I apologize in advance for the long wall of text; there is probably some way to condense this question to half the size without losing anything important, but I'm not sure which half to remove.

I recently finished writing my thesis, and this involved some last minute freaking out about the formatting of the front matter, since I kept putting off until "later". (I've seen a few horror stories on this website where some universities take a ruler to your thesis to make sure that your margins are within spec or something. The person at my school is fairly reasonable about such things. I remember showing him my attempt at the signature page and he said "perfect" despite the fact that it looked noticably different than the page from the guidelines. So it seems that I was freaking out for no good reason. This did not stop me from doing so for several more weeks. In some sense, the fact that I'm posting this question here shows that I haven't really stopped.) I could have used a template that my advisor had lying around, but it had various misfeatures that would have driven me crazy; what they were is probably irrelevant here. The point is, I know someone who did use, and was bitten by some of those misfeatures and needed my assistence to recover. Which makes me think I'm not the only one who would like something better. (In particular, simpler.) So, I resolved to write my own template in the hopes that someone else would find it useful someday.

I want to base it one of the standard classes (as opposed to something fancy like memoir or something in the KOMA bundle), and the only ones that support chapters are report and book. The question becomes: which one? When I first began working on my draft, I was using the report class since it had a built in abstract environment (abstract is required), but eventually I switched to book since the built in abstract environment was not doing what I want and thus had to hand format it. Now that I'm making a new template, I'm revisiting this decision. The trouble is that these two classes are so similar that it's hard to choose between them on any basis at all. Nevertheless, I compiled a list of differences between the two that seems relevant to this decision. Note: I'm ignoring differences that are only apparent in compatibility mode: ie \documentstyle instead of \documentclass. (For instance, book and report don't have the same text width in compatibility mode, but with LaTeX2e, they do.)

One argument in favor of book is that I've actually used it successfully myself, so I can say that it's proven to work. This argument seems pretty weak, since the classes are very similar and in any case I intend to test this on my thesis to make sure everything is still fine.

I've also done a bit of online research, and most How To Thesis Guides seem to favor report instead of book, although not all. But most of them give little to no justification for this choice.

In any case: the differences, with pros/cons:

The book class has support for the distinction between main matter and front matter, which the rules require. On the other hand, this seems like the kind of thing that can be 90% faked by \pagenumbering.

The book class is two-sided by default, the report class is one-sided. The rules require a single-sided thesis. On one hand, defaults are easily changed, but maybe I should take a hint that the class with defaults closer to what I need is the right one to use?

At 12pt base size using one-sided layout (I haven't checked the other size too carefully) there are some (very) minor differences in layout: \headsep is a few points smaller with book than report. (25pt vs 0.275in, or roughly 19.87pt.) Similarly, \marginparsep is a bit smaller in book (7pt vs 10pt), and \marginparwidth is thus 3pt wider in book (47pt vs 44pt). In practice, these are both much too small to make \marginpar useful, at least in my opinion (an inch or two is enough). The two important parameters are \textwidth and \textheight, and these are identical.

The report class has a built-in abstract environment. It's sufficiently different than what is neeeded that I have no temptation to reuse pieces of it. I can't decide if this is a point for or against using report. On the one hand, what's the sense of pulling in definitions I don't need. On the other, this seems to suggest that reports are the kind of thing that contain abstracts, while books are not, so in some vague sense report is the more sementically correct choice.

Slightly different titlepage environment. (book has an extra \cleardoublepage) This environment does crazy things to page numbers and I intend to avoid it. I have no idea who thought these page number games were a good idea. (Leslie Lamport probably?)

Different defaults for the page style: report uses plain, while book uses headings. Doesn't matter, both are not right. It turns out that myheadings, with no customizations does the right thing here. (Really, no setting marks or anything!)

Any thoughts? I can't help feeling like these two classes should be (or should have been; too late now I guess) merged into one; they are too similar.

  • 1
    Yes, very similar. And chances are pretty high you need to do customizations. Only difference that really matters, is the frontmatter stuff that is not around with report. I think you are freaking out about nothing. – Johannes_B May 26 '16 at 6:55
  • Note though, that providing a template is a big responsibility. – Johannes_B May 26 '16 at 6:56
  • 7
    Yes, you are overthinking: use book and its options. – egreg May 26 '16 at 6:58
  • 1
    book report and article are all generated from the same file with only a few differences, so it's not a case of needing to merge them they start as one file. – David Carlisle May 26 '16 at 7:00