Turing machine continuity w/ other definitions

[BassP97]

Hi all! Sipser (and other TOC sources) use a 7-tuple definition of a turing machine:

1. Start State
2. Reject State
3. Accept state
4. Transition function
5. Tape Alphabet
6. Input alphabet
7. Set of states

The open logic chapter on computation, in contrast, uses a 4-tuple definition. Is there a reason for this (maybe a difference in philosophy/computer science's respective disciplinary norms?) or should I (or someone else) rewrite the definition using the more conventional 7-tuple?

[rzach]

There is no single universally accepted definition. Eg, the OLP definition uses a one-sided tape and a tape-end marker, and allows the tape head to not move (so it is like the definition in Papadimitriou's Complexity Theory, for instance). Since the OLP (as of yet, in any case) has no content that would require separate input and tape alphabets, or a baked-in requirement that every machine has an accept and a reject state, I don't think that's necessary or useful. The sections comparing definitions could be expanded though. Let's keep the simpler definition for now, I'd say.