For instance:

• the datetime class lives in the datetime module; careful with your imports, you will want both!
• the datetime module has a time class, but there’s also a time module
• the documentation talks at different places about “time tuples” and also about “time.struct_time” objects; they’re the same thing
• to convert a unix timestamp to a time tuple representing UTC time, you use time.gmtime(); for the reverse conversion, you use calendar.timegm() (this is the only place you’ll use the calendar module)
• to get now you use:
  • time.localtime() if you want a timetuple
  • time.time() if you want a unix timestamp
  • datetime.now() if you want a datetime object
  • time.gmtime() if you want a timetuple representing UTC time
  • datetime.utcnow() if you want a datetime object representing UTC time
  • a crowbar if necessary
• to build a datetime object from a time-tuple, you can set the fields one-by-one… or you can slice the time-tuple and unpack the list: datetime(*a_time_tuple[:6])

In other words, there’s a lot of fiddly detail to remember. So I made a chart.

Click through for (non-blurry) PDF.

It’s formatted for a4 paper and prints nicely. I used the incredible LaTeX drawing system TikZ. (On my screen the arrows have a little glitch at start and finish, but it doesn’t appear in my printout. Possibly a TikZ bug? Not sure.)

Note that it doesn’t show everything you can do; apart from the more egregious omissions (like timezone handling on datetime objects, which actually seems quite straight-forward but didn’t fit nicely into the design), I didn’t include arrows for nonsensical operations. For example you could feed a UTC time-tuple into time.mktime() (which takes a localtime time-tuple to a unix timestamp), but the result would be nonsense: the number of seconds since a date similar to, but probably not quite the same as, the unix epoch.

Summary:

• XeTeX gives LaTeX access to non-TeX-installed fonts, and fontspec makes it easy.
• XeTeX also lets you write in unicode; biblatex will play nicely if you use biber instead of bibtex as the backend.
• fc-list lists installed fonts (giving the names that XeTeX needs).
• fontforge will display every glyph in the font, nice and big (and you can print the result).
• fontmatrix has a very slick interface for font-browsing, from “tag-this-font” to “inspect-this-glyph”.
• All of these except biblatex and biber (which are still in beta) are packaged for various Linux flavours; biblatex installation is easy, for biber I vaguely recall that I installed via CPAN, which is rather a pain. There may be easier alternatives by now though.

XeTeX lets you use all your fonts in LaTeX

XeTeX is a TeX replacement; you compile your document with xelatex if you’ve been using pdflatex, and nothing much changes apart from that. But behind the scenes quite a lot happens; among other things, you get to use the fontspec package, which works with the font names that the rest of your Linux system knows about. (Another, less positive, feature of the move is that you XeTeX is incompatible with microtype. Can’t have everything…)

\documentclass{article}
\usepackage{fontspec}
\usepackage{xunicode}
\usepackage{xltxtra}
\setmainfont[Mapping=tex-text]{Verdana}

If you pick a font with italic and bold alternatives, \textit{} and \textbf{} and so on will just work. Just like that. Which is nice.

If your font has Greek and Russian characters (Greek text characters, not a math alphabet), you can just type ‘em:

Lorem ipsem and so forth; ο ξυπόλητος πρίγκιπας; 
зону большого над, языке а; 
italic: \textit{in Latin και στα Ελληνικά},

Which is nice.

(I don’t know what the Russian says, nor do I remember where I got it. The Greek says “The barefoot prince” and “and in Greek”.)

That Mapping=tex-text says that XeTeX should recognise the old LaTeX style for quotes, long dashes, and so on. But you don’t have to use it: you can type them directly as unicode characters (which is nice).

When he goes---``Hello World!''\\
She replies—“Hello dear!”

And if the font you’re using has fun OpenType features (and that’s the whole point, right?), you can access them easily too:

\fontspec[Ligatures={Common, Rare}]{Impact}
\fontsize{12pt}{18pt}\selectfont Questo è strano assai!
\fontspec[Numbers={OldStyle}]{Impact}Old style: 1234567\\
\fontspec[Numbers={Lining}]{Impact}Lining: 1234567

See the fontspec documentation for much, much more. (If your pdf seems to have half its text missing, there’s something wrong with your pdf rendering library. I don’t remember what I did to fix this though; possibly just updated, or possibly changed to a different pdf viewer.)

biblatex with biber can handle unicode bibliographies

So part of the reason I was originally playing around with all this was to help Olga write documents mixing Greek and Latin characters, with both Greek and Latin in the bibliography, with a minimum of hassle. It turns out that bibtex doesn’t like unicode, in particular it doesn’t know what to do with Greek. But biber, which is being developed as a backend for the biblatex package, is designed for unicode from the ground up. Both biber and biblatex are beta release software and there are still a few kinks to iron out, but they can already do a lot.

List installed fonts with fc-list

If you’re anything like me, the first thing you’re going to want to do once you see how XeTeX and fontspec work is try out lots of fonts to see what you’ve got installed that you could use.

If you just want to know names, fc-list is your friend. Be prepared for a lot of output though…

Here’s a handy trick: use the pgffor package (part of the fantastic graphics package pgf) and you don’t have to constantly retype Lorem ipsem:

\usepackage{pgffor}
...
\foreach \f in {
  Droid Serif,
  Alfios,
  Alexander,
  Dustismo
  ...
  URW Bookman L%
}{
  \noindent\setmainfont{\f}\f:
  Lorem ipsem and so forth.
  \par\medskip
}

Viewing fonts (and individual glyphs) with fontforge and fontmatrix

Somewhere along the line I picked up a free font with heaps of fleurons, and a font of 300 ampersands each by a different designer (yeah, a gimmick, but the proceeds went to charity). For both of these, I wanted a way to see all the glyphs the font contains, large enough to appreciate them as individual characters, and also with the codepoint information so I know how to get access to whichever one I might want to use.

The internet was silent on how to achieve this (or my google-fu was weak), until JanekZ helped me out on typophile. It seems fontforge is the only way to get a conveniently-formatted listing of every glyph the font contains (“Encoding >> Compact”); it can generate a pdf for printing but the pdf doesn’t contain codepoint information. But for simply browsing, fontmatrix is a thing of beauty. Unfortunately it seems to be buggy: “print” doesn’t produce any output, and “view all mapped glyphs” refuses to scroll. But the stuff that works works beautifully.

(Thanks to Micha — I discovered this while helping him tidy up his texmf tree. And yes, I guess this is less down to (La)TeX and more down to my distribution, details jargon technicalities disclaimer.)

For instance, you can scan in somebody’s photocopy of a paper from 1978 using the office printer, which emails you a pdf of the scan, and run it through this file to get a single page of the original on each page of the pdf:

\documentclass{article}
\usepackage{pgffor}
\usepackage{pdfpages}
 
\begin{document}
 
% Paper starts on a right-hand page of the original book,
% we don't need the final page of the paper before it.
\includepdf[pages=1,
            angle=90,
            trim={0 0 0 135mm}
            ]{Kamp78SemanticsVersusPragmatics.pdf}
 
% For each double-page spread, rip out the left hand then
% the right hand side. The trim was set by eye.
\foreach \page in {2,...,17} {
  \includepdf[pages=\page,
              angle=90,
              trim={0 135mm 0 0}
              ]{Kamp78SemanticsVersusPragmatics.pdf}
  \includepdf[pages=\page,
              angle=90,
              trim={0 0 0 135mm}
              ]{Kamp78SemanticsVersusPragmatics.pdf}
}
 
\end{document}

(This also gives me a chance to show off try out my new code-formatting plugin. Pretty!)

Things to like about it:

• It’s damn elegant.
• It’s inspired by Bringhurst’s The Elements of Typographic Style, which remains the most eloquent and convincing argument for attention to typography I’ve ever read.
• Namecheck of the booktabs documentation (on why the tables don’t have vertical lines), which I’ve also tried to promote.
• “And no: the lines are not too short.”
• Sidenotes. (And footnotes. But sidenotes. Lovely.)
• The author wants postcards. And asks that you send complaints/suggestions/questions on the postcards, not by email. Nice filter.

There’s also the Ars Classica package, which patches ClassicThesis for a slightly different look. That’s all in Italian, but the code is almost readable anyway (much more LaTeX than TeX) which is encouraging for the prospect of coming up with a patch myself.

Reckon I’ll be sending a postcard.

I don’t usually do this, since my talks tend to rely a lot on … well, talking. The slides without the patter can vary between cryptic, irrelevant, and just plain wrong. (I quite enjoy letting the audience absorb some choice piece of wisdom on a slide then saying “This isn’t true.”) This time I’ve taken the trouble to add notes to each slide, explaining what I had to say and why I wanted to say it. (I’ve incorporated a bit of the discussion as well, but resisted the temptation to improve things with the benefit of hindsight.)

For others interested in how to do this, Beamer has a \note{} macro. If you issue it in a frame, its contents get collected into a notes “slide” that comes after the frame (it does the right thing with frames containing transitions: every transition semi-frame gets the same notes slide, so the notes stay the same while the frame changes). Do this for every frame and you get an alternating sequence frame/notes/frame/notes… The next thing to do is put each frame/notes pair on a single page. I couldn’t find a tidy (and free) way to do this, so I got my officemate to print-to-pdf on his Mac, which did the job. (That’s why the notes are under the slides, not beside: we couldn’t figure out the options that would do it the other way.)

Some questions for the laziweb:

1. Any good pdf manipulation programs, free for linux? For (for example) arranging several consecutive pages onto one page (like you can do with pstops).
2. How does the over-and-under positioning work (if anyone is actually reading the slides, that is) for you? I have a weird screen ratio that makes it perfect (I lie my screen on its side, for comfortably reading a4 pages), but I guess it’s less handy in 800×600. Do I need to worry excessively about this?
3. Beamer produces really big pdfs — 3.1M for that presentation. Any tips on cutting them down to size would be great.
4. I couldn’t seem to get beamer to make the notes slides plain, despite following what I thought were the instructions in the manual. Has anyone made this work, and if so how?

I’ve got a basic system that works pretty well, and now it has its own page on my site. Any suggestions would be welcome.

Snapped from the train on the way back from Paris. I guess I’ll post photos of that trip at some point too.

Why not?

The spacing is wrong. It’s usually only slightly, subtly wrong. But around an ‘f’, it’s grotesque. The reason is that in math mode TeX assumes individual letters represent individual variables, so it doesn’t use the kerning information that would otherwise tell it to close up the spacing around the ‘f’, since its ascender and descender stick out sideways. This behaviour is fine if the ‘f’ is a function variable, but it doesn’t work in ‘infold’, for instance.

Aren’t you being a bit picky?

Nope. I have in front of me page 307 of Type Theory & Functional Programming, which contains a function definition beginning in f old f st… or possible inf old f st, or (as in fact intended) infold f st, where infold is the function name, f is a function argument, and st is a state argument. The link takes you to an online PDF copy, go check it out. I’ll wait.

See it? If you look closely you might notice some weird spacing on the right-hand-side also — it would be truly picky to call that out, but the oddity on the left really interferes with the interpretation. (Why that page? It happened to come out on top of a signature as I was folding the book up for binding. And 307 is a nice number, don’t you think?)

What should it look like then?

Simple answer: use \mathit{} and \mathrm{} to get properly spaced words, italicised or roman, in math mode. (Thanks Tom for the reminder.)

Slightly more complicated and pedantic answer: it’s worth carefully considering whether you need italics at all. Traditionally these are used for variable quantities or entities; a function variable f which can take many values, a state variable st likewise (although I’d prefer s), and so on. It’s flatout wrong to use italics for well-known functions like sine and cosine, which is why LaTeX provides primitives \sin and \cos which typeset these in upright roman. So if you’re defining a function “infold”, which will always have the same definition, it might make sense to put that in roman too. Another case which crops up often (in the same line of TT&FP, for instance) is an equality symbol subscripted with ‘df’ or ‘def’ or similar, for “definitional”. Here there’s really no reason to use italics at all, and of course the spacing is again messed up by the ‘f’.

All right, now I am being picky.

Can I go now?

Yes yes, run along.

Oh, one more thing. I don’t mean to pick on Simon Thompson, who after all has written a textbook that I just downloaded and printed and expect to learn quite a lot from. I use that example because it shows that it’s not just students writing term papers who do this stuff, it crops up also in books with editors, that get published and that people pay money for. And it’s so easy to avoid, once you know there’s a problem. So do try, please? You’ll do wonders for my blood pressure.