Re: Predator Drones based on pirated, malfunctioning software?

                          Originally posted by don View Post
                          Old School

                          This meant war


                          New School

                          WASHINGTON — The C.I.A. has drastically increased its bombing campaign in the mountains of Pakistan in recent weeks, American officials said. As part of its covert war in the region, the C.I.A. has launched 20 attacks with armed drone aircraft thus far in September, the most ever during a single month, and more than twice the number in a typical month.

                          This means...something else

                          What else do you think it might mean, don?

                          It sure looks like increasing war with Pakistan to me.
                          Most folks are good; a few aren't.

                          Comment


                          • #28
                            Re: Predator Drones based on pirated, malfunctioning software?

                            Originally posted by ThePythonicCow View Post
                            What else do you think it might mean, don?

                            It sure looks like increasing war with Pakistan to me.
                            Pakistan is very vulnerable.

                            From a cocktail napkin perspective, I wonder IF Pakistan possesses a sustainable economy.

                            Combine the global financial crisis with the recent flooding and earthquake disasters in Pakistan and you've got some serious economic/financial problems that are kept at bay via considerable financial support from the US and China...as well as a reported 20%-ish theft of material transitting Pakistan for use in Afghanistan by coalition forces.

                            Without the foreign aid and theft would Pakistan currently have a sustainable economy?

                            So it's an environment that insurgent groups could easily leverage to wreak havoc.......as well as an environment that Pakistan's military leadership may feel compelled to create or encourage external blame or shift in focus.

                            I for one certainly hope the Commonwealth Games in India do not see a major attack by Pakistani ISI backed forces.

                            I'd happily bet anyone a case of good beer that the most likely next major conflict is between Pakistan and India....and as far as the next use of a nuclear weapon that will also be in either Pakistan or India in my opinion.

                            Comment


                            • #29
                              Re: Predator Drones based on pirated, malfunctioning software?

                              Originally posted by c1ue View Post
                              The article is about a widely used weapon system which appears to have been rushed into production, including use of a flawed and pirated software package.

                              Frankly a 42 foot error radius may not be that big an issue when 1000+ lb warheads are involved, but the fact that such a key component is literally halfa**ed ought to be of some concern.

                              Another example:

                              http://www.msnbc.msn.com/id/38857999/ns/us_news-life/
                              My point is that the article seems quite confused.

                              It implies there is only a single sensor or data source.

                              If you've got a bunch of overlapping sensor capabilities, does one sensor having a rather negligible error really matter?

                              If you've got 3 or 4 sensors that when combined offer a quite high level of accuracy, but with 1 possessing an acceptable error that can be effectively masked by the rest of the sensor suite......it quite possibly falls into the "good enough for operational use" category if the need is high enough.

                              It could be a question of a good but imperfect solution implemented today trumping a perfect solution implemented sometime next year.

                              Comment

                              Working...
                              X