0
$\begingroup$

We have a tag with a fair amount of questions (probably more where the tag is missing). Looking at them, it might as well be (in line with , , , , ...).

But then, we have always made a point of not excluding practical questions, as long as they go beyond pure programming. Note that a similar discussion about and (which currently both exist) has not yet been resolved.

Can we expect questions that would be aptly tagged by and not by ? Do we want both tags (and retag theoretic questions to ), or only one? If so, which?

$\endgroup$
4
  • $\begingroup$ Can you make the automata-theory a synonym for automata? It would help when migrating questions from cstheory. (I can post a new meta question if you think that would be a more proper way of discussing/doing this.) $\endgroup$
    – Kaveh
    Aug 14, 2012 at 23:24
  • $\begingroup$ @Kaveh: I don't think this meta discussion has had a conclusive result, and I also don't think the migration path should be a major reason for choosing a tag. I guess you can revive the discussion by performing a minor edit on the question or adding another answer. $\endgroup$
    – Raphael Mod
    Aug 14, 2012 at 23:38
  • $\begingroup$ I didn't meant migration path. I meant when I migrate a question from cstheory it is normally tagged with automata-theory and the tag is removed after I migrate it to cs.se because there is no such tag here. So I have to edit the question to add the automata tag. I was suggesting something similar to what we do for ds.algorithms. $\endgroup$
    – Kaveh
    Aug 14, 2012 at 23:44
  • $\begingroup$ @Kaveh: I understand that, but that does not change my opinion. We have to decide whether we want to use automata or automata-theory before setting up any synonyms (afaik that is, kind of, irreversible?). $\endgroup$
    – Raphael Mod
    Aug 15, 2012 at 6:03

3 Answers 3

Reset to default
1
$\begingroup$

Can we expect questions that would be aptly tagged by automata and not by automata-theory?

I can't see how. "Automata" have a rather clear meaning of "finite state machine" which I never encountered outside of automata-theory.

Do we want both tags (and retag theoretic questions to automata-theory), or only one? If so, which?

Only one, for the sake of simplicity and not to trouble a hesitant asker. for uniformity or for lazyness that is quite important in CS. probably won't exclude any question, practical or not.

exists because database is a real tag on SO.

is clearly justified for most questions and probably should be used for some questions ( 1 1 1 1 ) but some are ambiguous ( 1 1 ) the problem may be that the user expect and not .

I'm no expert but may not be used properly.

$\endgroup$
5
  • 1
    $\begingroup$ '"Automata" have a rather clear meaning of "finite state machine"' -- I have to strongly disagree. "Automata" also includes pushdown automata, Büchi automata, LBA, ..., in short all kinds of automata. "Finite state machine" is more restricted. $\endgroup$
    – Raphael Mod
    May 23, 2012 at 6:42
  • $\begingroup$ @Raphael: don't they have all a finite set of states? $\endgroup$
    – jmad
    May 23, 2012 at 7:11
  • 1
    $\begingroup$ Yes and no. There are only finitely many control states, but given a stack or tape, you have infinitely many states (control state x stack content). "Finite state machine" relates only to NFA and variants with output; see also Wikipedia. $\endgroup$
    – Raphael Mod
    May 23, 2012 at 7:15
  • $\begingroup$ @Raphael: fair enough. But my point is really that I don't see how automata can be used where automata-theory can't. $\endgroup$
    – jmad
    May 23, 2012 at 8:16
  • 1
    $\begingroup$ Compiler people use automata all the time; I am not sure that what they do is (called) "automata theory". They seem to be more concerned with efficient implementations. But I am no expert; maybe @AlextenBrink knows more? $\endgroup$
    – Raphael Mod
    May 23, 2012 at 8:23
1
$\begingroup$

I think that we can generally drop the distinction between "X" and "X theory". We should pick one, be consistent, and make a point of not interpreting one to exclude the other. Personally, I'd like to see consistency (i.e., either all "theory", or none), but realistically it's not all that bad if we mix and match, e.g., "automata" and "graph theory", or "automata theory" and "graphs".

$\endgroup$
1
  • $\begingroup$ I think there is a subtle difference in meaning; I'll add an answer. $\endgroup$
    – Raphael Mod
    Jun 25, 2012 at 7:51
1
$\begingroup$

can include practical aspects of automata that might not fit into , such as strategies for implementing some automaton. The difference is subtle enough so that we can not reasonably keep both. Therefore, I suggest that we keep the more inclusive as is.

$\endgroup$
2
  • $\begingroup$ I think there's an even more subtle distinction at play here: the distinction between a question's being in a tag, and a question's being about a tag. The question "What are the philosophical implications of computability?" is a question about computability, not in it, but I think it could be tagged with computability-theory. If I understand your position correctly, such a question would not be tagged as such, since computability-theory provides no tools to answer it... right? Or is it something weaker/stronger than that? Or am I off-base? $\endgroup$
    – Patrick87
    Jun 25, 2012 at 15:43
  • $\begingroup$ @Patrick87: I think that analogy works. Automata have the additional "problem" that there are applications. $\endgroup$
    – Raphael Mod
    Jun 25, 2012 at 18:30

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .