# Should we evangelize “correct use” in our answers?

This came up in the comments of this question. The OP asked how to fix one thing and gave an example where the problem arises (vectors with different components have different height). In that example the OP wasn't careful about other things: text in mathmode was not upright, and some braces were not scaled to encompass their contents.

In my answer I concentrated on the question at hand: how to make the vectors have the same length and I ignored the "problems" that the OP's MWE had. some of the comments I got seems to suggest that I should have corrected these issues in my answer (which used the MWE and fixed the relevant part).

What should be our approach? when someone asks about A and has issues B & C in the example, should we fix A, B, and C, or should we answer the question as it is posed? Perhaps we could comment on the original Question with regards to B and C instead of correcting it in the answer? Should we have a unified approach to this?

This is a judgement call. It may be that in their haste to provide a minimum working example, the questioner got sloppy. It may be that they really got it wrong.

The most important thing is to answer the actual question. The danger in correcting B and C is that the key change needed to correct A may get lost. However, I don't think that B and C should be ignored either: correcting those will also help the person asking (and anyone else who stumbles on the question). In an ideal world, I would probably try to phrase my answer like this:

Here's how to solve A:

Incidentally, I found that I got a nicer-looking result if I also did B and C.

Of course, that takes a little longer to do and if time and effort is short, just do A. The "incidentally" can also be a comment on the original question (or your answer).

If answering a question that wasn't asked (as in, if correcting something that the questioner didn't ask about), the most important thing is to be polite, which usually (and particularly when it comes to style) means: don't be dogmatic.

• The care required to achieve what you advocate would probably cause me to not bother answering at all. But OK. I see you point. – Yossi Farjoun Nov 1 '10 at 11:04
• @Yossi: Note that I said: "if time and effort is short, just do A.". If you are already inclined to point out B and C, then doing so politely is not going to take much time (I hope!). If it is easy to distinguish A in amongst B and C then you could say: "Here's how I would do it ... incidentally, I also did B and C.". – Loop Space Nov 1 '10 at 12:22

In this case I would just silently “correct” the other things without drawing special attention to it. So that the answer contains what you would type in a document without directly criticizing the OP. (I only added a comment to the question, because the comment might be misleading.)

If the issues B and C are more important than just missing \lefts I think it is useful to tell so somewhere. Depending on how directly they are related with issue A this might be in a comment or in the answer.

There are two distinct questions here.

1. Should we evangelise proper LaTeX usage? (prefer \[ to )
2. Should we evangelise good typography? (don't typeset your thesis in comic sans)

I think the answer to both is "yes but..." and endorse the suggestions in the other answers, which seem to apply to both

• Do people really try to typeset their thesis in Comic Sans??? – SamB Dec 17 '10 at 1:54

Yes, I agree with Andrew that the one thing that must happen is that the person's question should be answered. The rest is optional, but nice if you have time. Presumably the person is using *TeX because he wants more beautiful documents, so he would appreciate some typographical advice.