August 2, 2016 is the official release date for Microsoft’s Windows 10 Anniversary Update (aka Redstone or RS1). Although I do not use Windows OS, I am jumping for joy, for the benefit of those who do use this modern and world-class OS.
Thanks to our friends at Microsoft, the DirectWrite that ships with the Windows 10 Anniversary Update supports OpenType/CFF Collections (aka OTCs), such as those deployed as part of the Adobe-branded Source Han Sans and Google-branded Noto Sans CJK open source projects, to include their all-inclusive “one font to rule them all” Super OTCs.
This will be a short, sweet, and to-the-point article. Sorry, no graphics nor photos.
When developing name-keyed fonts, glyph names matter. They matter a lot. When developing new fonts, the glyph names should either be explicitly listed in AGLFN (Adobe Glyph List For New Fonts) or derivable via the AGL Specification. Glyph names that adhere to AGLFN or the AGL Specification result in fonts with well-formed 'cmap' tables, which means that their glyphs will behave better in a broader range of environments. I cannot stress the importance of this.
CIDs (Character IDs), on the other hand, represent a completely different beast. If a font is genuinely CID-keyed, it means that there are absolutely no glyph names, regardless of whether the source font or fonts that were used to build the CID-keyed font were named-keyed. Once a font resource becomes CID-keyed, the original glyph names are literally jettisoned, and the only way in which to map Unicode values to glyphs is via the 'cmap' table, which is usually done using a UTF-32 CMap resource. In other words, when developing fonts that are intended to be deployed in a CID-keyed fashion, the source glyph names play absolutely no role in how such fonts are processed.
By ESO/José Francisco Salgado (josefrancisco.org) — ALMA antennas under the Milky Way
Five years ago, I wrote this article that described how to manage XUID arrays. Then last year, I wrote this article that suggested that XUID arrays are no longer necessary.
Anyway, there are two messages that are being conveyed in today’s article.
The first message is short and sweet, and meant to be strong: Adobe advises against including XUID arrays in all new and updated font-related resources, meaning fonts themselves and their corresponding CMap resources. The good news is that omitting the XUID array represents one less thing to worry about during font development.
The second message is longer, meant to provide some background information, and describes why Adobe advises against including XUID arrays in font-related resources.
One of my more popular open source fonts is Adobe Blank, and to a less extent the related Adobe Blank 2 because it uses a 'cmap' table format, Format 13, that is not broadly supported. Actually, Adobe Blank provides absolutely nothing, because it maps all 1,111,998 Unicode code points to a range of 2,048 non-spacing and non-marking glyphs, yet such a font is useful for particular scenarios, such as addressing the FOUT (Flash Of Unstyled Text) problem.
Allow me to introduce Adobe NotDef, which is modeled after Adobe Blank in that it covers all of Unicode and maps to a range of 2,048 glyphs, but differs in that the functional glyphs are spacing and marking. The original suggestion for Adobe NotDef came from Dave Crossland. The glyphs match the shape and advance width of the standard Adobe .notdef glyph that is invoked in environments that do not support font fallback when the selected font does not include a glyph for a particular character, and as Dave wrote, Adobe NotDef is useful for font fallback purposes in that it can be used to prevent the display of non-standard .notdef glyphs that may be present in some fonts in the font fallback chain.
It seems that I am on roll, having released two new open source fonts on GitHub within the past week. The previous—and brief—article that was about the LOCL Test OpenType/CFF font simply pointed to the repository. This article will be longer. I promise.
Inspired by the font that I prepared for and referenced in the previous article, I decided to launch a dedicated open source project for this useful test font, LOCL Test.
Although this article shares its title with an article from four years ago that was about the excitement associated with attending ATypI Hong Kong 2012, this particular one will focus on efforts to properly support Hong Kong SAR (aka HK or Hong Kong) in the Adobe-branded Source Han Sans and Google-branded Noto Sans CJK typeface families, but also in infrastructure, such as OSes and apps.
In other words, this article is not about traveling to Hong Kong, but rather about properly supporting Hong Kong in OSes, apps, and fonts.
A peculiar series of events that took place on April 1st (no joke) and 2nd of this year led to the discovery of what can only be described as somewhat of a revelation: A small number of CJK Compatibility Ideographs are necessary for China. This is important, because I made the following statement on page 168 of CJKV Information Processing, Second Edition:
One of the most powerful font-development tools available today is tx (Type eXchange), which is included in AFDKO (Adobe Font Development Kit for OpenType) and whose sources are available on GitHub. Despite its two-letter name, this command-line utility is packed with an enormous amount of features and functionality.
Four years ago I wrote a similar article, but it seems like a good time to revisit tx and the useful things that it can do. I still recommend that its “-u” and -h” command-line options be used to explore its vast capabilities.
—Humans make mistakes—
—Anything made by humans has the potential to include mistakes—
The most important things about mistakes are that 1) we recognize them, lest they propagate; 2) we learn from them; 3) we make an effort not to repeat them; and 4) we try to fix them, if possible.
Some mistakes are more easily fixed than others. Mistakes that cannot be fixed must be worked around.
With that said, an interesting event of historical significance occurred in June of 2000:
The first version of the IVD (Ideographic Variation Database) was issued on 2007-12-14, meaning over eight years ago, and there have been three subsequent revisions, the latest being issued on 2014-05-16. There are currently three registered IVD collections: Adobe-Japan1, Hanyo-Denshi, and Moji_Joho. A significant number of IVSes are shared between the latter two IVD collections, 9,685 to be exact. While I cannot speak to the latter two IVD collections, the Adobe-Japan1 one is supported by hundreds of OpenType fonts via the Format 14 (Unicode Variation Sequences) ‘cmap‘ subtable. Furthermore, the number of apps and OSes that support UVSes has reached critical mass.
With all that said, there is a rather substantial missing link in terms of IVD support infrastructure: the all-important input method.
The next UTC (Unicode Technical Committee) meeting, the 147th one, takes place during the week of May 9th, and will be hosted at the Adobe headquarters in San José, California. All members of the Unicode Consortium, especially voting members, are encouraged to attend.
Much of the thinking that I did with regard to this unregistered—but hopefully soon-to-be-registered—IVD (Ideographic Variation Database) collection was done while visiting my parents in South Dakota, with one of the highlights of that trip being a scenic drive through Badlands National Park.
First and foremost, please forget, or at least ignore, most everything that was written in the 2016-02-13 and 2016-02-20 articles (which makes one wonder why I am linking to them, but I digress). Far too many things have changed, and what I present in this article represents the IVD collection that I hope will be registered later this year.
Continuing where I left off with the first article about this subject, I’d like to point out some of the implementation details and their ramifications in this article.
One of my longer term goals for the open source Source Han Sans project has been to eventually register a Pan-CJK IVD (Ideographic Variation Database) collection that would allow the regional variants to display and be preserved in “plain text” environments, and I think that I may have achieved a breakthrough the other day.
One of the fringe benefits of moving offices—especially when one has accumulated nearly 25 years of font-related material and it is thus not a pain-free exercise—is discovering historical documents, some of which turn out to be true gems. Our team is preparing to move from the Adobe East Tower to the West one, and part of the process is figuring which material to keep, and which to put into File 13. Anyway, I had been recently looking for a particular presentation that I prepared many years ago, and was fortunate enough to come across it while sifting through my accumulated materials.
In late 2015, I collaborated with Daisuke MIURA to submit a proposal (L2/15-328) to the UTC (Unicode Technical Committee) to encode the characters for four tally mark systems. The proposal was discussed during UTC #146, and the result was that the five ideographic tally mark characters were accepted. Good news.
The Script Ad Hoc Committee originally recommended in their report for UTC #146 (see page 9 of L2/15-037) that IDEOGRAPHIC TALLY DIGIT TWO not be encoded, because they felt that it could be unified with U+1D36E (COUNTING ROD TENS DIGIT SIX), but concerns over typographic consistency led to it being accepted as a separate character.
By default, the AFDKO makeotf tool includes Macintosh (platformID=1, encodingID=0, languageID=0) ‘name‘ table strings, and if specified in the “FontMenuNameDB” or “features” files, localized Macintosh ‘name’ table strings will also be included. The next release of AFDKO will include “-omitMacNames” as a new command-line option for makeotf whose purpose is to exclude Macintosh ‘name’ table strings, other than any that are explicitly specified in the “features” file.
IUC39 (The 39th Internationalization & Unicode Conference) took place in Santa Clara earlier this week, and Adobe was once again proud to be a Gold Sponsor. It was another outstanding and successful conference, and as usual, one of the greatest benefits of the conference—besides the many excellent presentations—was the opportunity for face-to-face exchanges with Unicode leaders, experts, and enthusiasts.
(The introductory graphic illustrates how the character 剣 (U+5263) is displayed using the fonts that are introduced in this article. The code point for this character maps to a glyph that displays as “63” in the FDArray Test 257 font, which is the hexadecimal equivalent of the decimal index of the FDArray element to which its glyph is assigned, which is 99. Likewise, the code point for this character maps to a glyph that displays as “52” in the FDArray Test 65535 font, which is the hexadecimal equivalent of the decimal index of the FDArray element to which its glyph is assigned, which is 82.)
I have built several CID-keyed OpenType/CFF fonts that are specifically designed to test various limits, by exercising various implementation limits, such as the number of glyphs (65,535 is the architectural limit), the number of FDArray elements (256 is the architectural limit), and the number of mappings in the ‘cmap‘ table (when the surrogates and non-characters are factored out, Unicode has 1,111,998 possible mappings in its 17 planes). I have sometimes made these fonts available, such as in this May of 2012 article that explains how such fonts can be built.
Anyway, I spent pretty much all day yesterday—except for a somewhat longer than usual lunch break that was actually used to watch The Martian (2015) with my wife—preparing a pair of open source CID-keyed OpenType/CFF fonts that exercise these limits but to different degrees, and I also managed to prepare and release the project on GitHub as FDArray Test.