Xref: utzoo comp.lang.c:11375 comp.arch:5541 Path: utzoo!attcan!uunet!husc6!bloom-beacon!bu-cs!purdue!decwrl!labrea!sri-unix!garth!smryan From: smryan@garth.UUCP (Steven Ryan) Newsgroups: comp.lang.c,comp.arch Subject: Re: Self-modifying code Message-ID: <989@garth.UUCP> Date: 16 Jul 88 21:41:47 GMT References: <5262@june.cs.washington.edu> <260@thor.wright.EDU> Reply-To: smryan@garth.UUCP (Steven Ryan) Organization: INTERGRAPH (APD) -- Palo Alto, CA Lines: 32 >As to the semantics of self-modifying code, isn't it identical to >a Turing machine's or am I missing something? No. The state mechanism of a Turing machine is fixed and finite. To show equivalence, it is necessary to show that self modifying code does not increase the number of states. Usually, people just appeal to Church-Turing and assume all the states are there, somewhere. For something like a load and go compiler, since every possible program is not hidden somewhere in the compiler, it would appear to escape CT. However the compiled program can be regarded as a transformation of the input tape into another region of the tape and its apparent execution as an interpretation of the tape by the compiler/loader/cpu state machine. Then again, if you believe really hard and clap your hands..... sm ryan ---------------------------------- Hugh: Why does he carry a sword? Fred: He has delusions of grandeur. him: Uh, Fred... (SWOOP) Fred: YEEK him: ...before you make a value judgement about a person's delusions... (poink) Fred: AWK! My head feathers! him: ...you'd better be absolutely sure they ARE delusions. (JAB) Fred: OUCH! -- Odd Bodkins