Over 400 people attended the 43rd ATypI meeting, which was only the third ATypI meeting in North America. It took place at the Bayside Expo, on the south side of Boston, from 7-10 October 1999. Although many people helped out with the organization, there is no doubt that the contents and spirit of the meeting were shaped by Matthew Carter and David Berlow, who introduced all the invited speakers and moderated all sessions that required moderating.
I was ecstatic after the first full day, Friday, because of the mix of subjects and the quality of the presentations. The second day, in contrast, had a few peaks, but a session called "Design extravaganza" spanning most of the afternoon, was a disaster, and even the organizers seemed to show some polite embarrassment (the last three speakers had no message, went over time, two of them were merely reading from prepared notes, and the last speaker managed to insult Gerard Unger and Tobias Frere-Jones by downplaying type revivals and questioning measures of readability). The specter of poorly scheduled parallel events was expertly avoided by careful planning, often pitting a technical session against an artsy presentation. As a result, the pace of the conference and the flow were smooth and consistent.
Meetings should be learning experiences, sprinkled with entertainment. And learn, most attendees did, from the main speakers such as Ed Tufte, the keynote speaker, Hrant Papazian, Gerard Unger and Lucas de Groot. Entertainment, boundless entertainment, was provided by Ed Benguiat, but there was also plenty of it, in a more subdued form, in the presentations of Roger Black and Ed Tufte.
As usual, the attendees got a lot of handouts, catalogs, locked font CDs and posters. There were beautiful fonts on display at the meeting, mostly the results of the Kyrillitsa 99 cyrillic type competition. The new wave of young Russian designers got well-deserved exposure: Illarion Gordon (who made the daring pLatinum font), Kyrill Sirotin (author of the more-art-than-font Pupygi and the related dingbat font Rybizma), Dmitry Kirsanov (doodle font Mas d'Azil), Lev Alborov (who designed Ger), and Serge Pichii (author of the dingbat font ITC Ancestor). Zapfino, ITC Japanese Garden (by Akira Kobayashi) and Risso Light (a great monospace font by Eric Fowles) rounded out the displays.
Let me first note the things I missed because of the scheduling or other commitments: I missed all films in the "film festival", I did not attend the presentations by John Benson, Michael Harvey, Scott Stowell, and Erik Spiekermann, I did not go to the auction, a 3-hour event given a life of its own by auctioneer extraordinaire Dave Farey, and I did not participate in the gala dinner. The dinner brings me to one of three sticking points. The conference fee was simply outrageous. Early prepayment was still a whopping 365 pounds, roughly 640 dollars. Later registrants paid up to 680 dollars. The income from fees was thus in excess of 260,000 dollars. Let us take away about 25,000 dollars for the coffee, cookies, lunchboxes, and room and equipment rentals. That still leaves about 230,000 dollars for paying invited speakers. So, I am led to conclude that this was a charity event to support the year-round expenses of ATypI, and to pay a few salaries. I saw a few designers in the audience for whom the fees must have been a real sacrifice. The second, more personal point, are the badges. It was a shame that the type, of all things, on the badges was too small to be readable from more than about 2 feet away, forcing people to stick their noses almost in the armpits of others. I was looking for certain people, and never found them in the crowd. So, please, next time, make the badges huge and use Jonathan Hoefler's "sumo" weights. The attendees were given a room with about 25 iMac computers. However, none had even the basic software to dial out by telnet, so many were shut off from our computers at home. The presentation of Roger Black would have been greatly improved had he been given access to the web from his computer. After all, his talk was about type on the net.
Finally, one of the great things about such meetings is the possibility to interact with the great minds in the field. Due to poor eyesight, I had to sit in the front row (in fact, the front row is wonderful at any event, try it!). I found myself either sitting besides the moderator, Matthew Carter, or upcoming speakers. During most of the second day, I was seated next to Don Knuth. We concluded that we probably were the only two mathematicians-turned-computer-scientists at the meeting, and no, we did not talk about mathematics. I complained to Don, and he agreed, that the square root in TeX is a typographic headache, and he explained some of the intricacies of its positioning. It is interesting to see how dress codes vary from society to society. While Don and I would fit in nicely in any mathematics meeting or any Salvation Army get-together, most male attendees at ATypI had the leather jacket and all-black outfits that are de rigueur at type meetings. In any case, the social aspect of such meetings as ATypI is not to be underestimated, and is undoubtedly an essential part of its huge success.
The keynote address by Ed Tufte was about a wonderful success story, his own life. He gave everyone a beautifully manicured poster that had all the things he wanted to talk about. No slides, movies, computers, or other paraphernalia. The talk was all Ed, speaking with his face and hands, and grabbing the attention like no one at this meeting. His deep soft voice gave his presentation a touch of class. Ed is the author of three books on the visualization of information. His first book, The Visual Display of quantitative Information, published in 1983, sold over 400,000 copies.
In a stroke of genius, he illustrated his main points about visualization with a 1869 drawing by French engineer Charles Joseph Minard that tried to explain the loss of life in Napoleon's army when it marched in 1812-1813 to Moscow. Of 422,000 soldiers, only 10,000 survived. Map, time-line, winter temperatures, river crossings, and numbers of casualties are all explained on this wonderful drawing that, without showing blood, and without a single line of text, is a wonderful testament against war. Good design is like clear thinking made visible.
The best a design can do is not screw up the information content. Stressing that in information display, you want to be self-effacing, Ed went on to show examples of bad design from the Northern German Renaissance to JPL/NASA fly-overs of Venus, to a squarish 1250 map of England (the island would have been longer if only the page would have permitted), to Galileo's 1613 text about the rings of Saturn with in-text figures (Saturn, a drawing, a word, a noun. The wonderful becomes familiar and the familiar wonderful).
Ed had few good words for web site designers (if your web site looks like a web site, you're in trouble), who usually clutter their pages with debris related to the operating system, commerce, and frivolous matters. He illustrated this with his design of a touch-screen information kiosk at the National Gallery. Of all the speakers at the meeting, Ed was the only one who gave us a permanent reminder of his talk in the form of his poster. It was perfection with a capital P.
Bubbly and always smiling, Roger Black is a household word in typography and publishing in the United States. In his talk, Roger complained that type on the web was "stuck". He stressed the need for font embedding and tried to allay the fears that embedding type and sending fonts with HTML files would unduly increase file sizes, citing his success with the web site he did for MSNBC. An argument in favor of font embedding is type branding---associating a particular type with a company or organization for easy recognition. He decried the slowness of w3c with respect to the introduction of SVG (scalable vector graphics), a technology that allows to have scalable fonts that are searchable (note; fonts in GIF format are not searchable!). XML, the extensible mark-up language, will according to Roger, give some of the type control back to the typographers.
Hrant Papazian is a Los Angeles-based specialist in non-Latin alphabets, but above all, he is a thinker. He explained his approach to alphabet reform. Imposing orthographic changes from the top-down is clearly slow and difficult, but if characters are changed in fonts by pushing readers gently, not too fast, to their limits, progress may be made. Hrant is concerned with type reading comfort, and argues on the basis of physionomic factors. He stressed the importance of outlines of words, and explained the difference between readability (a measure of deciphering word shapes) and legibility (a measure of deciphering letters).
Hrant proposed to make the basic lower case shapes as different as possible, in a process of slow deconstructivism. Starting from his work chart of 26 letters (his "conflict diagram"), grouped by "arch", "diagonal", "descenders", "ascenders", "v-group", "o-group", "i-group", with edges between close shapes (such as h and n), he showed how Lucida Book can be altered (in his MasLucida font) to make all glyphs more diverging. This is achieved by extra arches, tilting of strokes, additional nicks, lower case versions of capitals, and replacement of curves by angled lines. In this reform, he argued that we should work at the expectation level of the readers, and that the future is in the design of words, as the eye recognizes words in its first gaze. He ended by proposing the formation of an ATypI subgroup for alphabet reform. This talk was by far the most packed with fresh and interesting ideas.
According to Gerard Unger, William Eddison Dwiggins (1880-1957) was the greatest US type designer who ever lived. Dwiggins spent most of his life in Hingham near Boston, and was a world-renowned typographer and puppeteer. A prankster, he often wrote under the name Dr. Herman Püterschein and called his wife (Dorothy Abbe) Elza Püterschein. Dwiggins, who designed 18 faces, including the beautiful Caledonia, drew inspiration from many sources, including his designs of marionettes, and geometric shapes in general. Unger showed drawings such as Dwiggins's "geometric spinach". But above all, Unger underlined his love of type by concentrating on minute but important details in Dwiggins's designs and work methods. For example, as Unger discovered after a visit to the Boston Public Library where Dwiggins's work is exhibited, Dwiggins used to work on transparent paper that he would flip over and over, erasing previous designs to converge to final letter forms. Unger was quite surprised to discover that this technique used by himself had actually been around for such a long time. A great historical perspective, delivered by Unger in a classy low-key style.
Matthew Carter's introduction summed it up: Ed Benguiat created many ITC faces. You can't possibly categorize them. Ed is his own category. Speaking with a colorful New York (Brooklyn?) accent, and carrying a magic wand, Ed started off by showing his baby pictures and his place of birth. In true Jackie Gleason style, he delivered what was undoubtedly the funniest speech at any of the ATypI meetings ever held. He took us through a set of slides, each one of them leading up to a remark or story. His talk took us from BC (before computers) to his present type work. At one point, he showed the alphabet in a display. Noticing an omission, someone once asked Ed where the W was. Ed: None of your business. In any case, Ed is convinced that one should show a font in action, not in A-B-C-D format, which is like showing spaghetti in a box. This led him to lines like In New York, we put ketchup on pasta. Still, to cover his back, he added that his mother was a good cook. He also explained his rule of thumb: Never wear anything twice. He went on to explain why he will shave his mustache soon, just before a guy he helped send to jail for 8 years will be let out. Interestingly, that case hinged on a fake document typed with an IBM Souvenir ball, which was introduced, as Ed pointed out to the judge, years after the alleged date of the fake letter. After explaining why the New York Times logo is crooked on their trucks, and why no one wants an italic version of the Pope's cross, Ed blessed the audience with his magic wand, and left the room.
John Maeda is the Sony Career Development Professor of Arts and Sciences at the MIT Media Laboratory. He showed up with about 3 computers and a gaggle of graduate students, and dazzled the audience with demos of small computer programs and a few videos. Especially the lightning speed of the transitions was amazing. No fumbling here with set-ups and reboots---this was a flawless computer display. Themes included letters set to music, the sound created on the basis of letter shapes (for example, to simulate a scratchy pen), an original set of moving screen clocks, highly original letter animations, and a fantastic typewriter sound piece. Young, fresh, dynamic, something different.
At, noon there were "Master Meetings" in small rooms at the conference center. I missed the first such meeting on Friday because I mistakenly thought that the meetings were going to be held in the main room, where, in hindsight, they should have been held, given their enormous popularity and intrinsic interest. So, on day two, I managed to catch Lucas de Groot, the prolific Berlin-based Dutch designer of the Thesis family, which includes TheSans, TheMono, and many other subfamilies. Now, there is also TheAntiqua.
Lucas showed us how he managed to design all those multi-weight multi-style families. His main tool to date still is the 10-year old program Font Studio. He works a lot with interpolations and extrapolations of families based on certain formulas he deduced. He stressed the importance of having many sets of figures and of having small capitals besides the small caps, because languages like German use capitalized nouns and thus require capitals that are not too large. The many weights are necessary when one flips from dark to light in headlines. Also, bold characters appear smaller because their white spaces are smaller, and this should be compensated for by slightly larger x-heights. The use of a heavier weight for footnotes was also discussed, as well as the need to keep the percentage of ink (the "gray level") approximately constant when designing extended and condensed versions: thus, vertical strokes in extended fonts should be thicker.
Finally, Lucas delighted us with samples of his contracted work, such as FolhaSans (for the Folha Sao Paulo newspaper), SunSans and SunSerif (for Sun Microsystems), and a Volkswagen commission. Deploring the disappearance of multiple master fonts, Lucas admitted to not having used that technology in the first place. This session was almost like seeing Lucas at work at his terminal through a web-cam. A great view into the life of a humble and productive young man.
The conference was co-sponsored by Bitstream, Agfa-Monotype, Adobe, ITC, Linotype, Pyrus, Font Bureau, Carter&Cone, Textmatters, and Galapagos Design. The font production sessions moderated by David Berlow had technical panels that discussed various tools under development. The panels included Aiman Copty (Microsoft), David Lemon (Adobe), John Collins (Bitstream), Just van Rossum (Letterror, RoboFog Inc), Peter Lofting (Apple), Yuri Yarmola (Pyrus), and Greg Hitchcock and Simon Daniels (Microsoft). The companies presented their products and projects, but there was not enough time for questioning, unfortunately. The general feeling is that we are at a crucial point in type production and development, the main driving force being the introduction of OpenType as an umbrella format for truetype and type 1. Without exception, all companies are struggling with the problem, and try to cure the lack of support for OpenType.
OpenType works on the basis of tables. These tables are sometimes meant for the rasterizers (screen or printer) such as a table describing the outlines, a table controling the use of anti-aliasing and/or hinting, and a table describing the hinting. Other tables are meant for management: these have names, keywords, copyright information, a version history, panose numbers and the like. Still other tables will optionally be used by applications: these include character mapping tables (UNICODE support tables), metric tables (so that we can reserve boxes on output devices), ligature tables, tables for old style figures and small caps maps, tables with alternates, and additional tables for connected or multi-directional writing.
It has become frustratingly clear that each font is a complex beast, and as a result, very few full OpenType fonts are available: Adobe's Pro font series covers most european languages and has support for ISO-Latin 8859-1 and 8859-2, with a rich set of figures, support for the Euro, a multitude of symbols, and support for digital signatures that leave one's mark in the font. Nearly finished are Tekton Pro, Adobe Garamond Pro, Myriad Pro, and Minion Pro. For example, Minion Pro comes in 3 weights, 2 widths and 4 optical axes, and Myriad Pro has great Greek and Cyrillic versions. Adobe will produce more such fonts in 2000, and says that eventually, all fonts in their library will be in OpenType format. Most fonts shipping with Windows 2000 will be OpenType. Adobe's next version for ATM will recognize OpenType, and OpenType will be native in Windows 2000. So, what will we find at the various companies to make our life more pleasant?
Two more announcements rounded out the technical sessions. John Collins from Bitstream was raving about "MyFonts.com", which he likened to the amazon.com of fonts. This will be a warehouse where you can browse for fonts, test drive a font, buy fonts, get referrals to fonts, search for fonts by shape, kind, name, designer, or any characteristic. An automatic font matcher is included at this web site that is still in beta format (user: beta, password: letmein). To test drive, you will receive a collection of GIFs. David Berlow thought that doing this was a bit risky. John Collins asked the audience to send them CDs and fonts to be entered on the site.
Finally, the disappearance of Adobe's Multiple Master format created a few worries. However, Adobe announced that they will publish a tool that will allow users to create instances in type 1 or OpenType format from any multiple master font.
The open forum on legal issues focused on the details of copyright, trademark and patent. The most expert person there, at least for USA-realted problems, was Frank Martinez, a trademark lawyer, who in a previous life was a patent examiner. Active participants were Bruno Steinert (Linotype), Ira (Monotype), Dan Mills (Adobe), Erik Spiekermann (FontShop), and to a lesser extent, Hrant Papazian and Clive Bruton. It was more an information session rather than a brainstorming session about the future.
On trademarks, it was pointed out that in the US, one can do his/her own trademark search on the web---there is a URL with information that is up-to-date to about 3 months ago. This saves a lot on the costs, so that filing for trademark should be well under 1000 USD. In Europe, a EU search costs 2500 USD according to Steinert. In Europe, one can just register for 1200 USD plus attorney's fees, and worry about violations if and when they appear (dixit Steinert).
On copyright, there was discussion on what to submit. Frank Martinez deplored the US position on copyright, and said that the only protection of fonts there is through design patents. Cynthia (?) explained that the copyright decision in 1976 by the US Congress (refusing to address the copyright issue for fonts) was because fonts were considered industrial designs, not copyrightable in the US because they are "incremental improvements" (as opposed to creations). Bruno Steinert added that copyright protection in Germany is not as good as people think. For example, several years ago, a German Court found Candida too close to the standard alphabet to be protectable. As a result, he said, only calligraphic and script fonts can go to court for copyright protection. In the UK, where artistic overhead is not such a requirement, copyright protection is the strongest.
Finally, the forum considered patents. After a general introduction by Frank Martinez, Hrant Papazian complained that in the USA, there is a huge number of frivolous patents, and that patents should be for concepts. Someone else disagreed with this view, and Frank Martinez stated that the font people in the USA were forced to seek patents because of the 1976 decision on copyright by the US Congress. Don Mills clarified a bit later that the infamous SSi case (Adobe vs SSi) was in fact done on the basis of patents. Frank Martinez added that in any case, judges love documention in the form of trademarks and patents, and to be safe and successful in case of an injunction, he urged everyone to file for everything in sight.
Luc Devroye (copyright)
School of Computer Science
McGill University
Montreal, Canada H3A 2K6
luc@cs.mcgill.ca
http://luc.devroye.org
http://luc.devroye.org/fonts.html