Open Historical Type Revivals
On the occasion of doing a bit of a design refresh on this blog, using the amazing new Baskerville Punch typefaces, I was moved to write about the larger world of historical type being re-drawn, enhanced, and released online under open licences.
If you read this blog (or otherwise follow my work) you quickly learn that I am a typography geek. I always have been: the very start of my career coincided with the publication of Robert Bringhurst’s Elements of Typographic Style and I just devoured that book; I read it over and over again over the years. Not that I have ever identified as a designer: I don’t draw, and my ‘eye’ for composition is pretty rudimentary. I have, however, spent many years in the company of serious book designers: Ron Woodward, Roberto Dosil, and most recently Mauve Pagé; and I’ve learned a lot watching these people work and cribbing from them. More accurate to say I am a type historian, as my interest runs much more to history and the historical forms than to the latest and hottest. A giant part of my interest in SFU’s Aldine collection has to do with the watershed typographic moments on display in those books.
I have also spent many many years trying to find and champion open alternatives to Adobe’s grip on the world of design and printing. Adobe has been an enormously important company: shepherding “desktop publishing” into mainstream practice and becoming the pre-eminent font foundry (term used ironically) of the digital age. But it’s not the same nowadays; back in the 1990s Adobe won big by producing absolutely stellar design software and building a gigantic type library: they made all of that accessible to average Joes like me. But now, in the 21st century, they’ve largely become a massive data-extraction Internet company, a licensing agency (monthly subscriptions for the tools of your trade), with an incomprehensible and sinister embrace of AI. For a long time I’ve been working on how we do digital typography and publishing without Adobe. It’s an uphill battle… one does not simply reverse-engineer the Adobe Suite and release it free on the Internet.[1] Besides, there are more interesting and arguably better paths. That led me to push at decent web typography with CSS, and more recently, typesetting with Typst.
A few years back, I stumbled upon EB Garamond, a free and open-licensed font developed in the early 2010s by Georg Mayr-Duffner. It’s an extremely high-quality font, drawn from a 1592 specimen of Claude Garamond’s Roman and Robert Granjon’s italics. The font was subsequently rebuilt as a “variable font”[2] by Octavio Pardo with support from Google. The current version has a fabulously robust set of features, variants, and language support, and I think it nicely rivals Robert Slimbach’s Adobe Garamond, the world-beating reissue from the 1990s. Did I mention that EB Garamond is free-licensed? I used EB Garamond as the type for this blog since its launch in 2023 – albeit without half the features (e.g., old-style figures) engaged.
When we were building the initial website for Aldus@SFU I really wanted a nice Aldine-derived Roman for it. The mainstream font that fits this bill is Bembo, reissued by Monotype in the 1920s, used in several hundred Penguin novels, and nowadays available through Adobe, of course. There is also Poliphilus, which is more historically accurate but nowhere near as popular (also released by Monotype in the 1920s). But by chance (and a student paper) I discovered Cardo, an open-licensed redesign of Bembo by David Perry, and that led me to Fbb – an expansion of Cardo by Michael Sharpe – and from that to ETbb, yet another redesign of Bembo originally done by designer and chart-master Edward Tufte and later revised and expanded by none other than Michael Sharp and others. It is the best yet! We used ETbb for Aldus@SFU.
Where do all these free re-designs come from, you might ask – especially if you work primarily within the Adobe world. The answer is from the TeX community – the people who work with LaTeX (and other variants) to do fancy mathematical typesetting. This community has been around almost forever, and they occupy little-known (at least to outsiders) corners of the web which reveal treasures if you dig into them. The TeX community have always been about free software, and their approach to fonts is no different.
Notice what happens when you free-license a typeface: over a decade or two, half a dozen people make contributions to it, and you end up with something extremely good! And possibly even better next year… I started watching the Comprehensive TeX Archive Network, which hosts all sorts of TeX-based typographical goodies, including hundreds of fonts – all with really superior math support in them, because that’s what TeX people care about. But it’s worth noting that these fonts are released in standard TTF and OTF formats, so they work perfectly well in whatever software you’re using.
Once I had opened the door to this universe, I began to have serendipitous discoveries. Late this summer, a friend forwarded a post to me about Baskerville Punch, a stunning archaeological project at Cambridge to analyze and 3d-model the actual 18th-century punches of John Baskerville, widely hailed as the king of English type designers. The four fonts that have come out of this project are absolutely stunning: four distinct cuts (different weights for different sizes) of Baskerville’s type. For many decades versions and re-releases of Baskerville have been on the market – one of them (I believe it’s ITC New Baskerville) had the honour of being named the “typeface of truth” by Errol Morris – but this new release is four distinct cuts in one open-licensed package. Sorry, 20th-century foundries, but no one needs your Baskerville variants anymore; this project trumps them. I couldn’t believe my luck.
Having now acquired high-quality, open-licensed versions of the Aldine, the Garamond, and now the Baskerville, I started wondering: Is there a free Caslon out there? Sure enough, there’s a “Libre Caslon” available on CTAN, but it’s nothing to write home about: compared to Carol Twombly’s masterful 1990s redesign for Adobe, Libre Caslon doesn’t really rate. But wait…
Again serendipity. Reading the Wikipedia page on the “Long S” ( as in “Where the bee ſucks, there ſuck I…”) I noticed an example image from a font called “Junicode” – a weird name to be sure. I clicked through and discovered, again, treasures. Junicode is the Unicode version of a project extending way back (to the 1990s) by Peter S. Baker, a medievalist at Virginia, who found an old font used by Oxford University Press – a font that included Old English glyphs – and digitized it for his medievalist colleagues (it was originally named “Junius” after the person who commissioned the Saxon additions for the press). Again, a stunningly high-quality, open-licensed font, developed and enhanced over decades!
Junicode is not Caslon, but it’s reasonably contemporary, and frankly, side-by-side with Adobe Caslon it’s very close. A few telltale signs (the capital ‘T’ especially) will give it away, but for text setting, I see this as an absolutely suitable replacement… in fact, with slightly shorter capitals (Adobe Caslon’s capitals are taller than its lowercase ascenders) and a slightly less slanted italic, it might even be nicer; I’m not sure yet… it’s close, anyway. Plus, Junicode has more language support than you can imagine (I mean, Old English and two complete Greek fonts, Runes, Saxon, plus, plus…), all accessible as a open-licensed, variable font! Junicode may not be the best webfont due to their heft (these are fully ten times the filesize of the Baskerville Punch fonts), but for print/PDF work, this one is gold!
Concerned about copyright? How does the designer get paid? Let’s be serious here. Francesco Griffo (of the Aldine Press) died over 525 years ago. Claude Garamond, who died 450 years ago, slavishy imitated Griffo (albeit with celebrated improvements). William Caslon died 250 years ago, his designs copied by many during his lifetime.[3] John Baskerville died about ten years after that. So I think these things are unambiguously in the public domain. The closest thing to a claim might be Monotype for Bembo, but that was 1928! Vive les revivals libres.
Fonts mentioned here:
ETbb by David Perry, Michael Sharpe, and Edward Tufte
EB Garamond by Georg Mayr-Duffner, Octavio Pardo
Junicode by Peter S. Baker
Baskerville Punch by a big team at Cambridge & beyond
Er, except somebody apparently did that with “AI” in October of this year. Jury’s still out on whether it’s crap or not. ↩︎
In a Variable Font, instead of separate files for regular, bold italic, condensed, expanded, and so on, you get one (or two) larger files, and you can then control things like weight, horizontal stretch, and other design features numerically in any OpenType-compliant application (which, in 2026, should be anything you’re using). So weight is 400 for ‘regular,’ 600 for ‘semibold,’ 700 is ‘bold,’ but you can dial in 479 if you like. ↩︎
Typographic historian Ricardo Olocco, who has done more in the twenty-first century to examine early type design, type trading and markets, and type copying notes that “In their 1785 specimen, the Fry foundry of London claimed to have imitated Caslon’s types so exactly that they could not be distinguished from the originals.” (Olocco 2020 p84) Think of that next time you drink cocoa (yes, it’s those Frys) ↩︎