Thomas Phinney

So, lots of pressure here on Affinity to do this. I just want to comment a bit on something. And please note, this is from the perspective that I personally love variable fonts, have made a bunch of them, and can’t imagine personally paying to license any general design tool today that does not support them decently. Variable font support is NOT trivial It isn’t just a matter of enabling something. And no, switching out their back-end font engine for Harfbuzz would not be trivial, and even if it were, that would not solve the problem! Don’t get me wrong, either supporting Harfbuzz or supporting variable fonts in their own engine is a huge amount of work. So many assumptions need to be thrown away, in either case. APIs reworked... agh. But the other huge deal is: user interface aspects of variable fonts are a huge hassle. The names and contents of axes are completely arbitrary and not known in advance by app developers. What is the most axes a font can have? 64K. Maybe the app has some arbitrary limit lower than that. How fine should the stops be on an axis? How many predefined instances should they support per axis? How do they expose predefined instances for variable fonts and how do they make those interact with user-defined instances? Do they just use sliders for everything? Do they support decimals in axis positions or only whole numbers? Manual input of position coordinates, or only the sliders? And on and on. Without going into the same detail, impacts on layout are similarly huge. Writing this as former fonts product manager at Adobe (until December 2008), later product management for font management and web fonts at Extensis, and CEO of FontLab Anyway, I am eager to see this support in Serif’s Affinity products, but please understand it is a heck of a lot of work! (Is that enough italics and boldface for one post, or should I throw in some more gratuitously here?)

You haven’t heard the story before, because it is absolutely untrue. First, Adobe stopped making new fonts in Type 1 format very early in their switch to OpenType. Before they even shipped their first retail OpenType fonts in the year 2000. That included Type 1 MM fonts. Originally Adobe intended to have and ship OpenType Multiple Master fonts among their first OpenType fonts. The format was fully spec’d and fonts were being developed. Dan Mills at Adobe decided to drop the MM part for two reasons: 1) Microsoft had zero interest in MM OpenType at the time, and was proving a reluctant partner on that part of things. But that was not the same as pressure, and there was no quid pro quo that I ever heard at the time. 2) OpenType layout features and the rest of it was a hard enough sell without the MM part at the same time. Dan was concerned OpenType might not succeed if MM was bundled in there from the beginning. Some of us at Adobe were pretty unhappy when Dan told us (the type team) of his decision (in 1999 IIRC?). David Lemon and I talked about it at great length. But as much as we loved MM (and love variable fonts now) we were not convinced he was wrong. I believe I recall saying exactly that to David, and him agreeing, before we even left the room. I remember David Lemon and I talking about how maybe some day we could bring back MM in OpenType. It took 17 years, but it happened at ATypI in Warsaw in 2016. And that time around, Adobe (David), Microsoft (Greg), Apple (Ned) and Google (Behdad) all did it as a joint announcement—although Apple was coy about committing to support for the new stuff, as it was based fairly directly on the GX Variations tech, it seemed pretty clear they were down for it.