I haven’t been able to write much the last few weeks, but I need to break this streak of not-writing. So here is a list of changes I’ve made to the site.
I’m now using Inter for code and figure captions. This is the third sans-serif I’ve tried, replacing first Open Sans and then Source Sans (plus who knows how many others that never made it into a published version of the site.) Inter feels like a Univers-style neo-grotesque, and I’ve always loved pairing Garamond and Helvetica, so let’s see if this one sticks.
Inter comes with suggested letter-spacing metrics for different text sizes. It’s generally true that smaller type reads better with positive tracking, and display type looks good with negative tracking, but I’ve never seen it set up as a formula like this before. In any case, I haven’t felt the need to deviate, so I’m applying letter-spacing as suggested.
Font processing was taking too long, so I’ve split it out into a separate program.
I have a new auditing step that warns me if any pages use glyphs that aren’t available in the font, to avoid unexpected fallbacks to system fonts. It’s found some issues, especially on The Collected MOON FACTS that I’m not sure what to do with yet.
There’s a new text processing step that finds sequences of figures, as in 123-456-7890, and wraps them so I can add a little bit of letter spacing. I got the idea from The Elements of Typographic Style, by Robert Bringhurst.
I’ve reworked how list markers work. Rather than being left-aligned within a fixed width box, they’re now right-aligned and use an EN SPACE (U+2002) to position them.
The inline numbers for sidenotes now use EB Garamond’s scientific inferiors rather than its true superscripts. They’re positioned a little lower and don’t stick out as much. I’ve also added a THIN SPACE (U+2009) before them.
I really like this approach of using the various whitespace characters instead of adding padding or margin. I got the idea from “Micro-Typography: How To Space And Kern Punctuation Marks And Other Symbols” by Thomas Bohm, and I’m really excited for my copy of Detail in Typography by Jost Hochuli, translated by Charles Whitehouse, to arrive.
I’ve reworked how dark mode images are generated. Instead of inverting the luminance channel, I’m inverting all channels and then rotating the hue 180°. Then, the output range is compressed slightly, so black becomes equivalent to the slightly-lighter-than-black of the background, and white becomes equivalent to the slightly-darker-than-white of the text.
Mechanically, I’m also now using color lookup table rather than passing transformations to ImageMagick. ImageMagick is useful, but challenging to use and debug. Currently, only the map on the Archives page uses this feature.
Examples for most of these changes can be found on the design pages, especially Typefaces and Typography (at the time I’m writing this post, it’s the only design page. I hope there will be more soon.) I wanted to avoid a situation where any demos on this page break because of future changes to the site.
I’ve been making progress on more complex layouts for figures, partly to support an article I’ve been writing, but mostly because working on the blog engine is somehow easier than so many other things right now. Don’t misunderstand me: this is escapism. But I think it’s kind of beautiful, too.