Should we have a canonical/FAQ answer regarding minimization of "Boolean" (i.e. propositional) expressions?

Question

It seems that whenever there are some students taking a digital logic class, we get a lot of those relatively low-effort questions on SO, EE.SE and here. As far as I can tell, the reaction has been most negative on EE.SE, e.g. see this (that question is probably among the better ones on this topic.) I found that even repeatedly telling people in comments they can use this or that software to check their answers is tedious enough.

So perhaps we should have a canonical answer here with maybe an example and indication how to check work via software? Some questions are just "check my work" from people who have learned the basic minimization concepts like the one linked above, but others are even more basic, asked by students who are tasked to minimize via identities and who are yet unaware of minimization theory. At the other end of the spectrum, please note that sometimes questions ask for gate-number minimization (aka multi-level minimization), which a different/advanced topic, but often enough confused with the CNF/DNF minimization by people doling out answers and even by some voters (on SO anyway); this latter topic is rather more difficult to cover in a reasonable-sized FAQ entry, so perhaps just pointing out the difference and some refs would suffice for that.

Answer 1

A canonical answer would not be useful to this topic. Let's focus on the particular topic of questions asking about the minimization of a particular boolean expression, because, outside of this, we cannot anticipate, and have no interest in anticipating, what sorts of questions people might ask. I see two broad classes these questions might fall into:

• Those With Context (e.g. questions asking to verify a solution, clarifying particular steps, seeking alternate methods, finishing a partial solution etc.): These questions are best handled by giving instruction particular to the context presented in the questioner's question. Sure, they might know the broad rules, but not necessarily how to advantageously use them, or maybe not always recognize them. It's not going to help them to see an abstracted question - and frankly, it's kind of insulting to close a well-written and specific question as a duplicate of one which is the same topic with all the context stripped. Even if the question is just "Is this answer, after these derivations, correct?" one can usually write a better answer than, "No," as you might get from software.

• Those Without Context (e.g. questions including only a problem statement): I would assume any questioner would likely have seen similar content to whatever a canonical answer might contain - it'd be kind of mean for a lecturer to give such problems without giving any rules for manipulations. Instead of closing as a duplicate of something which is likely unhelpful to the OP and anyone else who sees the problem, we could leave a comment asking whether the OP is familiar with general procedures for such problems/the solution of similar problems and if so, what their particular trouble is. Even if we do have useful abstract duplicates, I think closing for that reason is only likely to work if it comes with the disclaimer of, "...but if this doesn't help you, you should improve your question by adding more context" - and in this particular case, given how mechanical the process tends to be (but how it is not necessarily best done by a uniform process, even if it can be), I doubt that many people will find a link to some examples and automatic checking would be of much help.