Let's assume that you typeset a computer-science book in German. The book will later be adapted to Springer's svmono.cls (see LaTeX templates in https://www.springer.com/gp/authors-editors/book-authors-editors/resources-guidelines/rights-permissions-licensing/manuscript-preparation/5636) and the PDF (and only the PDF) will go to Springer. You dislike UTF-8, since it incurs issues with certain packages such as listings (which can be worked around, but at a huge time cost: you'd escape a whole deal, so you might as well typeset your programs directly in LaTeX, and writing a whole parser in Python for minted is not really doable in the time you have). So, for a German book, you have a choice of two encodings: essentially, Latin-1 and CP437 (the German variant cp437de with ß as Eszett). You are well aware of Latin-1 (and that's the current encoding used for the book source). But you've never typeset anything in CP437. However, when looking at the codepage of CP437, it seems to offer you characters that occur more often in the book (e.g., some Greek letters and some maths) than it is the case for Latin-1. Nobody except you is going to see the sources.

Said that, does anyone have an experience of typesetting German computer-science books in CP437 that go to Springer later? If so, any advice? Would switching from Latin1 to CP437 pay off?

  • While I don’t have experience doing that, I’d suggest that, if you do, you use selinput. Then, your document should Just Work even if you change the encoding later. – Davislor Feb 2 at 1:53
  • For what it’s worth, my own personal preference is to use unicode-math when I can and legacy encodings when I have to, but many people disagree, and you stated your reasons why. – Davislor Feb 2 at 1:54
  • 5
    I strongly advise against using an unusual encoding unless you are absolutely sure that you will be the only one that will have to handle the tex files. If there is the possibility that you will need external support which involves the real files (e.g. here on tex.sx) use utf8. – Ulrike Fischer Feb 2 at 8:15
  • According to the documentation, it accepts the same encodings as inputenx. – Davislor Feb 2 at 22:27
  • 3
    But I agree with Ulrike Fischer (who, unlike me, has experience writing texts in German!). – Davislor Feb 2 at 22:29

I don’t know what Springer, specifically, requires. My recommendation is that, if you’re going to submit in some old, niche encoding, you load selinput. Make sure CP437, Latin-1, Latin-9 and UTF-8 are in the list of possible encodings.

That way, even if your editor converts the files to an encoding more convenient for them, your document should still compile.

  • 1
    @user49915 Thanks for pointing out my typo! – Davislor Feb 2 at 22:26
  • @user49915 I really have to advise you against choosing a code page from 1981 because it has “basic Greek.” It has an incomplete smattering of Greek letters that aren’t correctly-slanted and you probably would enter as \alpha and so on anyway. If you genuinely want to enter math symbols in your character set, you want UTF-8 with unicode-math. – Davislor Feb 3 at 0:38

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.