From: Charles A. Crayne on 1 Oct 2005 01:42 On 30 Sep 2005 16:40:45 -0700 "randyhyde(a)earthlink.net" <randyhyde(a)earthlink.net> wrote: :If you take *all* the possible bits in a :hypothetically maxed-out x86 system (including all secondary storage, :registers, memory, and anything else that can be set one way or the :other) Using your definition, not only is the number of machine states inconceivably large, it is also indeterminate, and varies over time. For example, my machine, like many others, has a full-time Internet connection. In theory, every machine state in every every machine which is also connected, or which could possibly be connected, to the Internet could be affected by a program on my machine. In addition, my machine, like most others, has removable media, and during the execution of any given program, could have an indefinite number of media swaps over the lifetime of a given program. Finally, my machine, line a number of others, has programs which run for a month or more. Since I can upgrade my machine with additional resources while these programs are running, the total number of machine states may grow over the lifetime of a given program, from this cause, as well. All in all, I don't consider your definition to be very useful. -- Chuck
From: wolfgang kern on 1 Oct 2005 04:01 <randyhyde(a)earthlink.net> wrote: | I guess you don't understand what a machine "state" is. So how could I write a disassembler which knows all machine-states in all possible variants and modes and various output options by just using 20KB then. | I'd suggest you study automata theory sometime. I better leave theories to were they belong to. All machines controlled by 'my' programs work very well without it. | ... (IIRC, the size of the universe is somewhere around 10^78 | particles or so, far less than the two to this very huge number). | So, Wolfgang: no, there is no way you could possibly even imagine all | the states that are possible to create with an x86 system. Mother always said: "never laugh about handicapped" ... So I follow this advice yet. So we all can see it as the obvious fact: Randy Hide's universe is limitied to 10^78 particels. (a very tiny world indeed). Mine is infinitive. And how many particles are to be found on just our planet? __ wolfgang
From: randyhyde@earthlink.net on 1 Oct 2005 13:52 Charles A. Crayne wrote: > On 30 Sep 2005 16:40:45 -0700 > "randyhyde(a)earthlink.net" <randyhyde(a)earthlink.net> wrote: > > :If you take *all* the possible bits in a > :hypothetically maxed-out x86 system (including all secondary storage, > :registers, memory, and anything else that can be set one way or the > :other) > > Using your definition, not only is the number of machine states > inconceivably large, Of course it is. That's the point we're trying to make you see here. > it is also indeterminate, and varies over > time. No it doesn't. A machine is a machine. If you add more states, you've got a *different* machine. As I said earlier, you don't seem to understand what is meant by the term "machine" in computer science. > > For example, my machine, like many others, has a full-time Internet > connection. In theory, every machine state in every every machine which is > also connected, or which could possibly be connected, to the Internet could > be affected by a program on my machine. Yes, it could. And each time you change the state, you have a different machine. But it doesn't really matter, even if we take a bare-bones PC with smallish amount of memory, the number of possible states is so incomprehensibly large that we couldn't possibly write a program to process all those possible states within a reasonable amount of time. As I believe it was Phil who said: there just aren't that many clock ticks in the universe. > > In addition, my machine, like most others, has removable media, and during > the execution of any given program, could have an indefinite number of > media swaps over the lifetime of a given program. The state of your machine includes all possible media that can be connected and used at any point in it's lifetime. How's that? All you're doing is increasing the number of states, making the problem even more intractible. But as I already pointed out, the number of states in a bare bones PC is already so astoundingly large that the problem is, for all intents and purpose, impossible to solve. True, in theory it *is* solvable, but you'd never benefit from the answer because the universe would implode upon itself long before your machine computed the result. > > Finally, my machine, line a number of others, has programs which run for a > month or more. Since I can upgrade my machine with additional resources > while these programs are running, the total number of machine states may > grow over the lifetime of a given program, from this cause, as well. And you have a different machine. Or, we can simply consider the *maximum* number of states your machine has over its lifetime to be the number of states the machine possesses. This doesn't change, one bit, the fact that the problem is intractible. > > All in all, I don't consider your definition to be very useful. Useful for what? You (and a few others) are the ones arguing that because the x86 is a finite state machine, we *can* solve the halting problem for programs written on that machine. I'm simply pointing out that, in theory, this statement is correct; but as you've so concisely stated, that fact isn't very useful. That's way it's a relatively safe assumption to pretend that the x86 *is* an infinite machine and assume that the theory applied to Turing machines applies to x86 systems. True, in theory this isn't quite accurate; but in practice it's close enough. Cheers, Randy Hyde
From: randyhyde@earthlink.net on 1 Oct 2005 13:58 wolfgang kern wrote: > <randyhyde(a)earthlink.net> wrote: > > | I guess you don't understand what a machine "state" is. > > So how could I write a disassembler which knows all > machine-states in all possible variants and modes > and various output options by just using 20KB then. Simple, your disassembler won't properly and automatically disassemble all code fed to it. The undecidability of code/data differentiation means that we can't make this decision for *every* possible input. It does not imply that if we limit ourselves to some subset of possible inputs we cannot do the job, nor does it imply that if we are willing to live with less that 100% correct output the job cannot be done. No doubt (as you've admitted in other posts in this thread), your disassembler doesn't handle all possible inputs. > > | I'd suggest you study automata theory sometime. > > I better leave theories to were they belong to. > All machines controlled by 'my' programs work very well without it. Ignorance is such sweet bliss, eh? > > | ... (IIRC, the size of the universe is somewhere around 10^78 > | particles or so, far less than the two to this very huge number). > > | So, Wolfgang: no, there is no way you could possibly even imagine all > | the states that are possible to create with an x86 system. > > Mother always said: "never laugh about handicapped" ... > So I follow this advice yet. > > So we all can see it as the obvious fact: > > Randy Hide's universe is limitied to 10^78 particels. > (a very tiny world indeed). Hmmm... I guess you have no concept of just how big 10^78 really is. > > Mine is infinitive. Whatever that means. Certainly you don't live in a universe with an infinite number of particles. > And how many particles are to be found on just our planet? I don't know the answer to that one, but I suspect you'd find the answer with a quick Google search. Cheers, Randy Hyde
From: wolfgang kern on 1 Oct 2005 15:48
<randyhyde(a)earthlink.net> wrote: | > So we all can see it as the obvious fact: | > Randy Hide's universe is limitied to 10^78 particels. | > (a very tiny world indeed). | | Hmmm... | I guess you have no concept of just how big 10^78 really is. Let's see if I can imagine it: a CPU with 256 address lines (just about five times AMD64) can address every particle in your universe. :):):) | > Mine is infinitive. | | Whatever that means. Certainly you don't live in a universe with an | infinite number of particles. I do. Regardless of Your believe. ... Show me the border of the universe ... | > And how many particles are to be found on just our planet? | | I don't know the answer to that one, but I suspect you'd find the | answer with a quick Google search. Why don't 'you' try it? __ wolfgang |